diff --git "a/perf-df-awq-1xA10.csv" "b/perf-df-awq-1xA10.csv" --- "a/perf-df-awq-1xA10.csv" +++ "b/perf-df-awq-1xA10.csv" @@ -1,15 +1,144 @@ -config.name,config.backend.name,config.backend.version,config.backend._target_,config.backend.task,config.backend.library,config.backend.model,config.backend.processor,config.backend.device,config.backend.device_ids,config.backend.seed,config.backend.inter_op_num_threads,config.backend.intra_op_num_threads,config.backend.model_kwargs.trust_remote_code,config.backend.processor_kwargs.trust_remote_code,config.backend.hub_kwargs.trust_remote_code,config.backend.no_weights,config.backend.device_map,config.backend.torch_dtype,config.backend.eval_mode,config.backend.to_bettertransformer,config.backend.low_cpu_mem_usage,config.backend.attn_implementation,config.backend.cache_implementation,config.backend.autocast_enabled,config.backend.autocast_dtype,config.backend.torch_compile,config.backend.torch_compile_target,config.backend.quantization_scheme,config.backend.quantization_config.bits,config.backend.quantization_config.version,config.backend.deepspeed_inference,config.backend.peft_type,config.scenario.name,config.scenario._target_,config.scenario.iterations,config.scenario.duration,config.scenario.warmup_runs,config.scenario.input_shapes.batch_size,config.scenario.input_shapes.num_choices,config.scenario.input_shapes.sequence_length,config.scenario.new_tokens,config.scenario.latency,config.scenario.memory,config.scenario.energy,config.scenario.generate_kwargs.max_new_tokens,config.scenario.generate_kwargs.min_new_tokens,config.launcher.name,config.launcher._target_,config.launcher.device_isolation,config.launcher.device_isolation_action,config.launcher.start_method,config.environment.cpu,config.environment.cpu_count,config.environment.cpu_ram_mb,config.environment.system,config.environment.machine,config.environment.platform,config.environment.processor,config.environment.python_version,config.environment.gpu,config.environment.gpu_count,config.environment.gpu_vram_mb,config.environment.optimum_benchmark_version,config.environment.optimum_benchmark_commit,config.environment.transformers_version,config.environment.transformers_commit,config.environment.accelerate_version,config.environment.accelerate_commit,config.environment.diffusers_version,config.environment.diffusers_commit,config.environment.optimum_version,config.environment.optimum_commit,config.environment.timm_version,config.environment.timm_commit,config.environment.peft_version,config.environment.peft_commit,report.traceback,config.backend.hub_kwargs.revision,config.backend.hub_kwargs.force_download,config.backend.hub_kwargs.local_files_only,report.prefill.memory.unit,report.prefill.memory.max_ram,report.prefill.memory.max_global_vram,report.prefill.memory.max_process_vram,report.prefill.memory.max_reserved,report.prefill.memory.max_allocated,report.prefill.latency.unit,report.prefill.latency.count,report.prefill.latency.total,report.prefill.latency.mean,report.prefill.latency.stdev,report.prefill.latency.p50,report.prefill.latency.p90,report.prefill.latency.p95,report.prefill.latency.p99,report.prefill.latency.values,report.prefill.throughput.unit,report.prefill.throughput.value,report.prefill.energy.unit,report.prefill.energy.cpu,report.prefill.energy.ram,report.prefill.energy.gpu,report.prefill.energy.total,report.prefill.efficiency.unit,report.prefill.efficiency.value,report.decode.memory.unit,report.decode.memory.max_ram,report.decode.memory.max_global_vram,report.decode.memory.max_process_vram,report.decode.memory.max_reserved,report.decode.memory.max_allocated,report.decode.latency.unit,report.decode.latency.count,report.decode.latency.total,report.decode.latency.mean,report.decode.latency.stdev,report.decode.latency.p50,report.decode.latency.p90,report.decode.latency.p95,report.decode.latency.p99,report.decode.latency.values,report.decode.throughput.unit,report.decode.throughput.value,report.decode.energy.unit,report.decode.energy.cpu,report.decode.energy.ram,report.decode.energy.gpu,report.decode.energy.total,report.decode.efficiency.unit,report.decode.efficiency.value,report.per_token.memory,report.per_token.latency.unit,report.per_token.latency.count,report.per_token.latency.total,report.per_token.latency.mean,report.per_token.latency.stdev,report.per_token.latency.p50,report.per_token.latency.p90,report.per_token.latency.p95,report.per_token.latency.p99,report.per_token.latency.values,report.per_token.throughput.unit,report.per_token.throughput.value,report.per_token.energy,report.per_token.efficiency,config.backend.quantization_config.exllama_config.version,config.backend.quantization_config.exllama_config.max_input_len,config.backend.quantization_config.exllama_config.max_batch_size -4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciCoder-1b,Deci/DeciCoder-1b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +config.name,config.backend.name,config.backend.version,config.backend._target_,config.backend.task,config.backend.library,config.backend.model,config.backend.processor,config.backend.device,config.backend.device_ids,config.backend.seed,config.backend.inter_op_num_threads,config.backend.intra_op_num_threads,config.backend.model_kwargs.trust_remote_code,config.backend.processor_kwargs.trust_remote_code,config.backend.hub_kwargs.trust_remote_code,config.backend.no_weights,config.backend.device_map,config.backend.torch_dtype,config.backend.eval_mode,config.backend.to_bettertransformer,config.backend.low_cpu_mem_usage,config.backend.attn_implementation,config.backend.cache_implementation,config.backend.autocast_enabled,config.backend.autocast_dtype,config.backend.torch_compile,config.backend.torch_compile_target,config.backend.quantization_scheme,config.backend.quantization_config.bits,config.backend.quantization_config.version,config.backend.deepspeed_inference,config.backend.peft_type,config.scenario.name,config.scenario._target_,config.scenario.iterations,config.scenario.duration,config.scenario.warmup_runs,config.scenario.input_shapes.batch_size,config.scenario.input_shapes.num_choices,config.scenario.input_shapes.sequence_length,config.scenario.new_tokens,config.scenario.latency,config.scenario.memory,config.scenario.energy,config.scenario.generate_kwargs.max_new_tokens,config.scenario.generate_kwargs.min_new_tokens,config.launcher.name,config.launcher._target_,config.launcher.device_isolation,config.launcher.device_isolation_action,config.launcher.start_method,config.environment.cpu,config.environment.cpu_count,config.environment.cpu_ram_mb,config.environment.system,config.environment.machine,config.environment.platform,config.environment.processor,config.environment.python_version,config.environment.gpu,config.environment.gpu_count,config.environment.gpu_vram_mb,config.environment.optimum_benchmark_version,config.environment.optimum_benchmark_commit,config.environment.transformers_version,config.environment.transformers_commit,config.environment.accelerate_version,config.environment.accelerate_commit,config.environment.diffusers_version,config.environment.diffusers_commit,config.environment.optimum_version,config.environment.optimum_commit,config.environment.timm_version,config.environment.timm_commit,config.environment.peft_version,config.environment.peft_commit,report.traceback,config.launcher.numactl,report.prefill.memory.unit,report.prefill.memory.max_ram,report.prefill.memory.max_global_vram,report.prefill.memory.max_process_vram,report.prefill.memory.max_reserved,report.prefill.memory.max_allocated,report.prefill.latency.unit,report.prefill.latency.count,report.prefill.latency.total,report.prefill.latency.mean,report.prefill.latency.stdev,report.prefill.latency.p50,report.prefill.latency.p90,report.prefill.latency.p95,report.prefill.latency.p99,report.prefill.latency.values,report.prefill.throughput.unit,report.prefill.throughput.value,report.prefill.energy.unit,report.prefill.energy.cpu,report.prefill.energy.ram,report.prefill.energy.gpu,report.prefill.energy.total,report.prefill.efficiency.unit,report.prefill.efficiency.value,report.decode.memory.unit,report.decode.memory.max_ram,report.decode.memory.max_global_vram,report.decode.memory.max_process_vram,report.decode.memory.max_reserved,report.decode.memory.max_allocated,report.decode.latency.unit,report.decode.latency.count,report.decode.latency.total,report.decode.latency.mean,report.decode.latency.stdev,report.decode.latency.p50,report.decode.latency.p90,report.decode.latency.p95,report.decode.latency.p99,report.decode.latency.values,report.decode.throughput.unit,report.decode.throughput.value,report.decode.energy.unit,report.decode.energy.cpu,report.decode.energy.ram,report.decode.energy.gpu,report.decode.energy.total,report.decode.efficiency.unit,report.decode.efficiency.value,report.per_token.memory,report.per_token.latency.unit,report.per_token.latency.count,report.per_token.latency.total,report.per_token.latency.mean,report.per_token.latency.stdev,report.per_token.latency.p50,report.per_token.latency.p90,report.per_token.latency.p95,report.per_token.latency.p99,report.per_token.latency.values,report.per_token.throughput.unit,report.per_token.throughput.value,report.per_token.energy,report.per_token.efficiency,config.backend.hub_kwargs.revision,config.backend.hub_kwargs.force_download,config.backend.hub_kwargs.local_files_only,config.backend.quantization_config.exllama_config.version,config.backend.quantization_config.exllama_config.max_input_len,config.backend.quantization_config.exllama_config.max_batch_size +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-1.4b,EleutherAI/pythia-1.4b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 64, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Salesforce/codegen-6B-nl,Salesforce/codegen-6B-nl,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: CodeGenForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-70m,EleutherAI/pythia-70m,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-110B,Qwen/Qwen1.5-110B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 32.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciCoder-1b,Deci/DeciCoder-1b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 68, in run report = scenario.run(backend) File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run _ = backend.generate(self.inputs, self.config.generate_kwargs) @@ -19,27 +148,27 @@ ChildProcessError: Traceback (most recent call last): return self.pretrained_model.generate(**inputs, **kwargs) File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1576, in generate - result = self._greedy_search( - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2494, in _greedy_search + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample outputs = self( File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl return self._call_impl(*args, **kwargs) File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1211, in forward + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1164, in forward outputs = self.model( File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl return self._call_impl(*args, **kwargs) File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1018, in forward + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 968, in forward layer_outputs = decoder_layer( File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl return self._call_impl(*args, **kwargs) File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 741, in forward + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 713, in forward hidden_states, self_attn_weights, present_key_value = self.self_attn( File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl return self._call_impl(*args, **kwargs) @@ -47,47 +176,93 @@ ChildProcessError: Traceback (most recent call last): return forward_call(*args, **kwargs) TypeError: DeciCoderAttention.forward() got an unexpected keyword argument 'cache_position' -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciLM-7B,Deci/DeciLM-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/recurrentgemma-2b,google/recurrentgemma-2b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/recurrentgemma-2b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aaa58-7af50ddb7799966c5ce4db38;be011b48-0af0-4651-b018-b39f726fad17) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/recurrentgemma-2b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ - self.load_model_with_no_weights() - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights - self.load_model_from_pretrained() - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained - self.pretrained_model = self.automodel_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 558, in from_pretrained - return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3544, in from_pretrained - config = cls._autoset_attn_implementation( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1463, in _autoset_attn_implementation - config = cls._check_and_enable_sdpa( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1619, in _check_and_enable_sdpa - raise ValueError( -ValueError: DeciLMForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/recurrentgemma-2b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-j-6b,EleutherAI/gpt-j-6b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,internlm/internlm-20b,internlm/internlm-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -95,28 +270,32 @@ ChildProcessError: Traceback (most recent call last): self.load_model_from_pretrained() File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained self.pretrained_model = self.automodel_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 558, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3544, in from_pretrained - config = cls._autoset_attn_implementation( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1463, in _autoset_attn_implementation - config = cls._check_and_enable_sdpa( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1619, in _check_and_enable_sdpa - raise ValueError( -ValueError: GPTJForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3626, in from_pretrained + model = cls(config, *model_args, **model_kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm-20b/80729bcf52fbc4553d965926b27304ac5e156d98/modeling_internlm.py"", line 906, in __init__ + self.model = InternLMModel(config) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm-20b/80729bcf52fbc4553d965926b27304ac5e156d98/modeling_internlm.py"", line 729, in __init__ + self.layers = nn.ModuleList([InternLMDecoderLayer(config) for _ in range(config.num_hidden_layers)]) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm-20b/80729bcf52fbc4553d965926b27304ac5e156d98/modeling_internlm.py"", line 729, in + self.layers = nn.ModuleList([InternLMDecoderLayer(config) for _ in range(config.num_hidden_layers)]) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm-20b/80729bcf52fbc4553d965926b27304ac5e156d98/modeling_internlm.py"", line 545, in __init__ + self.self_attn = INTERNLM_ATTENTION_CLASSES[config.attn_implementation](config=config) +KeyError: 'sdpa' -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-1.3B,EleutherAI/gpt-neo-1.3B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-2.7b,facebook/opt-2.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -126,26 +305,26 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3544, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained config = cls._autoset_attn_implementation( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1463, in _autoset_attn_implementation + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation config = cls._check_and_enable_sdpa( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1619, in _check_and_enable_sdpa + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa raise ValueError( -ValueError: GPTNeoForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` +ValueError: OPTForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neox-20b,EleutherAI/gpt-neox-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-125m,facebook/opt-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -155,57 +334,102 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3544, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained config = cls._autoset_attn_implementation( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1463, in _autoset_attn_implementation + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation config = cls._check_and_enable_sdpa( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1619, in _check_and_enable_sdpa + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa raise ValueError( -ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` +ValueError: OPTForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-34B,,cuda,0,42,,,,,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,main,False,False,MB,1721.061376,22129.672192,0.0,21483.225088,20799.036928,s,10,27.921223388671876,2.7921223388671876,0.002873024218980136,2.791891357421875,2.7956688476562497,2.7963155029296876,2.7968328271484375,"[2.793120361328125, 2.796962158203125, 2.790662353515625, 2.790180419921875, 2.7901220703125, 2.787002685546875, 2.79004345703125, 2.793321533203125, 2.795525146484375, 2.794283203125]",tokens/s,91.68652692484228,kWh,3.2928969429598913e-05,1.8046407099463975e-05,0.00015823184880759956,0.00020920722533666246,tokens/kWh,1223667.1060860218,MB,1726.52544,22129.672192,0.0,21483.225088,20902.142976,s,10,1662.0463125,166.20463124999998,0.013890741352160448,166.199171875,166.2239140625,166.22626953125,166.22815390625,"[166.211921875, 166.223390625, 166.191390625, 166.191359375, 166.191640625, 166.203515625, 166.194828125, 166.2176875, 166.191953125, 166.228625]",tokens/s,0.3790508093919315,kWh,0.0019620591764814327,0.001075380349219922,0.009375421333664404,0.012412860859365755,tokens/kWh,5075.381148131151,,s,629,1684.5351452636726,2.678116288177539,0.33164961442268687,2.637888427734375,2.63990244140625,2.640667578125,5.42840712890625,"[2.638138427734375, 2.638636962890625, 2.638865478515625, 2.6390283203125, 2.6389248046875, 2.63697509765625, 2.6374482421875, 2.637588623046875, 2.6383994140625, 2.637781982421875, 2.63811181640625, 2.639296630859375, 2.6380380859375, 2.637068359375, 2.638099365234375, 2.637255615234375, 2.637476806640625, 2.63743798828125, 2.637467529296875, 2.63781787109375, 2.637115478515625, 2.637315185546875, 2.63943896484375, 2.639993896484375, 2.6398291015625, 2.63933349609375, 2.63897802734375, 2.639859619140625, 2.63980126953125, 2.639171630859375, 2.637111328125, 2.63720556640625, 2.636434326171875, 2.63687158203125, 2.63707958984375, 2.639266845703125, 2.63845263671875, 2.637464599609375, 2.638060546875, 2.64020263671875, 2.638950439453125, 2.63784033203125, 2.63769091796875, 2.6381884765625, 2.639057861328125, 2.637621337890625, 2.638141357421875, 2.63823876953125, 2.639244384765625, 2.63790283203125, 2.6387138671875, 2.6390537109375, 2.638361572265625, 2.637244384765625, 2.639206298828125, 2.638740478515625, 2.639088623046875, 2.638509033203125, 2.637013916015625, 2.637980712890625, 2.637846435546875, 2.638864501953125, 5.43535302734375, 2.63853662109375, 2.63802978515625, 2.63697607421875, 2.637854736328125, 2.638664794921875, 2.63739794921875, 2.637233154296875, 2.637539306640625, 2.637358154296875, 2.637158447265625, 2.637360107421875, 2.63921875, 2.63846923828125, 2.638044189453125, 2.639015869140625, 2.636918701171875, 2.637613037109375, 2.637201416015625, 2.63785888671875, 2.6367763671875, 2.637592529296875, 2.63693310546875, 2.6386318359375, 2.63771337890625, 2.638298095703125, 2.637001708984375, 2.6378701171875, 2.6421064453125, 2.6414111328125, 2.641197021484375, 2.64100146484375, 2.63889111328125, 2.6406533203125, 2.6406552734375, 2.637020263671875, 2.637322265625, 2.63722900390625, 2.63769091796875, 2.6369853515625, 2.63765185546875, 2.63745751953125, 2.638650390625, 2.638825439453125, 2.637306884765625, 2.64330859375, 2.641152099609375, 2.6409931640625, 2.63775244140625, 2.639099853515625, 2.638685302734375, 2.6388388671875, 2.637993896484375, 2.639129638671875, 2.638023681640625, 2.638602294921875, 2.63790283203125, 2.63815380859375, 2.637550537109375, 2.63849169921875, 2.639201171875, 2.6400400390625, 2.638790771484375, 5.4299453125, 2.63796826171875, 2.638594970703125, 2.637592529296875, 2.6385439453125, 2.637388916015625, 2.638464111328125, 2.637737060546875, 2.63819775390625, 2.63681005859375, 2.637154296875, 2.63805224609375, 2.638739501953125, 2.63828173828125, 2.638024658203125, 2.637736083984375, 2.637816650390625, 2.63781884765625, 2.63653466796875, 2.637263916015625, 2.637474853515625, 2.63747998046875, 2.636569580078125, 2.64140478515625, 2.637125732421875, 2.638499755859375, 2.6369833984375, 2.637737060546875, 2.636338134765625, 2.63744921875, 2.637201416015625, 2.638740478515625, 2.638299072265625, 2.63929150390625, 2.637087646484375, 2.637667236328125, 2.6377021484375, 2.636683349609375, 2.637559814453125, 2.637581298828125, 2.6383564453125, 2.636505126953125, 2.63706005859375, 2.636032958984375, 2.6379765625, 2.637094970703125, 2.63819775390625, 2.64007568359375, 2.6376611328125, 2.6375966796875, 2.6389228515625, 2.63778515625, 2.639754150390625, 2.6381884765625, 2.63853466796875, 2.63933642578125, 2.639509521484375, 2.639793212890625, 2.640154541015625, 2.63959033203125, 2.636541015625, 2.6371533203125, 2.63670068359375, 5.43085791015625, 2.64038818359375, 2.640405517578125, 2.63921875, 2.638604248046875, 2.638392333984375, 2.63703857421875, 2.638612548828125, 2.638138427734375, 2.637631591796875, 2.637518798828125, 2.636412841796875, 2.639814697265625, 2.63709912109375, 2.638194580078125, 2.636851318359375, 2.637906982421875, 2.636904541015625, 2.63757421875, 2.637557861328125, 2.637801513671875, 2.637253662109375, 2.638193603515625, 2.63731396484375, 2.63754345703125, 2.637656005859375, 2.638063720703125, 2.6375537109375, 2.63655517578125, 2.636990478515625, 2.6373427734375, 2.638109619140625, 2.637427734375, 2.637330322265625, 2.63708154296875, 2.63813232421875, 2.636632080078125, 2.638185546875, 2.636789794921875, 2.6382294921875, 2.636559326171875, 2.63756494140625, 2.6364814453125, 2.637990966796875, 2.64270947265625, 2.63874755859375, 2.640064453125, 2.638127197265625, 2.637509521484375, 2.637388916015625, 2.637737060546875, 2.63737353515625, 2.639585205078125, 2.637530029296875, 2.638297119140625, 2.63809423828125, 2.637974609375, 2.63773583984375, 2.637834228515625, 2.63741552734375, 2.637762451171875, 2.637958251953125, 2.63785791015625, 5.428443359375, 2.63807275390625, 2.6370693359375, 2.638107666015625, 2.637665283203125, 2.63819677734375, 2.6379326171875, 2.63821923828125, 2.6370078125, 2.63802685546875, 2.638180419921875, 2.63954443359375, 2.6387373046875, 2.638023681640625, 2.636853271484375, 2.637129638671875, 2.637197265625, 2.636950439453125, 2.63758544921875, 2.638212158203125, 2.63828076171875, 2.6365849609375, 2.6370908203125, 2.638017578125, 2.63769384765625, 2.636833740234375, 2.638193603515625, 2.638341064453125, 2.640734130859375, 2.63990576171875, 2.639550537109375, 2.638340087890625, 2.63885107421875, 2.63923095703125, 2.6383955078125, 2.636854248046875, 2.636768310546875, 2.6369443359375, 2.637570068359375, 2.637836181640625, 2.63730078125, 2.63779541015625, 2.63750244140625, 2.637592529296875, 2.6366474609375, 2.638162841796875, 2.636866455078125, 2.6375146484375, 2.636231689453125, 2.6370693359375, 2.637305908203125, 2.63849169921875, 2.636894287109375, 2.636900390625, 2.637894775390625, 2.637490234375, 2.6377666015625, 2.64785107421875, 2.6376591796875, 2.637158447265625, 2.637832275390625, 2.63720849609375, 2.639467529296875, 5.425724609375, 2.637199462890625, 2.63809130859375, 2.638032958984375, 2.637854736328125, 2.638718017578125, 2.63857861328125, 2.637035400390625, 2.63773583984375, 2.638855224609375, 2.63739892578125, 2.638341064453125, 2.63885205078125, 2.639034423828125, 2.63807177734375, 2.637761474609375, 2.639326171875, 2.639801513671875, 2.639014892578125, 2.639296630859375, 2.637675537109375, 2.63769287109375, 2.637137939453125, 2.637119384765625, 2.6383955078125, 2.63736328125, 2.637094970703125, 2.637177734375, 2.6371953125, 2.637066162109375, 2.637592529296875, 2.639177734375, 2.637533203125, 2.6364169921875, 2.636937255859375, 2.6390283203125, 2.63927294921875, 2.63655126953125, 2.64357373046875, 2.637197265625, 2.637576171875, 2.638668701171875, 2.639214599609375, 2.639289306640625, 2.639532958984375, 2.638011474609375, 2.63690234375, 2.63817626953125, 2.636898193359375, 2.637287353515625, 2.637656982421875, 2.637981689453125, 2.636938232421875, 2.63751171875, 2.6373232421875, 2.63774609375, 2.637768798828125, 2.63883056640625, 2.639134765625, 2.641203125, 2.639278076171875, 2.638201904296875, 2.637488037109375, 5.42831396484375, 2.637304931640625, 2.638138427734375, 2.636865478515625, 2.639958984375, 2.63714404296875, 2.638241943359375, 2.63680810546875, 2.638630859375, 2.638310302734375, 2.640628662109375, 2.639406982421875, 2.63910302734375, 2.6390712890625, 2.640678955078125, 2.640310302734375, 2.63718701171875, 2.637646728515625, 2.637350830078125, 2.638884765625, 2.6375947265625, 2.637400146484375, 2.6368818359375, 2.642212890625, 2.637052978515625, 2.637252685546875, 2.636802001953125, 2.64029296875, 2.6377216796875, 2.6375556640625, 2.6366923828125, 2.63757421875, 2.638635009765625, 2.637696044921875, 2.637308837890625, 2.637477783203125, 2.637402099609375, 2.6365869140625, 2.637488037109375, 2.638253173828125, 2.6386328125, 2.637402099609375, 2.63834716796875, 2.6377451171875, 2.637392822265625, 2.636634033203125, 2.637593505859375, 2.637182861328125, 2.637476806640625, 2.6378896484375, 2.637464599609375, 2.639731689453125, 2.637717529296875, 2.63756591796875, 2.638674072265625, 2.63798974609375, 2.638200927734375, 2.63733251953125, 2.63752197265625, 2.63793359375, 2.636739501953125, 2.63832470703125, 2.637581298828125, 5.43222705078125, 2.638201904296875, 2.63801025390625, 2.639353759765625, 2.63851513671875, 2.637308837890625, 2.637641845703125, 2.6370263671875, 2.636823486328125, 2.638310302734375, 2.637739990234375, 2.636812255859375, 2.637005859375, 2.63739794921875, 2.637341796875, 2.636630126953125, 2.64057861328125, 2.638487548828125, 2.6377880859375, 2.63784033203125, 2.637433837890625, 2.638781494140625, 2.638268310546875, 2.63773486328125, 2.63734375, 2.64085302734375, 2.637772705078125, 2.63678369140625, 2.636781494140625, 2.637518798828125, 2.637106201171875, 2.636708984375, 2.63695166015625, 2.63906103515625, 2.6390087890625, 2.639414306640625, 2.63884912109375, 2.638530517578125, 2.639602783203125, 2.639301513671875, 2.638671875, 2.6394306640625, 2.640407470703125, 2.63750244140625, 2.640359375, 2.640257080078125, 2.641056884765625, 2.639901611328125, 2.64020263671875, 2.640825439453125, 2.637892578125, 2.63866357421875, 2.63870166015625, 2.638103515625, 2.641052734375, 2.639697998046875, 2.639138916015625, 2.63690234375, 2.63733447265625, 2.637642822265625, 2.63709912109375, 2.6370068359375, 2.638327880859375, 5.4337607421875, 2.63835546875, 2.638138427734375, 2.637888427734375, 2.63785986328125, 2.63816796875, 2.637724609375, 2.63817431640625, 2.63734375, 2.637358154296875, 2.637413330078125, 2.63850390625, 2.63788134765625, 2.638455810546875, 2.63848046875, 2.637498291015625, 2.637077392578125, 2.640005126953125, 2.638171142578125, 2.63721875, 2.638088134765625, 2.637340576171875, 2.637073486328125, 2.637824951171875, 2.63834619140625, 2.63819580078125, 2.63836474609375, 2.63822021484375, 2.637216796875, 2.63757421875, 2.63648046875, 2.6365439453125, 2.63712255859375, 2.63768994140625, 2.636919921875, 2.637600830078125, 2.637790283203125, 2.63785986328125, 2.636936279296875, 2.637821044921875, 2.637937744140625, 2.638752685546875, 2.63861767578125, 2.637978515625, 2.63923291015625, 2.639403076171875, 2.63928515625, 2.63895751953125, 2.63758447265625, 2.637350830078125, 2.63724853515625, 2.6378095703125, 2.638350341796875, 2.637401123046875, 2.636908447265625, 2.63790478515625, 2.638477294921875, 2.638087158203125, 2.639475830078125, 2.639469482421875, 2.640037841796875, 2.637907958984375, 2.636857421875, 5.4322666015625, 2.640575439453125, 2.641816650390625, 2.640384033203125, 2.639365234375, 2.639097900390625, 2.6402724609375, 2.638215087890625, 2.638160888671875, 2.637603759765625, 2.637083740234375, 2.6377421875, 2.637223876953125, 2.6389013671875, 2.638928955078125, 2.637473876953125, 2.637214599609375, 2.636859375, 2.637595703125, 2.63667822265625, 2.6370498046875, 2.638342041015625, 2.63901904296875, 2.637927490234375, 2.63882958984375, 2.640554931640625, 2.64306884765625, 2.639297607421875, 2.6393896484375, 2.64045166015625, 2.63986376953125, 2.639677490234375, 2.64067578125, 2.64060009765625, 2.640331787109375, 2.640183349609375, 2.639626220703125, 2.640194580078125, 2.63760498046875, 2.639280029296875, 2.6397255859375, 2.641039306640625, 2.6397880859375, 2.6377646484375, 2.639326171875, 2.63863916015625, 2.636462158203125, 2.637134765625, 2.6374892578125, 2.637487060546875, 2.63678466796875, 2.63695361328125, 2.636927978515625, 2.636856201171875, 2.63660546875, 2.637761474609375, 2.637675537109375, 2.6367119140625, 2.637622314453125, 2.6384404296875, 2.637370361328125, 2.63602880859375, 2.636900390625]",tokens/s,0.37339678057090686,,,,, -4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-125m,EleutherAI/gpt-neo-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/gemma-7b,google/gemma-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/gemma-7b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aaa06-6d9562733c7f0a2005faff54;36051902-3fdd-4bea-bcaf-26e4ffbf97f0) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/gemma-7b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ - self.load_model_with_no_weights() - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights - self.load_model_from_pretrained() - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained - self.pretrained_model = self.automodel_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained - return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3544, in from_pretrained - config = cls._autoset_attn_implementation( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1463, in _autoset_attn_implementation - config = cls._check_and_enable_sdpa( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1619, in _check_and_enable_sdpa - raise ValueError( -ValueError: GPTNeoForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/gemma-7b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-6B,,cuda,0,42,,,,,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,main,False,False,MB,1571.749888,5448.925184,0.0,4802.47808,4489.12128,s,10,5.0903373718261715,0.5090337371826171,0.0024963690031113454,0.5090882415771485,0.5114007537841796,0.5133009628295898,0.514821130065918,"[0.515201171875, 0.5058653564453125, 0.5075380249023438, 0.5109784851074218, 0.5074734191894531, 0.5066469421386719, 0.5090344543457032, 0.5092752685546875, 0.5091822204589844, 0.5091420288085937]",tokens/s,502.91362104386286,kWh,5.98198832737075e-06,3.277856253707796e-06,2.697877158300277e-05,3.623861616408132e-05,tokens/kWh,7064287.412104325,MB,1571.749888,5448.925184,0.0,4802.47808,4557.793792,s,10,299.8803828125,29.98803828125,0.03808075694133287,29.9707392578125,30.046216015625,30.058939062500002,30.0691175,"[29.972552734375, 30.043388671875, 30.071662109375, 30.010330078125, 29.96, 29.95865625, 29.96925, 29.972228515625, 29.963451171875, 29.95886328125]",tokens/s,2.1008376543053404,kWh,0.00035374844759702685,0.00019388459414373753,0.001556149911585597,0.0021037829533263617,tokens/kWh,29946.05498651303,,s,629,303.97862612915014,0.48327285553124066,0.06045359810307915,0.47556915283203127,0.47775867309570313,0.47801220703124997,0.9833881201171876,"[0.4750715026855469, 0.4762265625, 0.4773417053222656, 0.47625729370117187, 0.47571148681640624, 0.47557223510742186, 0.47629312133789065, 0.47471615600585937, 0.475789306640625, 0.47461376953125, 0.4746967163085937, 0.474913818359375, 0.47592657470703126, 0.47551071166992187, 0.47595623779296875, 0.47574118041992186, 0.4750817260742187, 0.47586611938476564, 0.4771154479980469, 0.47594082641601565, 0.4757462768554688, 0.47569818115234375, 0.4764293212890625, 0.475978759765625, 0.47632794189453126, 0.47545343017578123, 0.47498138427734377, 0.47522305297851564, 0.47550360107421874, 0.47533575439453124, 0.4750652770996094, 0.4751790161132812, 0.47525888061523436, 0.4754544677734375, 0.4752803955078125, 0.4756971435546875, 0.4763504638671875, 0.4758845520019531, 0.47809332275390626, 0.475462646484375, 0.47741644287109375, 0.47680307006835937, 0.47630642700195314, 0.47653070068359377, 0.47590911865234375, 0.4751380615234375, 0.47556710815429687, 0.475146240234375, 0.47564901733398435, 0.4757647399902344, 0.47530291748046877, 0.4750807189941406, 0.47518719482421873, 0.47629925537109374, 0.4754646911621094, 0.4754155578613281, 0.4753121337890625, 0.4750469055175781, 0.47505612182617185, 0.4759122009277344, 0.47506842041015623, 0.4758763427734375, 0.9829324951171875, 0.4772618103027344, 0.47679385375976563, 0.4752302551269531, 0.47515850830078127, 0.4750141296386719, 0.4753879089355469, 0.47504486083984376, 0.47488204956054686, 0.47477862548828126, 0.4748114013671875, 0.47600946044921877, 0.4750940246582031, 0.4750837707519531, 0.4750745544433594, 0.4750592041015625, 0.47484109497070315, 0.4761292724609375, 0.4751933288574219, 0.4750182495117187, 0.475357177734375, 0.47510528564453125, 0.4762142639160156, 0.47762432861328125, 0.4771512451171875, 0.478097412109375, 0.47743179321289064, 0.4781097106933594, 0.4774246520996094, 0.47769497680664064, 0.4778270568847656, 0.4778741760253906, 0.47795404052734375, 0.47758233642578124, 0.47754238891601564, 0.47758438110351564, 0.4775854187011719, 0.4776581115722656, 0.4781363220214844, 0.47784140014648435, 0.4785827941894531, 0.4778670043945312, 0.478013427734375, 0.4813271179199219, 0.4787435607910156, 0.4778823547363281, 0.4775577697753906, 0.47766937255859376, 0.4776509399414062, 0.47802163696289063, 0.4775925903320313, 0.4771829833984375, 0.47707955932617185, 0.4779130859375, 0.47752191162109375, 0.4774225769042969, 0.47764581298828124, 0.47809228515625, 0.47725054931640626, 0.47956683349609375, 0.47532647705078124, 0.47634738159179685, 0.4757237854003906, 0.9846497192382813, 0.4779151306152344, 0.47775845336914063, 0.4778516540527344, 0.4773468017578125, 0.47776461791992186, 0.47770932006835937, 0.4769669189453125, 0.4777676696777344, 0.4777635803222656, 0.47618765258789064, 0.4749752197265625, 0.4755538024902344, 0.4747980651855469, 0.4747796630859375, 0.4748114013671875, 0.47680514526367185, 0.4753622741699219, 0.47532440185546876, 0.4750438537597656, 0.4750325622558594, 0.47742669677734373, 0.4790394897460937, 0.47991500854492186, 0.4776028137207031, 0.4770672607421875, 0.47781991577148436, 0.4773304443359375, 0.4778946533203125, 0.4778496398925781, 0.4774788818359375, 0.4779632568359375, 0.47771136474609377, 0.47775640869140623, 0.47768267822265625, 0.47796121215820314, 0.4775454711914062, 0.4780103759765625, 0.4779438171386719, 0.4773253173828125, 0.4780001220703125, 0.47749530029296877, 0.478482421875, 0.4767160339355469, 0.47870156860351565, 0.47761203002929686, 0.47860427856445314, 0.47783221435546874, 0.4779366149902344, 0.4776212463378906, 0.4775301208496094, 0.477328369140625, 0.478376953125, 0.4772362365722656, 0.47751168823242185, 0.47775955200195314, 0.47782699584960936, 0.47732632446289064, 0.47806362915039063, 0.4772454528808594, 0.4775782775878906, 0.47722698974609373, 0.477765625, 0.9892290649414063, 0.47758950805664063, 0.47766937255859376, 0.47711026000976564, 0.4785479736328125, 0.4779674072265625, 0.47848751831054687, 0.4773918762207031, 0.47714407348632815, 0.475968505859375, 0.47733248901367187, 0.4777830505371094, 0.4774912109375, 0.4774410095214844, 0.4776990661621094, 0.47711026000976564, 0.47647540283203127, 0.4762900390625, 0.47571044921875, 0.47531622314453126, 0.47745126342773436, 0.4778526611328125, 0.4754595947265625, 0.4753295593261719, 0.47519024658203124, 0.47728536987304687, 0.4776847229003906, 0.47796734619140624, 0.47679693603515627, 0.4769525756835937, 0.47623678588867185, 0.47598489379882813, 0.47588760375976563, 0.47573504638671876, 0.4752127990722656, 0.47496600341796874, 0.4763709411621094, 0.47549542236328124, 0.4751493225097656, 0.4747591552734375, 0.47531109619140627, 0.47625933837890627, 0.47604122924804687, 0.4752691345214844, 0.478866455078125, 0.4751441650390625, 0.4752322692871094, 0.47499981689453125, 0.47643954467773436, 0.4755630187988281, 0.4757166137695312, 0.47565005493164064, 0.476015625, 0.4754155578613281, 0.4754124755859375, 0.47525070190429686, 0.47504281616210936, 0.4759347229003906, 0.47554763793945315, 0.47518923950195313, 0.47525274658203126, 0.475404296875, 0.47679489135742187, 0.9835653076171875, 0.4750100708007812, 0.47497518920898435, 0.47521792602539065, 0.47529168701171876, 0.47549435424804687, 0.47515341186523435, 0.47609548950195313, 0.47539712524414063, 0.47541351318359376, 0.47511859130859374, 0.47484622192382814, 0.47537664794921874, 0.475104248046875, 0.47634738159179685, 0.4752896118164063, 0.47548724365234374, 0.4761077880859375, 0.4751452026367188, 0.4771768188476562, 0.4755333251953125, 0.4755230712890625, 0.4750796813964844, 0.4751923217773438, 0.4752496643066406, 0.4748308410644531, 0.47501516723632814, 0.47498751831054686, 0.4759531555175781, 0.4754883117675781, 0.47538067626953123, 0.4753909912109375, 0.47539712524414063, 0.4753735656738281, 0.475494384765625, 0.47589376831054686, 0.4761006164550781, 0.47518206787109374, 0.4750602111816406, 0.47489434814453124, 0.47634228515625, 0.47604940795898437, 0.47567974853515627, 0.47519744873046876, 0.4751749267578125, 0.47551077270507813, 0.4754002685546875, 0.47505914306640623, 0.47493426513671877, 0.475788330078125, 0.4752465515136719, 0.47501516723632814, 0.47556915283203127, 0.4761507873535156, 0.47565823364257814, 0.47601458740234376, 0.47671194458007815, 0.478308349609375, 0.4755906677246094, 0.47542578125, 0.4754708557128906, 0.4771061706542969, 0.4754810791015625, 0.9827368774414063, 0.47548416137695315, 0.4752916564941406, 0.4750274658203125, 0.4750274658203125, 0.47477658081054686, 0.4755199890136719, 0.47510833740234376, 0.4752332763671875, 0.47511962890625, 0.4750837707519531, 0.47506329345703124, 0.4751718444824219, 0.477048828125, 0.4756756591796875, 0.475315185546875, 0.47508685302734377, 0.474925048828125, 0.476084228515625, 0.4754022521972656, 0.47531417846679686, 0.47490765380859373, 0.47505612182617185, 0.4752547912597656, 0.47493939208984376, 0.4751216735839844, 0.4752363586425781, 0.47553741455078125, 0.47550872802734373, 0.4752025451660156, 0.47503768920898437, 0.47521588134765624, 0.47512063598632814, 0.4751523742675781, 0.4767621154785156, 0.4764487609863281, 0.47532235717773436, 0.4752414855957031, 0.47862374877929686, 0.47638223266601565, 0.47562542724609375, 0.47544525146484373, 0.47627877807617186, 0.47506329345703124, 0.47538995361328124, 0.4754227294921875, 0.47551693725585936, 0.4754288635253906, 0.47596337890625, 0.4758896789550781, 0.4756899719238281, 0.4755988464355469, 0.4755916748046875, 0.4752066650390625, 0.4755640258789062, 0.47575653076171875, 0.47548416137695315, 0.4752998962402344, 0.4755332336425781, 0.4752363586425781, 0.4765552673339844, 0.4758056945800781, 0.47643853759765625, 0.9858508911132813, 0.4760975341796875, 0.4763627624511719, 0.47618048095703125, 0.47583026123046873, 0.4758814697265625, 0.47627365112304687, 0.47550360107421874, 0.4748226623535156, 0.4754380798339844, 0.4754176025390625, 0.47634841918945314, 0.47666278076171875, 0.4759449462890625, 0.47538177490234373, 0.475430908203125, 0.4754606018066406, 0.4765736999511719, 0.47562957763671876, 0.47619277954101563, 0.4762623901367187, 0.47586407470703124, 0.4763156433105469, 0.47800421142578126, 0.47563058471679687, 0.47566949462890623, 0.4764661865234375, 0.47568487548828126, 0.4763525085449219, 0.47546881103515626, 0.4754503784179688, 0.47535821533203126, 0.4751155090332031, 0.476685302734375, 0.4755333251953125, 0.4750540771484375, 0.4750325622558594, 0.4753387451171875, 0.47632281494140627, 0.47543603515625, 0.47514727783203126, 0.47536947631835935, 0.4750110778808594, 0.4750960693359375, 0.47505816650390625, 0.47526605224609375, 0.47514215087890627, 0.47597158813476564, 0.47550054931640623, 0.47509503173828127, 0.47519845581054687, 0.4751452026367188, 0.4753704833984375, 0.474967041015625, 0.47536639404296877, 0.47516876220703125, 0.47490457153320315, 0.4751523742675781, 0.47530905151367187, 0.4764610595703125, 0.47634228515625, 0.47607601928710935, 0.4758845520019531, 0.9857208251953125, 0.4755548095703125, 0.47583026123046873, 0.4755302734375, 0.4761640625, 0.4756316223144531, 0.475904052734375, 0.4751758728027344, 0.47539813232421874, 0.475536376953125, 0.4755210266113281, 0.475514892578125, 0.4755937805175781, 0.47548818969726564, 0.47541656494140627, 0.47544216918945315, 0.477454345703125, 0.4756387939453125, 0.4758138732910156, 0.47603302001953124, 0.47580465698242186, 0.47644158935546876, 0.4751994934082031, 0.47532339477539065, 0.4750745544433594, 0.47546981811523437, 0.4754565124511719, 0.47503768920898437, 0.47527935791015624, 0.4750858154296875, 0.4753018798828125, 0.4747683715820312, 0.4749916076660156, 0.47602789306640625, 0.4755906677246094, 0.4751790161132812, 0.47570123291015626, 0.477876220703125, 0.477338623046875, 0.4760637512207031, 0.4767999877929687, 0.47610983276367186, 0.47703347778320315, 0.47675802612304685, 0.47620712280273436, 0.4752158508300781, 0.47609036254882814, 0.4771328125, 0.4762552185058594, 0.4760504455566406, 0.47554458618164064, 0.475283447265625, 0.47543603515625, 0.4751523742675781, 0.475335693359375, 0.4752138366699219, 0.47528140258789064, 0.475109375, 0.47681332397460935, 0.47562240600585937, 0.47532339477539065, 0.47533465576171874, 0.47510833740234376, 0.984848388671875, 0.4750960693359375, 0.47506024169921873, 0.4759930419921875, 0.47555789184570313, 0.4750254211425781, 0.47533978271484373, 0.4758394775390625, 0.475610107421875, 0.47590911865234375, 0.4757596130371094, 0.4766791687011719, 0.4766371765136719, 0.4757176208496094, 0.4764549255371094, 0.4773990478515625, 0.4758917236328125, 0.476790771484375, 0.47595416259765627, 0.4752404479980469, 0.47515545654296876, 0.475030517578125, 0.4752916564941406, 0.4751278076171875, 0.4755138549804688, 0.47549234008789065, 0.47554150390625, 0.4758425598144531, 0.47556607055664063, 0.4752015380859375, 0.47505612182617185, 0.4750540771484375, 0.47509811401367186, 0.4751697998046875, 0.4750817260742187, 0.4752066650390625, 0.4769587097167969, 0.4757596130371094, 0.47552410888671875, 0.4753387451171875, 0.4751247253417969, 0.47600946044921877, 0.47491787719726564, 0.47531417846679686, 0.47513394165039063, 0.4757739562988281, 0.4750469055175781, 0.475030517578125, 0.4758507385253906, 0.4755599365234375, 0.47600741577148437, 0.4752414855957031, 0.4751247253417969, 0.47610470581054687, 0.4753530883789063, 0.475404296875, 0.4752005004882813, 0.476626953125, 0.4756654052734375, 0.47670578002929687, 0.4761466979980469, 0.4752547912597656, 0.4752209777832031, 0.9850972290039063, 0.47500698852539064, 0.47586099243164065, 0.475536376953125, 0.4751933288574219, 0.4753070068359375, 0.47523434448242186, 0.475157470703125, 0.4753039245605469, 0.4752762756347656, 0.4760606689453125, 0.47565216064453125, 0.4750796203613281, 0.47518106079101563, 0.4763607177734375, 0.47558758544921875, 0.47543499755859375, 0.47558349609375, 0.4752384033203125, 0.47536639404296877, 0.47527423095703125, 0.47547494506835936, 0.4750335998535156, 0.4753049621582031, 0.47516571044921874, 0.4754155578613281, 0.4772812805175781, 0.4751769714355469, 0.4758814697265625, 0.4761343994140625, 0.4757074279785156, 0.4766033630371094, 0.47504281616210936, 0.47508480834960937, 0.4752629699707031, 0.47697714233398436, 0.4764979248046875, 0.4756357421875, 0.4753837890625, 0.4751473083496094, 0.475273193359375, 0.475177978515625, 0.47527835083007813, 0.4753541259765625, 0.4753623046875, 0.4753930358886719, 0.4750796813964844, 0.47559576416015625, 0.47536334228515625, 0.4764241943359375, 0.47548724365234374, 0.47530087280273436, 0.47603302001953124, 0.4752906188964844, 0.47511962890625, 0.47501516723632814, 0.4768563232421875, 0.4758026123046875, 0.47588864135742187, 0.4753950805664062, 0.4752138366699219, 0.47535000610351563, 0.4750120849609375]",tokens/s,2.069224432025556,,,,, -4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-2.7B,EleutherAI/gpt-neo-2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,huggyllama/llama-13b,huggyllama/llama-13b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,2981.855232,9259.450368,0.0,8613.003264,8211.364864,s,10,10.951241455078126,1.0951241455078125,0.002069112877122291,1.0949886474609376,1.0979957763671875,1.0981948486328124,1.0983541064453124,"[1.0979515380859375, 1.0983939208984375, 1.0932901611328125, 1.0924078369140624, 1.0937889404296874, 1.092544677734375, 1.0941727294921875, 1.0958045654296875, 1.096221435546875, 1.0966656494140625]",tokens/s,233.76345143161097,kWh,1.2900002946456276e-05,7.068370924207557e-06,5.9710658879597037e-05,7.967903275026086e-05,tokens/kWh,3212890.4074724964,MB,2986.262528,9330.753536,0.0,8684.306432,8503.627264,s,10,640.8562343749999,64.0856234375,0.007507663361876726,64.08716796875,64.093946875,64.09397148437499,64.09399117187499,"[64.08627734375, 64.0873125, 64.09369921875, 64.09394140625, 64.0870234375, 64.09399609375, 64.08822265625, 64.078796875, 64.0717890625, 64.07517578125]",tokens/s,0.9830597975135754,kWh,0.000756652800159322,0.0004147102533278666,0.0034761336142379974,0.004647496667725186,tokens/kWh,13555.684813613145,,s,629,649.691616149902,1.0328960511127223,0.130104123840355,1.0171883544921876,1.0177021728515625,1.01789716796875,2.1117173046874997,"[1.0174935302734376, 1.0171514892578124, 1.0171883544921876, 1.017997314453125, 1.0175242309570312, 1.0175477905273438, 1.0175784912109376, 1.017881591796875, 1.0172057495117188, 1.0168115234375, 1.0167367553710938, 1.0173132934570313, 1.0168719482421875, 1.0171760864257813, 1.0166824951171876, 1.0168237915039062, 1.0169927978515625, 1.0175252685546874, 1.0170921020507813, 1.0169978637695312, 1.0169354248046876, 1.0170029907226563, 1.0173900756835939, 1.017280517578125, 1.016859619140625, 1.0172426147460938, 1.016791015625, 1.0168565673828125, 1.0172825317382812, 1.01675927734375, 1.0171392211914063, 1.0170572509765625, 1.0173419799804688, 1.0176777954101563, 1.0177576904296874, 1.0170183715820313, 1.0177105712890624, 1.0174996337890625, 1.0173941650390625, 1.0173890380859374, 1.017080810546875, 1.0169651489257812, 1.0173880615234374, 1.0172467041015625, 1.0171791381835937, 1.016754150390625, 1.0172498168945312, 1.0174290161132813, 1.0169159545898439, 1.0170194091796876, 1.016964111328125, 1.0168411865234375, 1.0170531616210938, 1.0171678466796874, 1.0177003784179688, 1.0177116088867189, 1.0175713500976562, 1.0176112670898438, 1.01741259765625, 1.01737060546875, 1.0172507934570312, 1.0171494140625, 2.116391845703125, 1.01690673828125, 1.0169548950195313, 1.0181068725585938, 1.0178170776367188, 1.0172692260742187, 1.0174985961914063, 1.017291748046875, 1.0175293579101563, 1.0175150146484375, 1.016748046875, 1.0170767211914062, 1.0172333984375, 1.0175641479492188, 1.0175477905273438, 1.0169292602539062, 1.016869873046875, 1.0169712524414063, 1.0170634155273437, 1.0173245239257813, 1.0168094482421874, 1.016943603515625, 1.0171064453125, 1.0171781005859375, 1.017275390625, 1.0166548461914062, 1.017407470703125, 1.0169978637695312, 1.016574951171875, 1.0173306884765625, 1.0168370971679688, 1.016875, 1.0170009765625, 1.0174505004882812, 1.0167992553710938, 1.0171494140625, 1.0171607055664063, 1.0168944702148437, 1.0170101928710937, 1.017059326171875, 1.017080810546875, 1.0172037353515626, 1.0174771118164063, 1.0174095458984376, 1.0170245361328125, 1.019852783203125, 1.0168125610351562, 1.0170029907226563, 1.0175538940429687, 1.0178017578125, 1.0176777954101563, 1.016995849609375, 1.0182564086914063, 1.0176224975585937, 1.0173235473632813, 1.01770751953125, 1.0170552368164063, 1.01711669921875, 1.0171351318359374, 1.0170449829101562, 1.016859619140625, 1.0170890502929688, 1.0175088500976563, 2.11154345703125, 1.0170613403320312, 1.0173931274414063, 1.0170582885742188, 1.01701123046875, 1.01686474609375, 1.0172272338867188, 1.0170921020507813, 1.0172078247070313, 1.0167490844726563, 1.0166835327148438, 1.0169088134765625, 1.0174351196289062, 1.0175641479492188, 1.017354248046875, 1.0168699340820313, 1.0168974609375, 1.0171945190429688, 1.0176737670898437, 1.0177105102539064, 1.017945068359375, 1.0175324096679688, 1.0171945190429688, 1.0176880493164062, 1.0172406005859376, 1.0171146240234374, 1.0182410278320313, 1.0179573974609375, 1.0179983520507812, 1.0176327514648438, 1.0174893798828124, 1.0176010131835938, 1.0182543334960938, 1.0177720336914062, 1.0173716430664062, 1.017248779296875, 1.016859619140625, 1.0170859375, 1.0173767700195313, 1.0172733154296876, 1.0168115234375, 1.0170787963867187, 1.0168923950195312, 1.0172713012695314, 1.0172866821289062, 1.0175477905273438, 1.0167623901367187, 1.0171729736328126, 1.0170921020507813, 1.0174443359375, 1.0174791870117188, 1.017380859375, 1.0176481323242188, 1.0177054443359375, 1.0170101928710937, 1.01709619140625, 1.0171812133789062, 1.0176399536132812, 1.0173388671875, 1.017680908203125, 1.0173245239257813, 1.0175344848632812, 1.016958984375, 2.111784912109375, 1.0173778076171875, 1.0168678588867188, 1.0177362060546875, 1.01760205078125, 1.0175693359375, 1.0170152587890624, 1.0167971801757814, 1.0172047119140626, 1.0175958862304688, 1.016896484375, 1.0167449340820311, 1.017417724609375, 1.0172200927734374, 1.016796142578125, 1.0171340942382812, 1.016943603515625, 1.01682177734375, 1.0168862915039063, 1.0173562622070313, 1.017080810546875, 1.0174678955078125, 1.0173972778320313, 1.0175897827148437, 1.017354248046875, 1.0172293090820312, 1.0175057983398437, 1.0172456665039062, 1.0170859375, 1.0177197875976562, 1.017154541015625, 1.0173880615234374, 1.0175590209960939, 1.01743408203125, 1.016933349609375, 1.0172835693359374, 1.017112548828125, 1.0172262573242188, 1.0170572509765625, 1.017354248046875, 1.01725390625, 1.0173184204101562, 1.018461181640625, 1.0179942626953125, 1.0194544677734374, 1.0179215087890625, 1.0178590698242187, 1.0177402954101562, 1.017133056640625, 1.0170347290039063, 1.0172620849609375, 1.0173992919921875, 1.0176665649414063, 1.01707470703125, 1.0167613525390624, 1.0173092041015626, 1.0173604125976563, 1.0174525146484374, 1.017565185546875, 1.0170890502929688, 1.01747509765625, 1.017607177734375, 1.0175682373046875, 2.112203857421875, 1.017459716796875, 1.0166384887695312, 1.0169661865234374, 1.0174985961914063, 1.0174044189453124, 1.016933349609375, 1.0176819458007813, 1.0176041259765625, 1.0174843139648437, 1.017217041015625, 1.0171586303710938, 1.01719140625, 1.017375732421875, 1.0171525268554686, 1.0169210815429688, 1.017154541015625, 1.0176266479492186, 1.0178191528320313, 1.0170921020507813, 1.0168934326171875, 1.0171627807617187, 1.0176942138671874, 1.017333740234375, 1.0175324096679688, 1.0170572509765625, 1.0169077758789062, 1.017111572265625, 1.0173767700195313, 1.0171566162109376, 1.0170460205078125, 1.0172252197265625, 1.0173870239257812, 1.0172252197265625, 1.0171300048828125, 1.0168084716796875, 1.017064453125, 1.0177720336914062, 1.016975341796875, 1.0169774169921875, 1.0169395141601563, 1.0170787963867187, 1.0171217651367188, 1.0177638549804688, 1.0169712524414063, 1.016958984375, 1.01680126953125, 1.0175764770507814, 1.0173480834960937, 1.0168197021484375, 1.017017333984375, 1.0169978637695312, 1.0169896850585938, 1.0174003295898437, 1.0170214233398438, 1.01707568359375, 1.0175559692382812, 1.01893017578125, 1.0172241821289063, 1.0170316772460937, 1.01722216796875, 1.0172119140625, 1.0173163452148437, 2.1100595703125, 1.0167705688476563, 1.0177894287109375, 1.0171873168945313, 1.0167449340820311, 1.0167142333984376, 1.016875, 1.0168862915039063, 1.0167982177734376, 1.0173235473632813, 1.0169313354492187, 1.0169343872070313, 1.0175621337890626, 1.017955322265625, 1.0172938232421875, 1.0173009643554687, 1.017554931640625, 1.0171617431640625, 1.0173388671875, 1.0177402954101562, 1.0175774536132813, 1.0175682373046875, 1.0177136840820313, 1.0181427001953125, 1.0177576904296874, 1.0176788330078126, 1.017692138671875, 1.017617431640625, 1.0178262939453124, 1.0180515747070313, 1.0169354248046876, 1.0168514404296876, 1.0176296997070313, 1.0172395629882813, 1.0167859497070313, 1.0172784423828125, 1.0170859375, 1.0170224609375, 1.0186465454101563, 1.0175170288085937, 1.0169405517578125, 1.0179154052734376, 1.0179379272460938, 1.01726318359375, 1.0171544799804688, 1.0171340942382812, 1.0169609985351562, 1.0169047241210938, 1.017776123046875, 1.017396240234375, 1.0173921508789063, 1.0171002807617187, 1.0185277709960938, 1.017439208984375, 1.017312255859375, 1.017007080078125, 1.0172160034179687, 1.0172160034179687, 1.017218017578125, 1.0172958984375, 1.0172764282226563, 1.0170203857421876, 1.017691162109375, 2.111909912109375, 1.0176123046875, 1.0175139770507813, 1.0174228515625, 1.0174617309570313, 1.0175221557617187, 1.0169579467773437, 1.0177576904296874, 1.0171986083984375, 1.0168289184570312, 1.0170685424804689, 1.0173532104492187, 1.0175938720703126, 1.017185302734375, 1.01743408203125, 1.0171443481445313, 1.016764404296875, 1.0172661743164062, 1.0177013549804688, 1.0171986083984375, 1.017049072265625, 1.0170828857421874, 1.0174166870117187, 1.0170664672851562, 1.0174453735351563, 1.0168790893554687, 1.0178897705078125, 1.0175170288085937, 1.0171791381835937, 1.0170726928710938, 1.0169456176757812, 1.0167869262695313, 1.0170419311523438, 1.0173788452148438, 1.017333740234375, 1.0175242309570312, 1.018076171875, 1.01758056640625, 1.0170736694335938, 1.0170521850585938, 1.0168186645507813, 1.0173572998046876, 1.0176532592773437, 1.01709619140625, 1.0169476928710937, 1.0168115844726562, 1.0169046630859375, 1.0173613891601563, 1.0177177734375, 1.0168975219726561, 1.016826904296875, 1.0167296142578126, 1.017049072265625, 1.0175293579101563, 1.0174402465820314, 1.0174054565429687, 1.0173552856445311, 1.017469970703125, 1.0172620849609375, 1.0167603149414062, 1.0176635131835938, 1.0175570068359374, 1.0175396118164062, 2.113271728515625, 1.0164520874023437, 1.0175570068359374, 1.0175221557617187, 1.0169467163085937, 1.01684326171875, 1.01673779296875, 1.0166527709960937, 1.016406005859375, 1.01699072265625, 1.0166343383789063, 1.0166343383789063, 1.016573974609375, 1.0167418823242187, 1.0167675170898438, 1.0167920532226562, 1.0167357177734375, 1.0169302978515624, 1.0171954956054687, 1.0174054565429687, 1.0170368041992188, 1.0171791381835937, 1.0176737060546874, 1.0177310791015626, 1.0174033813476562, 1.0174044189453124, 1.017459716796875, 1.0175098876953126, 1.0173675537109375, 1.0176378784179687, 1.017259033203125, 1.0173460693359375, 1.0176511840820313, 1.0174822387695313, 1.0173767700195313, 1.0175150146484375, 1.0174218139648437, 1.0178508911132813, 1.0178406372070312, 1.0176204833984375, 1.0173767700195313, 1.0178017578125, 1.0179000244140626, 1.0172119140625, 1.0165678100585938, 1.0165330200195313, 1.0163292236328125, 1.0167285766601561, 1.0169968872070312, 1.0166988525390626, 1.0169210815429688, 1.0170480346679687, 1.0171238403320313, 1.0167817993164063, 1.0167633666992189, 1.0170337524414061, 1.0168043823242188, 1.0171392211914063, 1.017153564453125, 1.0167838745117188, 1.0169251708984375, 1.0174423217773438, 1.01701220703125, 2.113585205078125, 1.0164449462890626, 1.016585205078125, 1.0164551391601562, 1.01684326171875, 1.016680419921875, 1.016753173828125, 1.0173470458984375, 1.0173030395507812, 1.0169528198242188, 1.0168330078125, 1.0170194091796876, 1.01732763671875, 1.0167736206054687, 1.0168514404296876, 1.0169835815429686, 1.0166835327148438, 1.01719140625, 1.0171043701171876, 1.0166845703125, 1.017260009765625, 1.0176849975585938, 1.0174935302734376, 1.017529296875, 1.0172733154296876, 1.0169231567382813, 1.0173982543945312, 1.0170020141601563, 1.016616943359375, 1.0167767333984374, 1.0169467163085937, 1.0172088623046875, 1.01715966796875, 1.016932373046875, 1.0170357666015626, 1.0171320190429687, 1.01718017578125, 1.0174566650390624, 1.0167255249023437, 1.0167971801757814, 1.0167633666992189, 1.0170337524414061, 1.0170050659179688, 1.0170439453125, 1.01673779296875, 1.0168330078125, 1.0170828857421874, 1.017059326171875, 1.0168739624023437, 1.0171300048828125, 1.0171238403320313, 1.016974365234375, 1.0169026489257813, 1.0171504516601562, 1.0168207397460938, 1.0168319702148438, 1.0172692260742187, 1.0173624877929688, 1.016796142578125, 1.0168391723632813, 1.0171893920898438, 1.017365478515625, 1.016943603515625, 2.1139599609375, 1.0168893432617188, 1.0171945190429688, 1.016859619140625, 1.0165811157226563, 1.0167500610351563, 1.01697021484375, 1.016826904296875, 1.0168453369140624, 1.0170132446289062, 1.0168053588867187, 1.0173511962890625, 1.01734912109375, 1.0167859497070313, 1.0169036865234375, 1.0171238403320313, 1.0169866333007813, 1.0171996459960937, 1.0168934326171875, 1.0169763793945312, 1.0167183227539063, 1.0170234985351563, 1.0178928833007812, 1.0167654418945313, 1.0169620361328124, 1.0170040283203126, 1.0167449340820311, 1.0173562622070313, 1.016680419921875, 1.0172979125976562, 1.0165924072265624, 1.017101318359375, 1.0172999877929687, 1.0170224609375, 1.0172262573242188, 1.0168402099609375, 1.0170101928710937, 1.0167879638671875, 1.01705419921875, 1.0175529174804687, 1.0173265991210938, 1.01718017578125, 1.0174719848632812, 1.0170245361328125, 1.0174248657226563, 1.016958984375, 1.0166087646484374, 1.0166466674804688, 1.0171975708007812, 1.01707470703125, 1.0169210815429688, 1.0175795288085938, 1.0173245239257813, 1.0168084716796875, 1.0169722900390625, 1.0170470581054687, 1.0169948120117187, 1.0169896850585938, 1.0172006225585937, 1.0173716430664062, 1.0176337890625, 1.0176676025390625, 1.0172323608398437]",tokens/s,0.9681516343515072,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-7B,Qwen/Qwen-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -213,28 +437,28 @@ ChildProcessError: Traceback (most recent call last): self.load_model_from_pretrained() File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained self.pretrained_model = self.automodel_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained - return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3544, in from_pretrained - config = cls._autoset_attn_implementation( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1463, in _autoset_attn_implementation - config = cls._check_and_enable_sdpa( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1619, in _check_and_enable_sdpa - raise ValueError( -ValueError: GPTNeoForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 550, in from_pretrained + model_class = get_class_from_dynamic_module( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 498, in get_class_from_dynamic_module + final_module = get_cached_module_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 323, in get_cached_module_file + modules_needed = check_imports(resolved_module_file) + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 181, in check_imports + raise ImportError( +ImportError: This modeling file requires the following packages that were not found in your environment: transformers_stream_generator. Run `pip install transformers_stream_generator` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciCoder-1b,Deci/DeciCoder-1b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-72B,Qwen/Qwen1.5-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -242,34 +466,41 @@ ChildProcessError: Traceback (most recent call last): self.load_model_from_pretrained() File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained self.pretrained_model = self.automodel_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 558, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained - hf_quantizer.postprocess_model(model) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model - return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading - model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 432, in post_init_awq_exllama_modules - model = exllama_post_init(model) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init - submodule.post_init() - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init - self.q4 = exl_ext.make_q4( -NameError: name 'exl_ext' is not defined + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 96.00 MiB. GPU -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 -4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciLM-7B,Deci/DeciLM-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciLM-7B,Deci/DeciLM-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) - File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): - File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -279,26 +510,26 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 558, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3544, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained config = cls._autoset_attn_implementation( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1463, in _autoset_attn_implementation + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation config = cls._check_and_enable_sdpa( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1619, in _check_and_enable_sdpa + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa raise ValueError( ValueError: DeciLMForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 -4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-j-6b,EleutherAI/gpt-j-6b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-j-6b,EleutherAI/gpt-j-6b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) - File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): - File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -308,26 +539,27 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3544, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained config = cls._autoset_attn_implementation( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1463, in _autoset_attn_implementation + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation config = cls._check_and_enable_sdpa( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1619, in _check_and_enable_sdpa + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa raise ValueError( ValueError: GPTJForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 -4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-1.3B,EleutherAI/gpt-neo-1.3B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-0.5B,Qwen/Qwen1.5-0.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,1441.677312,1709.703168,0.0,1063.256064,942.605312,s,10,0.8730960388183594,0.08730960388183594,0.002035304872210584,0.08713673782348633,0.08910322875976563,0.09054942092895508,0.09170637466430664,"[0.09199561309814454, 0.0843604507446289, 0.08696444702148437, 0.08719612884521484, 0.08721517181396485, 0.08707734680175781, 0.08465267181396484, 0.0866138916015625, 0.08823846435546875, 0.08878185272216797]",tokens/s,2932.0943930345643,kWh,1.0099359104323139e-06,5.533989598310048e-07,2.5805410008034e-06,4.143875871066718e-06,tokens/kWh,61777912.265045814,MB,1441.677312,1709.703168,0.0,1063.256064,942.607872,s,10,53.98236181640625,5.398236181640624,0.0409841937132303,5.390563720703125,5.449904833984375,5.452959252929687,5.455402788085937,"[5.38043359375, 5.40069384765625, 5.41887353515625, 5.44691552734375, 5.456013671875, 5.44922607421875, 5.37513134765625, 5.35530859375, 5.36842431640625, 5.33134130859375]",tokens/s,11.670478630457612,kWh,6.202919682701882e-05,3.3993588802392576e-05,0.00015045469253839705,0.00024647747816780845,tokens/kWh,255601.4467054386,,s,629,54.67969024658198,0.08693114506610816,0.010438416792585154,0.08622489929199219,0.08683417816162109,0.08736747589111328,0.17189485961914075,"[0.08353689575195312, 0.08669286346435547, 0.08684236907958984, 0.08664883422851563, 0.08732160186767578, 0.0871107177734375, 0.08675321960449218, 0.08646348571777343, 0.08578867340087891, 0.08561151885986328, 0.08533920288085937, 0.08318355560302734, 0.08327986907958984, 0.08331366729736328, 0.08347750091552734, 0.08323891448974609, 0.08337407684326172, 0.08320102691650391, 0.08308223724365234, 0.08320921325683593, 0.08336589050292968, 0.08362290954589843, 0.08334028625488281, 0.08379084777832031, 0.08326246643066407, 0.08351641845703126, 0.08637849426269531, 0.0860948486328125, 0.08636211395263672, 0.0862208023071289, 0.0863825912475586, 0.08643379211425781, 0.08343961334228515, 0.08492339324951172, 0.08729497528076172, 0.08630579376220703, 0.08625049591064453, 0.08637340545654297, 0.08612451171875, 0.08631295776367187, 0.0867215347290039, 0.08637235260009765, 0.08630169677734376, 0.08631193542480468, 0.08447795104980468, 0.08495410919189453, 0.08645938873291016, 0.08668978881835937, 0.08635699462890625, 0.08412057495117188, 0.0861143035888672, 0.08629350280761719, 0.08610099029541016, 0.0859535369873047, 0.086508544921875, 0.08645426940917969, 0.0861286392211914, 0.08638976287841797, 0.08641228485107422, 0.08600883483886719, 0.08566067504882813, 0.08624947357177734, 0.16856166076660156, 0.08411341094970703, 0.0832890853881836, 0.08320819091796874, 0.08320409393310547, 0.08308633422851562, 0.08329011535644532, 0.08373766326904297, 0.08319993591308594, 0.08375193786621093, 0.08421683502197265, 0.08632524871826172, 0.0862371826171875, 0.08634674835205078, 0.08365670776367187, 0.0833986587524414, 0.08349798583984375, 0.08350003051757812, 0.08335769653320313, 0.08395059204101563, 0.0839208984375, 0.08677375793457032, 0.0867154541015625, 0.08665900421142578, 0.0864194564819336, 0.08657305908203125, 0.08616550445556641, 0.08598429107666015, 0.08726729583740235, 0.08673792266845703, 0.08624230194091796, 0.0869222412109375, 0.08644096374511719, 0.0866344985961914, 0.08689356994628906, 0.08668672180175781, 0.08644608306884766, 0.08659046173095702, 0.08682291412353516, 0.08598323059082032, 0.08748851013183594, 0.08676249694824219, 0.08683213043212891, 0.08647679901123047, 0.08678195190429687, 0.08769843292236328, 0.08739942169189453, 0.08640614318847656, 0.08639897918701171, 0.08644403076171875, 0.08633344268798829, 0.08639488220214844, 0.08622694396972656, 0.0867215347290039, 0.0865638427734375, 0.08646451568603515, 0.08639078521728516, 0.08660889434814453, 0.08628224182128906, 0.08633753967285156, 0.08671437072753906, 0.08621875, 0.08635084533691406, 0.17366220092773438, 0.08666828918457031, 0.08657817840576172, 0.08662118530273437, 0.08642969512939454, 0.08634265899658203, 0.0862525405883789, 0.08606208038330078, 0.08631501007080078, 0.08632012939453125, 0.08620851135253907, 0.08628530883789062, 0.08648499298095703, 0.08680038452148438, 0.08667135620117188, 0.08662732696533203, 0.08640409851074218, 0.08657817840576172, 0.08667033386230469, 0.08634060668945312, 0.08654131317138672, 0.08610816192626954, 0.08561766052246093, 0.08718950653076171, 0.08702668762207032, 0.08669388580322265, 0.08728883361816406, 0.08666214752197265, 0.08715264129638672, 0.08645017242431641, 0.08640409851074218, 0.08617984008789062, 0.08535756683349609, 0.0866324462890625, 0.08659967803955078, 0.08661299133300782, 0.08663346862792969, 0.08641843414306641, 0.08643379211425781, 0.08636006164550782, 0.08653209686279296, 0.0865771484375, 0.08655257415771485, 0.08642047882080078, 0.08629657745361329, 0.08657100677490234, 0.0862955551147461, 0.08327884674072265, 0.08376831817626954, 0.08357273864746094, 0.08349286651611328, 0.08337715148925781, 0.0860549087524414, 0.08430796813964844, 0.08347443389892578, 0.08344064331054687, 0.08356454467773437, 0.08309248352050781, 0.08637235260009765, 0.08680242919921875, 0.0863078384399414, 0.08653311920166015, 0.08646041870117188, 0.1741107177734375, 0.08636214447021484, 0.08619312286376953, 0.08634162902832031, 0.08638976287841797, 0.08647885131835938, 0.08649215698242188, 0.08664064025878906, 0.08642771148681641, 0.08624428558349609, 0.08617369842529297, 0.08573542022705079, 0.08636109161376954, 0.08620134735107422, 0.086329345703125, 0.08637644958496093, 0.08652902221679687, 0.08682086181640625, 0.08738508605957031, 0.0865249252319336, 0.0863815689086914, 0.0862894058227539, 0.08674918365478515, 0.08637337493896484, 0.08735027313232421, 0.08632422637939453, 0.08617984008789062, 0.08628428649902344, 0.08778956604003907, 0.08680754852294922, 0.08660582733154297, 0.08646348571777343, 0.08672358703613281, 0.0874486083984375, 0.08708297729492187, 0.08639078521728516, 0.08637337493896484, 0.08624742126464843, 0.08626278686523438, 0.08637651062011718, 0.08677881622314453, 0.08631501007080078, 0.08682803344726563, 0.0861112289428711, 0.08643583679199218, 0.08626380920410157, 0.08634572601318359, 0.086181884765625, 0.08633036804199219, 0.08633856201171874, 0.08504627227783203, 0.08628121948242187, 0.08638771057128906, 0.08684748840332031, 0.0863825912475586, 0.08656588745117187, 0.08662528228759765, 0.08654131317138672, 0.08605388641357421, 0.08622694396972656, 0.0862955551147461, 0.08632524871826172, 0.08606412506103515, 0.1738434600830078, 0.0866170883178711, 0.08668978881835937, 0.0868823013305664, 0.08633548736572266, 0.08676761627197266, 0.08665599822998046, 0.08651468658447266, 0.08632012939453125, 0.08640306854248046, 0.08641433715820312, 0.0875315170288086, 0.08692838287353516, 0.0863477783203125, 0.0873861083984375, 0.08702054595947266, 0.08652390289306641, 0.08673382568359375, 0.08648191833496094, 0.0868485107421875, 0.08659967803955078, 0.0865249252319336, 0.08798515319824218, 0.08668672180175781, 0.0869713897705078, 0.08641024017333984, 0.08619417572021484, 0.08626278686523438, 0.08664371490478516, 0.0865577621459961, 0.08645420837402344, 0.08650342559814453, 0.08614502716064452, 0.0860057601928711, 0.08658636474609376, 0.08687513732910156, 0.08634572601318359, 0.08673484802246094, 0.086614013671875, 0.08654745483398438, 0.08647782135009766, 0.08652288055419922, 0.08698777770996094, 0.08646348571777343, 0.08669286346435547, 0.08713728332519531, 0.08633344268798829, 0.08693965148925781, 0.0867000961303711, 0.08665286254882812, 0.08626483154296875, 0.0867430419921875, 0.08634982299804687, 0.08623616027832032, 0.0861685791015625, 0.08641228485107422, 0.08606719970703125, 0.08660889434814453, 0.08628736114501953, 0.08668057250976563, 0.08639794921875, 0.08680242919921875, 0.08641741180419922, 0.1737769012451172, 0.08605696105957031, 0.08600371551513672, 0.08643071746826173, 0.08781107330322266, 0.0866519012451172, 0.08646246337890626, 0.08642668914794922, 0.08613164520263672, 0.08653823852539062, 0.08642047882080078, 0.0866324462890625, 0.08596377563476562, 0.0863815689086914, 0.08631398773193359, 0.08695295715332031, 0.08793907165527344, 0.08786227416992187, 0.08662630462646484, 0.08779571533203125, 0.08724275207519532, 0.08729190063476562, 0.0877844467163086, 0.08636006164550782, 0.0858818588256836, 0.08638771057128906, 0.086949951171875, 0.08719251251220703, 0.08616550445556641, 0.08660377502441406, 0.08642457580566407, 0.08629452514648438, 0.0865054702758789, 0.08606105804443359, 0.08653311920166015, 0.08622796630859375, 0.0860200958251953, 0.08642253112792969, 0.08642355346679688, 0.0863641586303711, 0.08620851135253907, 0.0867583999633789, 0.08634880065917969, 0.08620851135253907, 0.08670003509521484, 0.08732057952880859, 0.08644608306884766, 0.08616960144042969, 0.08388813018798828, 0.08584806060791016, 0.08618905639648437, 0.0861788787841797, 0.0874474868774414, 0.08617574310302735, 0.08607027435302735, 0.0862064666748047, 0.08648703765869141, 0.08633241271972657, 0.08608153533935547, 0.08628838348388672, 0.08599142456054687, 0.08555622100830078, 0.08707379150390625, 0.1688248291015625, 0.08354815673828125, 0.08360345458984375, 0.08642150115966797, 0.08649830627441406, 0.0861808624267578, 0.08627609252929687, 0.08619622039794922, 0.0862033920288086, 0.08743424224853516, 0.08643276977539062, 0.08662425231933593, 0.08616754913330078, 0.08617164611816407, 0.0859135971069336, 0.08686489868164063, 0.08547532653808594, 0.08608972930908203, 0.08635596466064453, 0.08650649261474609, 0.08669593811035156, 0.0856258544921875, 0.08557164764404297, 0.08602413177490234, 0.08610304260253906, 0.08613478088378906, 0.0858818588256836, 0.08467046356201172, 0.08328396606445312, 0.08320614624023437, 0.08348467254638672, 0.0833433609008789, 0.08343142700195312, 0.08517120361328125, 0.08331775665283203, 0.0835225601196289, 0.08325939178466797, 0.0829675521850586, 0.08333618927001953, 0.0835962905883789, 0.08292864227294922, 0.08303923034667969, 0.0840273895263672, 0.08592588806152343, 0.08553369903564453, 0.08563097381591797, 0.08609180450439453, 0.08550192260742187, 0.08622489929199219, 0.08613069152832031, 0.08599961853027344, 0.08535244750976563, 0.0853544921875, 0.08618495941162109, 0.08482406616210937, 0.08524288177490234, 0.08417894744873047, 0.08835072326660157, 0.08645426940917969, 0.08594847869873047, 0.08594425964355469, 0.08639794921875, 0.0862003173828125, 0.17308876037597656, 0.0861470718383789, 0.08535763549804687, 0.08573228454589844, 0.08777318572998047, 0.0860579833984375, 0.08572621154785157, 0.086181884765625, 0.08619213104248047, 0.08773222351074218, 0.08648397064208985, 0.08551526641845703, 0.08338022613525391, 0.08344371032714844, 0.08378880310058594, 0.08320511627197266, 0.08437350463867188, 0.08477490997314453, 0.08600780487060547, 0.08639385223388672, 0.08647065734863281, 0.08623411560058594, 0.08597503662109375, 0.08415744018554687, 0.08344166564941406, 0.08589516448974609, 0.08594534301757813, 0.08557977294921874, 0.08637545776367188, 0.08613065338134765, 0.08599756622314453, 0.08644300842285156, 0.08653823852539062, 0.08349593353271484, 0.08332083129882813, 0.08336589050292968, 0.08353075408935547, 0.08340172576904296, 0.08323072052001954, 0.08343756866455078, 0.08349286651611328, 0.0832890853881836, 0.0835041275024414, 0.08353897857666015, 0.08367203521728515, 0.08369664001464844, 0.08330963134765625, 0.08333510589599609, 0.08366182708740234, 0.08354617309570313, 0.08324396514892578, 0.08303308868408203, 0.08764422607421875, 0.084233154296875, 0.08548761749267578, 0.08656588745117187, 0.08594841766357422, 0.08611126708984375, 0.08648700714111328, 0.08606719970703125, 0.08635289764404297, 0.08466329956054687, 0.08477286529541016, 0.17457868957519532, 0.08449132537841797, 0.08596371459960937, 0.08606515502929687, 0.08642355346679688, 0.08605696105957031, 0.08621673583984375, 0.08608969879150391, 0.08441241455078125, 0.08336179351806641, 0.08336179351806641, 0.08567501068115234, 0.08582860565185547, 0.083346435546875, 0.08477388763427735, 0.0841707534790039, 0.08664268493652344, 0.08617881774902343, 0.0860057601928711, 0.08586239624023438, 0.08551219177246094, 0.08600678253173828, 0.08574361419677734, 0.08587980651855469, 0.0835389404296875, 0.08575692749023438, 0.08598118591308594, 0.085970947265625, 0.08765235137939453, 0.08452095794677734, 0.08327577972412109, 0.08366387176513672, 0.08551423645019532, 0.08583475494384765, 0.08568831634521484, 0.08532889556884765, 0.0860057601928711, 0.08593408203125, 0.08623308563232422, 0.08616754913330078, 0.08630989074707031, 0.08582454681396484, 0.08615420532226563, 0.08506265258789063, 0.08292249298095702, 0.08326246643066407, 0.08425062561035156, 0.08737894439697266, 0.08324813079833984, 0.08346521759033203, 0.08319385528564453, 0.08310169219970703, 0.08324198150634765, 0.08323072052001954, 0.08317030334472657, 0.08586962890625, 0.08609273529052734, 0.08769741058349609, 0.08657305908203125, 0.08599244689941406, 0.08583679962158203, 0.08583782196044921, 0.08327680206298828, 0.17533644104003906, 0.08587776184082031, 0.08605081939697265, 0.0857548828125, 0.08574873352050781, 0.0859535369873047, 0.08613581085205078, 0.08587161254882812, 0.08592998504638671, 0.08673795318603515, 0.08629654693603515, 0.08595455932617188, 0.08510157012939454, 0.08576204681396485, 0.08615424346923828, 0.08628736114501953, 0.08378880310058594, 0.08452607727050782, 0.08607129669189453, 0.0862586898803711, 0.08617062377929688, 0.08422502136230468, 0.08376422119140625, 0.0832890853881836, 0.08342630767822265, 0.0846397476196289, 0.08346630096435546, 0.08343545532226562, 0.08319590759277344, 0.08344371032714844, 0.08367001342773438, 0.0833259506225586, 0.0832511978149414, 0.08327372741699218, 0.0835389404296875, 0.0833966064453125, 0.08333004760742188, 0.0835758056640625, 0.08591462707519532, 0.08338739013671875, 0.08385945892333985, 0.08679219055175781, 0.0835594253540039, 0.08360345458984375, 0.08362598419189453, 0.08332288360595703, 0.08334137725830078, 0.08309037017822266, 0.08337209320068359, 0.08316102600097657, 0.08313855743408204, 0.08323993682861328, 0.08327168273925781, 0.08664064025878906, 0.08615526580810547, 0.08606719970703125, 0.08610099029541016, 0.08608665466308593, 0.08620543670654297, 0.0832573471069336, 0.08329011535644532, 0.08356147003173828, 0.08407449340820312]",tokens/s,11.50335704469939,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-160m,EleutherAI/pythia-160m,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -337,26 +569,26 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3544, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained config = cls._autoset_attn_implementation( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1463, in _autoset_attn_implementation + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation config = cls._check_and_enable_sdpa( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1619, in _check_and_enable_sdpa + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa raise ValueError( -ValueError: GPTNeoForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 -4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neox-20b,EleutherAI/gpt-neox-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Salesforce/codegen-16B-nl,Salesforce/codegen-16B-nl,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -366,26 +598,26 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3544, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained config = cls._autoset_attn_implementation( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1463, in _autoset_attn_implementation + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation config = cls._check_and_enable_sdpa( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1619, in _check_and_enable_sdpa + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa raise ValueError( -ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` +ValueError: CodeGenForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 -4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-34B,01-ai/Yi-34B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-MoE-A2.7B,Qwen/Qwen1.5-MoE-A2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -395,32 +627,37 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained - hf_quantizer.postprocess_model(model) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model - return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading - model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 432, in post_init_awq_exllama_modules - model = exllama_post_init(model) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init - submodule.post_init() - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init - self.q4 = exl_ext.make_q4( -NameError: name 'exl_ext' is not defined + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3640, in from_pretrained + hf_quantizer.preprocess_model( + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 182, in preprocess_model + return self._process_model_before_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 85, in _process_model_before_weight_loading + model, has_been_replaced = replace_with_awq_linear( + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 178, in replace_with_awq_linear + _, has_been_replaced = replace_with_awq_linear( + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 178, in replace_with_awq_linear + _, has_been_replaced = replace_with_awq_linear( + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 178, in replace_with_awq_linear + _, has_been_replaced = replace_with_awq_linear( + [Previous line repeated 1 more time] + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 165, in replace_with_awq_linear + model._modules[name] = target_cls( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemm.py"", line 102, in __init__ + assert out_features % (32 // self.w_bit) == 0 +AssertionError -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 -4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-125m,EleutherAI/gpt-neo-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-1.3b,EleutherAI/pythia-1.3b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -430,26 +667,26 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3544, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained config = cls._autoset_attn_implementation( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1463, in _autoset_attn_implementation + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation config = cls._check_and_enable_sdpa( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1619, in _check_and_enable_sdpa + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa raise ValueError( -ValueError: GPTNeoForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 -4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-6B,01-ai/Yi-6B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-2.7b,EleutherAI/pythia-2.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -459,32 +696,28 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained - hf_quantizer.postprocess_model(model) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model - return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading - model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 432, in post_init_awq_exllama_modules - model = exllama_post_init(model) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init - submodule.post_init() - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init - self.q4 = exl_ext.make_q4( -NameError: name 'exl_ext' is not defined + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 -4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-2.7B,EleutherAI/gpt-neo-2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-14B,Qwen/Qwen1.5-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,3984.965632,12732.33408,0.0,12085.886976,11337.364992,s,10,10.92237109375,1.092237109375,0.0021117743883837157,1.0918812255859374,1.0945805541992188,1.0956890197753906,1.0965757922363282,"[1.0967974853515625, 1.094334228515625, 1.0893026123046874, 1.090765625, 1.0907347412109376, 1.09039990234375, 1.0912408447265625, 1.09275048828125, 1.0935235595703126, 1.0925216064453125]",tokens/s,234.38134247836382,kWh,1.2875690162181854e-05,7.054504655134223e-06,6.32269116926043e-05,8.315710650992037e-05,tokens/kWh,3078510.1928655975,MB,3984.965632,12732.33408,0.0,12085.886976,11686.79936,s,10,637.5050507812499,63.75050507812499,0.0049395574759148195,63.749460937500004,63.757564453125,63.759594726562504,63.7612189453125,"[63.74835546875, 63.7478046875, 63.75041796875, 63.7437265625, 63.75711328125, 63.74664453125, 63.761625, 63.75044140625, 63.74853515625, 63.75038671875]",tokens/s,0.9882274645949038,kWh,0.0007525312550531495,0.00041245463154615205,0.0037184816692274027,0.004883467555826704,tokens/kWh,12900.669305119396,,s,629,646.3186791381836,1.027533671125888,0.12973576665422035,1.0118389892578126,1.012325793457031,1.0126516357421875,2.103177431640625,"[1.0117550048828126, 1.0119987182617187, 1.0120816650390625, 1.0120765991210938, 1.0125403442382812, 1.011704833984375, 1.0115768432617187, 1.0117868041992188, 1.01175390625, 1.0114559936523437, 1.0114959106445311, 1.0115798950195312, 1.0117611694335937, 1.01186865234375, 1.0123243408203124, 1.0116085815429687, 1.01121435546875, 1.0115245971679687, 1.0116761474609375, 1.011673095703125, 1.0115952758789062, 1.0115348510742188, 1.011820556640625, 1.0118010864257811, 1.0121104125976563, 1.0117160034179689, 1.0116536254882813, 1.0114908447265625, 1.0122445068359376, 1.0119515991210937, 1.0117140502929687, 1.011957763671875, 1.0116392822265625, 1.0119588012695313, 1.0129981689453125, 1.011984375, 1.0124021606445313, 1.01186865234375, 1.0122096557617188, 1.0120274047851563, 1.0117069091796875, 1.0121103515625, 1.0121615600585938, 1.0125332641601563, 1.0121226196289062, 1.01194140625, 1.0114242553710937, 1.0118287353515625, 1.011631103515625, 1.0119935913085938, 1.0117672729492186, 1.0120878295898437, 1.0116167602539063, 1.0122476196289063, 1.0120057983398438, 1.0121021728515625, 1.011689453125, 1.0124891967773437, 1.0117969970703125, 1.0116424560546875, 1.0114559326171875, 1.0116290283203124, 2.106623046875, 1.0121912841796874, 1.012216796875, 1.0120878295898437, 1.01224755859375, 1.0123724365234374, 1.0121103515625, 1.011904541015625, 1.0120908813476563, 1.0119536743164061, 1.0122240600585937, 1.0117017211914063, 1.0118287353515625, 1.0117017822265626, 1.0116792602539062, 1.0114119873046874, 1.0114088745117187, 1.01127783203125, 1.011852294921875, 1.01174169921875, 1.011968994140625, 1.0117713623046876, 1.0116188354492188, 1.0120038452148437, 1.011915771484375, 1.0112901000976562, 1.0112747802734374, 1.0113668823242188, 1.01144677734375, 1.0113177490234375, 1.0119198608398436, 1.0120233154296876, 1.0116013793945313, 1.0115082397460937, 1.0116761474609375, 1.0115245971679687, 1.0114263305664062, 1.0112634887695313, 1.011557373046875, 1.0114109497070312, 1.0121984252929688, 1.0121513061523437, 1.01176318359375, 1.0114232177734375, 1.011726318359375, 1.011304443359375, 1.0149703979492188, 1.0115143432617189, 1.012031494140625, 1.0119803466796875, 1.0133125, 1.0118717651367188, 1.0115481567382814, 1.0116792602539062, 1.0120653076171875, 1.0115502319335938, 1.0118072509765625, 1.0117089233398437, 1.0117550048828126, 1.0120355834960937, 1.0122034912109374, 1.011757080078125, 1.0119905395507813, 2.102846435546875, 1.0113546142578125, 1.0115645141601564, 1.011766357421875, 1.0118368530273438, 1.011746826171875, 1.0118656005859374, 1.0115020751953125, 1.0114826049804688, 1.0119618530273438, 1.01157373046875, 1.0117109985351562, 1.0118194580078126, 1.0120345458984374, 1.0119188232421874, 1.0118768920898438, 1.0121174926757812, 1.0112481079101563, 1.011708984375, 1.011843017578125, 1.011999755859375, 1.0120202026367187, 1.0120222778320314, 1.0137774047851562, 1.011525634765625, 1.0120479125976563, 1.011984375, 1.0118092651367188, 1.0117805786132812, 1.0120990600585937, 1.0119854125976562, 1.01212158203125, 1.0122670288085938, 1.0120601806640626, 1.0118174438476562, 1.0118410034179688, 1.0115552978515625, 1.011240966796875, 1.011661865234375, 1.0114876708984375, 1.011441650390625, 1.01127783203125, 1.0116198120117188, 1.0113597412109374, 1.0116414184570313, 1.0120242309570313, 1.0122291259765626, 1.013486572265625, 1.0116536254882813, 1.0121564331054687, 1.0123980712890626, 1.0119321899414062, 1.012010009765625, 1.0122987670898438, 1.0121973876953125, 1.0126295166015624, 1.0126663818359376, 1.0119352416992187, 1.0116690063476563, 1.0120150756835937, 1.0120540771484374, 1.01182666015625, 1.0116351928710938, 2.10330615234375, 1.0118348999023437, 1.0117386474609376, 1.0117161254882812, 1.01167822265625, 1.0121646118164063, 1.0119137573242187, 1.01201611328125, 1.0119761962890625, 1.0122608032226563, 1.0121830444335937, 1.0119669799804687, 1.0134896850585937, 1.0120386352539064, 1.0123622436523438, 1.0113034057617187, 1.0113535766601562, 1.011219482421875, 1.0112798461914063, 1.0111006469726562, 1.0113710327148437, 1.0122311401367188, 1.0115225830078125, 1.0115430297851562, 1.0117457885742187, 1.0114498291015626, 1.0119761962890625, 1.0120970458984375, 1.0119505615234374, 1.0119619140625, 1.0121400146484374, 1.0117672729492186, 1.0114600830078124, 1.0114866943359375, 1.011557373046875, 1.011900390625, 1.0118594360351563, 1.01203662109375, 1.0116956176757812, 1.0115460815429687, 1.0117755126953125, 1.0116608276367187, 1.01161474609375, 1.0111918334960937, 1.0146948852539062, 1.01151953125, 1.011521484375, 1.0112010498046875, 1.01165771484375, 1.0114376220703125, 1.01186865234375, 1.0118800048828125, 1.0115624389648437, 1.0114273071289062, 1.011641357421875, 1.0113966064453126, 1.0116741333007813, 1.011451904296875, 1.0115655517578126, 1.011557373046875, 1.012046875, 1.0120038452148437, 1.011968017578125, 2.10362060546875, 1.0117232666015625, 1.0121318359375, 1.0118696899414064, 1.0120601806640626, 1.0119393310546876, 1.012031494140625, 1.0123212890625, 1.0119556884765626, 1.01214208984375, 1.0124891967773437, 1.0119710693359374, 1.0119178466796874, 1.0113935546875, 1.0116761474609375, 1.0115359497070313, 1.011611572265625, 1.01361767578125, 1.0116761474609375, 1.0114703369140625, 1.0116137084960937, 1.011589111328125, 1.0115932006835937, 1.0114774780273437, 1.0115348510742188, 1.0114713745117188, 1.0118184814453124, 1.0116055297851563, 1.01148876953125, 1.0116792602539062, 1.0118379516601563, 1.0119649047851562, 1.0117376708984376, 1.011568603515625, 1.0120294189453125, 1.01218505859375, 1.0119823608398437, 1.0122332153320313, 1.011926025390625, 1.0118748168945313, 1.0124400634765625, 1.01216357421875, 1.011631103515625, 1.0115266723632812, 1.0125834350585938, 1.0121605224609376, 1.0119229736328126, 1.0119270629882813, 1.01148876953125, 1.0118287353515625, 1.011999755859375, 1.011862548828125, 1.0116915283203125, 1.0118143920898437, 1.011768310546875, 1.0122393798828124, 1.0125045776367188, 1.0162032470703124, 1.0122127075195313, 1.0123765869140624, 1.0122915649414062, 1.012389892578125, 1.0124400634765625, 2.102681640625, 1.0123212280273437, 1.0120990600585937, 1.0122465209960938, 1.0117007446289064, 1.012094970703125, 1.0112553100585937, 1.0112625122070313, 1.0114201049804687, 1.0113854370117188, 1.01134228515625, 1.0119854125976562, 1.0117590942382813, 1.0113126220703126, 1.0112696533203125, 1.011483642578125, 1.0116823120117187, 1.0114539794921875, 1.0115082397460937, 1.0114959106445311, 1.0115481567382814, 1.0120242919921876, 1.0115635375976562, 1.0115850219726563, 1.0117254028320313, 1.0121840209960937, 1.0118276977539062, 1.0116751098632812, 1.0118184814453124, 1.0114385986328125, 1.0118215942382813, 1.0118092651367188, 1.0117406616210938, 1.011473388671875, 1.0117386474609376, 1.0118164672851562, 1.01148876953125, 1.0114867553710938, 1.0149037475585938, 1.0118359375, 1.0122781982421876, 1.0117642211914062, 1.0119249877929688, 1.011294189453125, 1.0117027587890626, 1.0116690063476563, 1.0123939819335936, 1.0120908813476563, 1.0122567749023437, 1.0120653076171875, 1.0122485961914063, 1.012369384765625, 1.012421630859375, 1.0119987182617187, 1.0126663818359376, 1.0115481567382814, 1.0117089233398437, 1.0119823608398437, 1.0117867431640626, 1.0115614624023437, 1.0122546997070312, 1.0117652587890624, 1.012073486328125, 2.104281005859375, 1.011789794921875, 1.0118133544921875, 1.0122199096679687, 1.01174169921875, 1.0117130126953124, 1.0126079711914062, 1.0127093505859375, 1.0125383911132813, 1.0122393798828124, 1.01216357421875, 1.0123182373046875, 1.0122199096679687, 1.0121246948242189, 1.0119669799804687, 1.0121113891601563, 1.0128097534179688, 1.01222705078125, 1.012335693359375, 1.0115757446289062, 1.0118441162109375, 1.0118328247070312, 1.011631103515625, 1.0136719360351563, 1.0121174926757812, 1.01250048828125, 1.012552734375, 1.0127390747070313, 1.0122342529296875, 1.0118225708007813, 1.0121768798828126, 1.0121298217773438, 1.0121860961914062, 1.0118389892578126, 1.012316162109375, 1.0120714111328124, 1.0119198608398436, 1.0119669799804687, 1.0115543823242188, 1.0116760864257812, 1.0120601806640626, 1.0119342041015624, 1.0120653076171875, 1.0115552978515625, 1.0121380004882812, 1.0123212890625, 1.0120806274414063, 1.0117672729492186, 1.0117805786132812, 1.0118287353515625, 1.0118615112304687, 1.0117703857421876, 1.0114365234375, 1.0116065063476563, 1.0118512573242187, 1.0121676635742187, 1.0124697875976563, 1.0122608642578126, 1.0120376586914062, 1.0119403686523438, 1.0120181884765624, 1.0120386352539064, 1.01167822265625, 2.1052109375, 1.0120621948242188, 1.01188916015625, 1.0118225708007813, 1.01174169921875, 1.012010009765625, 1.0117171020507814, 1.0117847290039061, 1.0114641723632813, 1.01161474609375, 1.0114928588867187, 1.0118482055664062, 1.0115112915039062, 1.0117078857421875, 1.0117007446289064, 1.011862548828125, 1.0131630249023438, 1.0117345581054686, 1.0117489013671874, 1.0116443481445312, 1.0115972900390624, 1.0121307983398438, 1.01167822265625, 1.0120202026367187, 1.0114754638671875, 1.012262939453125, 1.0119618530273438, 1.0120448608398438, 1.01155224609375, 1.0120181884765624, 1.0120068969726563, 1.0124298095703126, 1.0118533325195314, 1.0121185302734375, 1.0118779296875, 1.0119566650390626, 1.0118379516601563, 1.0119434204101563, 1.0118062133789063, 1.0120775756835938, 1.0119024658203124, 1.0125721435546875, 1.0118563842773438, 1.0118348999023437, 1.0117273559570312, 1.01222705078125, 1.011905517578125, 1.0115379028320313, 1.0121062622070311, 1.0116116333007812, 1.011979248046875, 1.012168701171875, 1.0120171508789062, 1.01224658203125, 1.0122731323242187, 1.0115819091796876, 1.0118994140625, 1.0117642211914062, 1.0116351928710938, 1.011962890625, 1.0119782104492188, 1.01182568359375, 1.012173828125, 2.105998291015625, 1.0114345092773438, 1.0115174560546876, 1.0114590454101562, 1.0117990112304687, 1.011937255859375, 1.0119331665039062, 1.011800048828125, 1.0117058715820313, 1.0113648681640626, 1.0117181396484376, 1.0116608276367187, 1.0114611206054687, 1.0114600830078124, 1.0114508666992188, 1.0114754638671875, 1.0120274047851563, 1.0133237915039062, 1.0123274536132814, 1.0115994262695311, 1.0119392700195313, 1.011694580078125, 1.0119700317382812, 1.0124237060546875, 1.0117488403320312, 1.0116751098632812, 1.0119721069335939, 1.011651611328125, 1.01169970703125, 1.0113812255859376, 1.0121625366210938, 1.0115175170898438, 1.0117620849609374, 1.011493896484375, 1.0118062133789063, 1.0119086303710938, 1.01222607421875, 1.0115870971679688, 1.0119854125976562, 1.0115471801757812, 1.0122720336914062, 1.0117857055664063, 1.0124462280273439, 1.0115880737304688, 1.0122782592773438, 1.0121298217773438, 1.0129663696289062, 1.011736572265625, 1.0120570678710938, 1.0118963623046875, 1.0122628784179688, 1.0121093139648438, 1.0120888061523436, 1.011789794921875, 1.0123253784179687, 1.0119198608398436, 1.0131046142578124, 1.0120990600585937, 1.0116629028320312, 1.0112337646484375, 1.0118236083984375, 1.0113085327148437, 1.0116392822265625, 2.104349609375, 1.0117294311523437, 1.0120775756835938, 1.0116546630859375, 1.01159423828125, 1.0118696899414064, 1.0115491943359376, 1.0113505249023438, 1.0112051391601562, 1.0120068969726563, 1.0118911743164063, 1.011857421875, 1.01180419921875, 1.0114385375976562, 1.011646484375, 1.011788818359375, 1.0118615112304687, 1.0118656005859374, 1.0119014282226562, 1.0118348999023437, 1.011794921875, 1.011900390625, 1.0119556884765626, 1.0122608642578126, 1.0124257202148437, 1.0121984252929688, 1.0144102172851563, 1.011820556640625, 1.0118195190429689, 1.0117386474609376, 1.0117069091796875, 1.0122393798828124, 1.011646484375, 1.0119147338867187, 1.011462158203125, 1.012073486328125, 1.0115911865234375, 1.011684326171875, 1.0119321899414062, 1.0118062133789063, 1.0121994018554688, 1.0123673706054688, 1.0117744750976563, 1.011873779296875, 1.0116044921875, 1.011989501953125, 1.0118911743164063, 1.01174169921875, 1.011662841796875, 1.0115194702148438, 1.0117089233398437, 1.0126878662109375, 1.0129653930664062, 1.0122413940429686, 1.0118389892578126, 1.0119505615234374, 1.0117611694335937, 1.01174169921875, 1.0118338623046874, 1.0117755126953125, 1.0121144409179688, 1.0123131103515626, 1.0117294311523437]",tokens/s,0.973204117879934,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen2-beta-14B,Qwen/Qwen2-beta-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,3934.998528,12732.33408,0.0,12085.886976,11337.370624,s,10,10.913714233398439,1.0913714233398437,0.0019187724556758634,1.0909278564453126,1.0936456420898437,1.0942043518066407,1.0946513195800782,"[1.0947630615234376, 1.093521484375, 1.0894732666015625, 1.0895545654296874, 1.089643798828125, 1.0893114013671874, 1.0899912109375, 1.091864501953125, 1.093216552734375, 1.0923743896484375]",tokens/s,234.56725595451456,kWh,1.2859468128946092e-05,7.046563530275307e-06,6.250710556117145e-05,8.241313722039284e-05,tokens/kWh,3106300.871854854,MB,3934.998528,12732.33408,0.0,12085.886976,11686.804992,s,10,636.61697265625,63.661697265625,0.007269593222884917,63.66015234375,63.672558203125,63.6732771484375,63.6738523046875,"[63.6561171875, 63.66687109375, 63.65788671875, 63.66241796875, 63.65402734375, 63.67399609375, 63.6723984375, 63.656875, 63.65174609375, 63.66463671875]",tokens/s,0.9896060379467405,kWh,0.0007518513685133722,0.0004120810364553653,0.003630345543163038,0.004794277948131776,tokens/kWh,13140.664909624129,,s,629,645.425664550782,1.0261139341029908,0.12959104210993952,1.0104422607421875,1.0109677490234374,1.0113603271484375,2.100010478515625,"[1.0103787231445311, 1.0104750366210937, 1.010914306640625, 1.0098585815429688, 1.0104279174804687, 1.010324462890625, 1.0105446166992187, 1.010070556640625, 1.0100776977539063, 1.0105692138671876, 1.0102251586914062, 1.0103941040039062, 1.0105692138671876, 1.0099476318359375, 1.010234375, 1.0106050415039063, 1.0103367919921875, 1.0100828247070313, 1.0105968627929687, 1.0105446166992187, 1.0105169677734376, 1.0106787719726562, 1.0105886840820313, 1.0103572387695312, 1.0107811889648437, 1.010609130859375, 1.010071533203125, 1.0102784423828124, 1.0099865112304687, 1.010282470703125, 1.0107012939453126, 1.0107606811523437, 1.0109655151367187, 1.0103797607421876, 1.0103644409179688, 1.0103593139648437, 1.0102118530273438, 1.0102118530273438, 1.0101124877929688, 1.0109286499023438, 1.0103910522460937, 1.0106552124023438, 1.011262451171875, 1.0107944946289062, 1.0104873046875, 1.0106204223632813, 1.0099425048828126, 1.0105272216796874, 1.00997119140625, 1.0104494018554688, 1.0101258544921874, 1.010229248046875, 1.0103654174804688, 1.0102947998046874, 1.0099885864257812, 1.0105845947265626, 1.0102702026367187, 1.0102896728515625, 1.0102702026367187, 1.0103797607421876, 1.0103336791992188, 1.0109767456054688, 2.1042431640625, 1.0108385009765626, 1.0106071166992188, 1.0101176147460937, 1.0099619750976563, 1.01039306640625, 1.0101083984375, 1.0104063720703125, 1.0112655639648438, 1.0104002685546876, 1.0102466430664063, 1.0105128784179687, 1.0102671508789063, 1.0101801147460938, 1.0101739501953124, 1.0103828735351563, 1.0103910522460937, 1.0109398803710938, 1.0110330810546875, 1.0103602905273437, 1.0109020385742187, 1.0112982788085938, 1.010745361328125, 1.0104309692382814, 1.0102783813476564, 1.010271240234375, 1.0101309204101563, 1.0107698974609376, 1.0114027709960938, 1.01064501953125, 1.0102230834960937, 1.0105385131835938, 1.0106644287109374, 1.0104248046875, 1.0108897094726563, 1.010545654296875, 1.0105814819335937, 1.0107955322265625, 1.0110156860351562, 1.0106214599609376, 1.01065625, 1.01066650390625, 1.0106920776367188, 1.0103101196289062, 1.0108272705078125, 1.0104053955078125, 1.0105886840820313, 1.010440185546875, 1.0111314086914063, 1.010808837890625, 1.0105968627929687, 1.010450439453125, 1.0106644287109374, 1.010177001953125, 1.0106286010742187, 1.01064599609375, 1.01049853515625, 1.0107146606445312, 1.0105742797851562, 1.0107597045898438, 1.0104780883789062, 1.01089794921875, 1.0108590087890625, 2.10035302734375, 1.0105681762695313, 1.0103634033203126, 1.0108939819335938, 1.0105936279296874, 1.01033984375, 1.0102159423828125, 1.0105128784179687, 1.0100869140625, 1.0100756225585938, 1.0101299438476563, 1.0103162841796876, 1.0100142211914063, 1.010387939453125, 1.0102691650390625, 1.0100736083984374, 1.0103839111328126, 1.0102691650390625, 1.0101923828125, 1.0100633544921875, 1.0100695190429687, 1.0103623657226561, 1.0100408325195311, 1.0103367919921875, 1.0102200317382812, 1.0101944580078126, 1.0105272216796874, 1.0104760131835937, 1.010466796875, 1.0102435913085936, 1.010566162109375, 1.010461669921875, 1.0103326416015626, 1.0106859741210938, 1.0104954833984374, 1.0103705444335938, 1.0109921264648438, 1.0105057373046875, 1.0102046508789062, 1.01007666015625, 1.0101852416992188, 1.0097950439453125, 1.0098770141601563, 1.0107269287109375, 1.0103705444335938, 1.0104033203125, 1.0112020263671875, 1.0103848876953125, 1.0109419555664063, 1.0105897216796875, 1.0109766845703125, 1.010745361328125, 1.0108344116210937, 1.011294189453125, 1.0108703002929686, 1.0104647827148439, 1.0111211547851562, 1.010250732421875, 1.0104473876953124, 1.0103490600585938, 1.0106961669921875, 1.010408447265625, 1.0106736450195313, 2.09991064453125, 1.0102855834960938, 1.010808837890625, 1.0101883544921875, 1.0098963623046875, 1.0099046630859374, 1.0102015380859375, 1.0112860107421875, 1.0114365234375, 1.0114171142578126, 1.011398681640625, 1.0108375854492186, 1.011353515625, 1.0101831665039063, 1.0099578857421876, 1.0098401489257813, 1.0104351806640626, 1.0098911743164063, 1.01047705078125, 1.0109163818359375, 1.0101514282226562, 1.0102528076171875, 1.0105303344726562, 1.010460693359375, 1.0104515380859376, 1.0105640258789061, 1.0108231811523438, 1.0104279174804687, 1.0102528076171875, 1.0102149047851563, 1.0106552124023438, 1.0108037109375, 1.0104227905273437, 1.010044921875, 1.0103521118164063, 1.0104022827148438, 1.0106019897460938, 1.0105252075195312, 1.0107361450195314, 1.0110341186523437, 1.0109389038085939, 1.0102589721679687, 1.0107658081054687, 1.01032958984375, 1.0109112548828125, 1.0104422607421875, 1.0104935302734375, 1.0104278564453124, 1.0109645385742188, 1.0107493896484374, 1.01064501953125, 1.0110023803710937, 1.0106593017578125, 1.0102763671875, 1.01026611328125, 1.0103255004882812, 1.0103726196289062, 1.010171875, 1.0105169677734376, 1.0103091430664062, 1.0100910034179686, 1.0103613891601562, 1.010420654296875, 2.10003857421875, 1.0099517211914062, 1.010044921875, 1.0100490112304688, 1.010212890625, 1.0100838623046875, 1.0100510864257812, 1.0103336791992188, 1.0101473388671875, 1.0099937133789063, 1.0099507446289062, 1.0100213623046874, 1.0101913452148437, 1.010255859375, 1.0110208129882812, 1.0101217041015624, 1.0101422119140624, 1.0106214599609376, 1.0106583251953125, 1.0105067749023438, 1.0107003173828124, 1.0106572875976563, 1.0101596069335939, 1.0103726196289062, 1.0107791137695312, 1.0102958374023439, 1.0105692138671876, 1.010567138671875, 1.0101862182617187, 1.0100357055664062, 1.0101422119140624, 1.0100562133789062, 1.0101636962890626, 1.0102886352539062, 1.0103009033203125, 1.0100858764648437, 1.0102108154296876, 1.0106234741210938, 1.0103255004882812, 1.0102271728515626, 1.0103224487304687, 1.0102650756835938, 1.0104852294921876, 1.0101156005859375, 1.0108948364257813, 1.0105333862304688, 1.0106009521484376, 1.010366455078125, 1.0103654174804688, 1.0105374755859375, 1.0105303344726562, 1.010418701171875, 1.0103674926757813, 1.0106214599609376, 1.0106808471679687, 1.0101104736328126, 1.0104002685546876, 1.010713623046875, 1.010361328125, 1.0107811889648437, 1.0112348022460937, 1.0109830322265625, 1.010874267578125, 2.099938232421875, 1.0099097900390626, 1.010356201171875, 1.0107750244140625, 1.0106531982421876, 1.0112010498046875, 1.0109235229492188, 1.0112593994140624, 1.011103759765625, 1.0114088745117187, 1.0117509155273436, 1.0115809326171874, 1.0110750732421876, 1.0116761474609375, 1.0117703857421876, 1.01070849609375, 1.0111211547851562, 1.0108140869140625, 1.010822021484375, 1.0101463012695313, 1.010524169921875, 1.010092041015625, 1.0108528442382811, 1.0102384643554687, 1.0105374755859375, 1.0103961791992186, 1.0102907104492187, 1.0105763549804687, 1.0101422119140624, 1.0106972045898437, 1.010524169921875, 1.0101381225585937, 1.01081396484375, 1.0101923828125, 1.010398193359375, 1.0102046508789062, 1.0100275268554688, 1.0102159423828125, 1.01076171875, 1.0107730102539063, 1.0104595947265624, 1.0105743408203125, 1.0107811889648437, 1.0105466918945312, 1.010503662109375, 1.0107811889648437, 1.0115594482421875, 1.0113648681640626, 1.0109235229492188, 1.0113966064453126, 1.0115973510742187, 1.0105558471679688, 1.0106337280273439, 1.0101053466796874, 1.0102333374023438, 1.0100643920898438, 1.0104063720703125, 1.0102118530273438, 1.0105886840820313, 1.0109163818359375, 1.01098291015625, 1.0103121948242189, 1.0104627075195312, 2.1005556640625, 1.0100828247070313, 1.0103214111328125, 1.0101217041015624, 1.0104524536132813, 1.0106480712890624, 1.009934326171875, 1.0109644775390625, 1.01163427734375, 1.011917724609375, 1.0118707275390626, 1.0117642211914062, 1.0120970458984375, 1.0110791625976563, 1.0113136596679688, 1.0106326904296874, 1.010629638671875, 1.0110320434570312, 1.0109531860351562, 1.0102200317382812, 1.0101647338867188, 1.01035107421875, 1.0101319580078125, 1.0104688720703126, 1.0105006103515626, 1.01138330078125, 1.010892822265625, 1.0108292846679687, 1.0109214477539064, 1.0098134765625, 1.0101145629882813, 1.0102262573242187, 1.0102721557617187, 1.010165771484375, 1.0108426513671875, 1.0108047485351563, 1.010640869140625, 1.0113280029296876, 1.0108528442382811, 1.010176025390625, 1.0101810913085938, 1.0100869140625, 1.0103460083007811, 1.010440185546875, 1.0107904052734376, 1.0101801147460938, 1.0104586181640625, 1.0103050537109375, 1.0111918334960937, 1.0107914428710938, 1.0108416137695313, 1.0108283081054688, 1.0106810302734375, 1.0106591186523437, 1.0115072021484375, 1.0104279174804687, 1.0108436279296875, 1.0108283081054688, 1.01055078125, 1.0102097778320311, 1.010460693359375, 1.0103121948242189, 1.0103951416015624, 2.102578125, 1.0104268798828124, 1.0102036743164062, 1.010355224609375, 1.0103173217773438, 1.0101637573242188, 1.01056201171875, 1.0103060302734375, 1.0101749877929687, 1.0104330444335938, 1.0104094848632812, 1.0101801147460938, 1.0102282104492188, 1.0102640380859376, 1.0102354125976563, 1.01035107421875, 1.0101268310546876, 1.01020263671875, 1.0103245849609375, 1.010389892578125, 1.0105927734375, 1.0102783813476564, 1.0102159423828125, 1.0109481201171875, 1.0104248046875, 1.0102097778320311, 1.0102999267578125, 1.0106634521484374, 1.01051904296875, 1.0106911010742188, 1.0106603393554687, 1.0103428955078124, 1.010208740234375, 1.0103828735351563, 1.0103501586914063, 1.0101267700195313, 1.010428955078125, 1.0099415283203126, 1.0103070678710937, 1.010746337890625, 1.0105733032226563, 1.0100193481445312, 1.0105620727539062, 1.0108969116210937, 1.010830322265625, 1.010567138671875, 1.0105261840820312, 1.0104903564453125, 1.010502685546875, 1.010763916015625, 1.0106510009765626, 1.01055078125, 1.010597900390625, 1.010534423828125, 1.01037158203125, 1.0100613403320313, 1.0104227905273437, 1.0101196899414062, 1.0103214111328125, 1.0106286010742187, 1.0109337768554687, 1.0104739990234375, 1.0107742309570313, 2.104293212890625, 1.010323486328125, 1.0104320068359376, 1.0102138671875, 1.0099435424804688, 1.0105354614257813, 1.0101022338867187, 1.0104801025390624, 1.0105220947265625, 1.0102742919921874, 1.0100582275390626, 1.0102661743164063, 1.0102158813476563, 1.0104473876953124, 1.0102271728515626, 1.0102630615234376, 1.0100213623046874, 1.0100828247070313, 1.0104155883789063, 1.01005517578125, 1.010208740234375, 1.0101239013671874, 1.0102445068359376, 1.0100910034179686, 1.0104801025390624, 1.0102210693359375, 1.0102210693359375, 1.0105426025390625, 1.0104248046875, 1.0105814819335937, 1.0103746337890624, 1.0100991821289063, 1.0102630615234376, 1.0103203735351562, 1.0105231323242188, 1.0105108642578124, 1.0102191162109375, 1.0101605224609376, 1.0104053955078125, 1.0104586181640625, 1.0102518920898438, 1.0101390380859374, 1.0103285522460939, 1.0101493530273438, 1.0104166259765626, 1.0106941528320312, 1.0107689208984374, 1.0102313232421876, 1.0105620727539062, 1.01051708984375, 1.0104871826171875, 1.01003369140625, 1.0101371459960937, 1.0099342651367187, 1.0106521606445313, 1.0104391479492187, 1.0105620727539062, 1.010513916015625, 1.010798583984375, 1.0104074096679687, 1.0103634033203126, 1.0103224487304687, 1.010681884765625, 2.103435302734375, 1.010229248046875, 1.0106736450195313, 1.0107843017578124, 1.0107432861328125, 1.0110986328125, 1.0106265869140625, 1.010587646484375, 1.0101248168945312, 1.0100542602539062, 1.0107769775390625, 1.0108426513671875, 1.0104627075195312, 1.0106603393554687, 1.0108734741210939, 1.01037255859375, 1.0107083740234375, 1.0104166259765626, 1.0103275756835937, 1.0101319580078125, 1.0107750244140625, 1.0107258911132813, 1.0107606811523437, 1.0102907104492187, 1.0104279174804687, 1.0105446166992187, 1.0105364379882813, 1.0103009643554688, 1.010494384765625, 1.0100807495117188, 1.0106644287109374, 1.0108026733398439, 1.0108375244140626, 1.010492431640625, 1.0103746337890624, 1.0101227416992187, 1.0101484375, 1.0099373168945311, 1.0103756713867187, 1.0104944458007812, 1.0109050903320314, 1.0101473388671875, 1.0104903564453125, 1.0099251098632813, 1.0105374755859375, 1.010134033203125, 1.0101810913085938, 1.0101248168945312, 1.0104063720703125, 1.0102702026367187, 1.0105569458007813, 1.0107965698242187, 1.01083544921875, 1.0107893676757813, 1.01082421875, 1.0101309204101563, 1.0103572387695312, 1.010207763671875, 1.010361328125, 1.0107053833007813, 1.0118082275390625, 1.01166796875, 1.01172021484375]",tokens/s,0.9745506485828798,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-30b,facebook/opt-30b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -494,26 +727,26 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3544, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained config = cls._autoset_attn_implementation( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1463, in _autoset_attn_implementation + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation config = cls._check_and_enable_sdpa( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1619, in _check_and_enable_sdpa + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa raise ValueError( -ValueError: GPTNeoForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` +ValueError: OPTForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 -4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciCoder-1b,Deci/DeciCoder-1b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-1.3B,EleutherAI/gpt-neo-1.3B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) - File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): - File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -521,34 +754,104 @@ ChildProcessError: Traceback (most recent call last): self.load_model_from_pretrained() File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained self.pretrained_model = self.automodel_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 558, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained - hf_quantizer.postprocess_model(model) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model - return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading - model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 432, in post_init_awq_exllama_modules - model = exllama_post_init(model) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init - submodule.post_init() - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init - self.q4 = exl_ext.make_q4( -NameError: name 'exl_ext' is not defined + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 -4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciLM-7B,Deci/DeciLM-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,databricks/dbrx-base,databricks/dbrx-base,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/databricks/dbrx-base/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aa523-62c6c43d32f792b12bc008c6;20fcdfd6-cd19-4baa-b381-01182c8f2baa) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/databricks/dbrx-base/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like databricks/dbrx-base is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,huggyllama/llama-7b,huggyllama/llama-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,2013.822976,5480.382464,0.0,4833.93536,4503.282688,s,10,5.711301574707031,0.5711301574707031,0.0015604684158887843,0.5710088500976562,0.5724867248535156,0.573356430053711,0.5740521942138671,"[0.5715228271484375, 0.5742261352539062, 0.5684718017578125, 0.5703966064453125, 0.570494873046875, 0.5693699340820313, 0.5704375610351563, 0.5720609130859375, 0.57229345703125, 0.5720274658203125]",tokens/s,448.2340787846278,kWh,6.721677934681928e-06,3.683185084507083e-06,3.147118258432934e-05,4.187604560351835e-05,tokens/kWh,6113280.189438215,MB,2014.928896,5480.382464,0.0,4833.93536,4688.699392,s,10,334.74296874999993,33.474296875,0.0037315247196552594,33.473017578124995,33.479158984375,33.4801732421875,33.4809846484375,"[33.47309765625, 33.47263671875, 33.4715390625, 33.4729375, 33.47534765625, 33.47058984375, 33.47893359375, 33.4811875, 33.4778203125, 33.46887890625]",tokens/s,1.8820410249468458,kWh,0.00039524016447641235,0.00021662581418956808,0.0018242358853134645,0.002436101863979445,tokens/kWh,25860.987560301615,,s,629,339.3508396606445,0.5395084891266209,0.06790528508400501,0.531294189453125,0.5317521484375,0.5319190551757812,1.1017576318359374,"[0.5316085815429688, 0.5315625, 0.5310873413085937, 0.5312276611328125, 0.5310341186523437, 0.5311795043945312, 0.5312860107421875, 0.53180517578125, 0.5313863525390625, 0.5315205078125, 0.5309071655273437, 0.5309808349609375, 0.5307955322265625, 0.5312429809570313, 0.530924560546875, 0.5313556518554687, 0.5308876953125, 0.531230712890625, 0.5310802001953125, 0.5317345581054688, 0.5312921752929688, 0.5313382568359375, 0.5308221435546875, 0.5310607299804687, 0.5308549194335938, 0.5315635375976563, 0.5311006469726562, 0.5312788696289062, 0.5312388916015625, 0.5312849731445313, 0.5308907470703125, 0.5311416015625, 0.5308528442382813, 0.5311743774414063, 0.531367919921875, 0.5315932006835937, 0.5311682739257813, 0.5317539672851562, 0.5317908325195313, 0.531473388671875, 0.53108837890625, 0.53146728515625, 0.531220458984375, 0.5315338134765625, 0.5310525512695312, 0.531900390625, 0.5317529296875, 0.5317069091796875, 0.53103515625, 0.53153076171875, 0.5311344604492187, 0.5311918334960938, 0.5308630981445313, 0.5311641845703124, 0.5309531860351563, 0.531577880859375, 0.5316065063476563, 0.5314017333984375, 0.5316864013671875, 0.5318963012695312, 0.5315604248046875, 0.53161279296875, 1.1062353515625, 0.5315419921875, 0.5316771850585937, 0.5312737426757812, 0.5311528930664062, 0.5314109497070313, 0.5312112426757812, 0.5310986328125, 0.5311057739257813, 0.53089892578125, 0.5317980346679687, 0.531167236328125, 0.53102490234375, 0.5308078002929687, 0.5314508666992187, 0.5309224853515625, 0.5310689086914062, 0.5308067626953125, 0.5315430297851562, 0.5309972534179688, 0.5317294311523437, 0.5312696533203125, 0.5314498291015625, 0.5310535888671875, 0.5314406127929687, 0.5309706420898438, 0.5315850219726562, 0.5310156860351563, 0.5313699951171875, 0.5310637817382813, 0.5316474609375, 0.5311948852539062, 0.53110888671875, 0.53115185546875, 0.531330078125, 0.5310965576171875, 0.5311754150390625, 0.5309706420898438, 0.5319178466796874, 0.532337646484375, 0.5315972900390625, 0.531103759765625, 0.5311692504882812, 0.5309296875, 0.5313453979492188, 0.53089794921875, 0.532917236328125, 0.5309666137695312, 0.5310463256835938, 0.5309081420898437, 0.5315809326171875, 0.5311968994140625, 0.5312440185546875, 0.5310320434570313, 0.5312368774414062, 0.5308692626953125, 0.5317509155273438, 0.5314529418945313, 0.5314232177734375, 0.5311610717773437, 0.531356689453125, 0.5314232177734375, 0.5313045043945313, 1.1011102294921875, 0.5311477661132813, 0.5314263305664062, 0.531145751953125, 0.5313720092773437, 0.5311477661132813, 0.5315952758789062, 0.5309235229492187, 0.5311928100585938, 0.531162109375, 0.5311211547851562, 0.53140478515625, 0.5317939453125, 0.53121435546875, 0.5318656005859375, 0.531483642578125, 0.5315000610351562, 0.5312716674804687, 0.5315645141601563, 0.531324951171875, 0.5316557006835938, 0.5313607788085938, 0.5315901489257813, 0.5309573364257812, 0.5312061157226563, 0.5308528442382813, 0.5311078491210938, 0.5310279541015624, 0.5312245483398438, 0.5308385009765625, 0.5313310546875, 0.53131982421875, 0.5318225708007812, 0.5312051391601562, 0.5311702880859375, 0.5310709838867187, 0.5313167114257813, 0.5311600341796875, 0.5312184448242188, 0.531252197265625, 0.5313607788085938, 0.5311314086914063, 0.5312163696289063, 0.5310596923828125, 0.531451904296875, 0.5310115966796874, 0.5313280029296875, 0.53153076171875, 0.5314334716796875, 0.5314508666992187, 0.5315317993164063, 0.5316935424804687, 0.5310279541015624, 0.5312266235351563, 0.5312102661132813, 0.531051513671875, 0.5313505249023438, 0.5309439697265625, 0.53108837890625, 0.5311426391601562, 0.5314908447265625, 0.5311590576171875, 0.5316188354492187, 1.1020093994140625, 0.5310126342773438, 0.5312286987304687, 0.5310289916992188, 0.531694580078125, 0.5315460815429688, 0.5315625, 0.5309337768554687, 0.5313526000976563, 0.5313218383789062, 0.5313914794921875, 0.53136279296875, 0.5317498779296875, 0.5313587036132813, 0.5314078979492187, 0.531162109375, 0.5312286987304687, 0.531251220703125, 0.5314600830078124, 0.5311160278320313, 0.53119384765625, 0.5310310668945313, 0.5316546630859375, 0.531409912109375, 0.531567626953125, 0.5313894653320312, 0.5315020751953125, 0.531135498046875, 0.5310525512695312, 0.5308528442382813, 0.5313935546875, 0.5313812255859375, 0.5312890625, 0.530966552734375, 0.531019775390625, 0.5310494995117188, 0.5313024291992188, 0.5313894653320312, 0.531294189453125, 0.5311016845703125, 0.5310525512695312, 0.5311856689453125, 0.531736572265625, 0.531641357421875, 0.5321881713867187, 0.5313126220703125, 0.5313751220703125, 0.5310965576171875, 0.531794921875, 0.5309716186523438, 0.5311610717773437, 0.5310167236328125, 0.5313290405273438, 0.5308538818359375, 0.5310453491210938, 0.5313873901367188, 0.5312819213867187, 0.5309849853515625, 0.531398681640625, 0.5310084838867187, 0.5321144409179688, 0.531578857421875, 0.5316700439453125, 1.10214453125, 0.530830322265625, 0.5315020751953125, 0.5308692626953125, 0.5310894165039063, 0.5316024169921875, 0.5316566772460938, 0.5311129760742187, 0.5313003540039063, 0.5309224853515625, 0.5312870483398437, 0.5310771484375, 0.5316566772460938, 0.5308528442382813, 0.5312696533203125, 0.5311334228515625, 0.5315010375976562, 0.5314529418945313, 0.5316433715820312, 0.5310289916992188, 0.5314703369140625, 0.5313914794921875, 0.5314805908203125, 0.5312102661132813, 0.5313157348632812, 0.5312808837890625, 0.5316044921875, 0.5312051391601562, 0.5314078979492187, 0.5316720581054688, 0.5313822631835937, 0.5311006469726562, 0.53110888671875, 0.5308703002929688, 0.5316167602539063, 0.5314221801757812, 0.531800048828125, 0.5311160278320313, 0.5314561157226563, 0.531411865234375, 0.5317734375, 0.5317805786132812, 0.5310750732421875, 0.5310975952148438, 0.53131982421875, 0.53121435546875, 0.531240966796875, 0.530924560546875, 0.5315112915039063, 0.5309634399414063, 0.53119384765625, 0.5311242065429688, 0.5314119873046875, 0.5314744262695312, 0.5316505737304688, 0.5320745239257813, 0.531435546875, 0.53260595703125, 0.5313802490234375, 0.5314990234375, 0.531493896484375, 0.531061767578125, 0.5313065185546875, 1.1010294189453125, 0.5310013427734375, 0.53124609375, 0.5309849853515625, 0.5314959106445313, 0.5311447143554687, 0.531198974609375, 0.531135498046875, 0.5317611694335938, 0.5310453491210938, 0.5312368774414062, 0.5310105590820312, 0.5315205688476563, 0.5312000122070313, 0.531442626953125, 0.5309010009765625, 0.53104638671875, 0.5309081420898437, 0.5313760986328125, 0.5308211059570312, 0.5312214965820312, 0.5310167236328125, 0.5315963134765626, 0.531188720703125, 0.5312471313476562, 0.5308661499023437, 0.5318123779296875, 0.5314027709960938, 0.5315277099609375, 0.5311959228515625, 0.5315491943359375, 0.5311867065429687, 0.5311385498046876, 0.5308538818359375, 0.5313402709960937, 0.5308897094726562, 0.53110888671875, 0.5310873413085937, 0.532853759765625, 0.5313351440429688, 0.531751953125, 0.5309685668945312, 0.5313668823242188, 0.5309788208007813, 0.5314406127929687, 0.5319198608398438, 0.5315972900390625, 0.5316341552734375, 0.531399658203125, 0.5313290405273438, 0.5312604370117188, 0.531052490234375, 0.5311856689453125, 0.5311262817382812, 0.5311273193359375, 0.5309450073242188, 0.5313003540039063, 0.5310310668945313, 0.531162109375, 0.5316484985351563, 0.5313546142578125, 0.531051513671875, 0.5314641723632813, 1.10335595703125, 0.5312788696289062, 0.5317376098632812, 0.5317816162109374, 0.5310443725585937, 0.5310945434570312, 0.531515380859375, 0.5312010498046875, 0.5312890625, 0.5315604248046875, 0.5310709838867187, 0.5312757568359375, 0.5313177490234375, 0.5313668823242188, 0.5315430297851562, 0.5313382568359375, 0.531462158203125, 0.5310822143554688, 0.5318031616210938, 0.5317324829101563, 0.5316137084960938, 0.5309522094726562, 0.5314713745117188, 0.531346435546875, 0.5311826171875, 0.5308856201171875, 0.5316904907226563, 0.5310525512695312, 0.5313648681640625, 0.5309481201171875, 0.5315020751953125, 0.5309798583984375, 0.5310975952148438, 0.5311068115234375, 0.53123583984375, 0.5308887329101563, 0.5311610717773437, 0.5312593994140625, 0.5315113525390625, 0.5320181274414062, 0.53134130859375, 0.5311498413085938, 0.5313812255859375, 0.5318276977539063, 0.5313341674804688, 0.5315061645507813, 0.5313771362304688, 0.5312665405273438, 0.53157373046875, 0.5311734008789063, 0.5318348999023438, 0.5309931640625, 0.5319618530273438, 0.5310545654296875, 0.531431396484375, 0.5314805908203125, 0.5322670288085938, 0.5313526000976563, 0.5318143920898437, 0.5314058227539062, 0.5316392822265625, 0.5314242553710937, 0.5316956176757812, 1.1039549560546875, 0.5310596923828125, 0.531430419921875, 0.5309603881835937, 0.5313013916015625, 0.5310545654296875, 0.5314334716796875, 0.5311129760742187, 0.5314866943359375, 0.5309327392578125, 0.5313535766601563, 0.5317621459960937, 0.5320017700195312, 0.5313474731445312, 0.5321390380859375, 0.5313034057617188, 0.5315491943359375, 0.5315614624023437, 0.5321195678710937, 0.5317386474609375, 0.531430419921875, 0.53125732421875, 0.5315266723632812, 0.5314437255859376, 0.5318656005859375, 0.5314652099609375, 0.5318246459960938, 0.5309419555664062, 0.5313320922851562, 0.5310873413085937, 0.5319567260742187, 0.5315020751953125, 0.5317171020507813, 0.5314866943359375, 0.531135498046875, 0.5316986694335938, 0.5315419921875, 0.531198974609375, 0.5315286865234375, 0.5312481079101562, 0.53157373046875, 0.5313710327148438, 0.531556396484375, 0.5311170043945312, 0.5320908813476563, 0.5312860107421875, 0.5311273193359375, 0.53108837890625, 0.5317069091796875, 0.5314140014648437, 0.5318666381835937, 0.5313013916015625, 0.5315419921875, 0.5312225341796875, 0.531567626953125, 0.5311692504882812, 0.53146728515625, 0.5312901000976562, 0.5316116333007812, 0.5311057739257813, 0.5314898071289063, 0.5310587158203125, 0.5314539794921875, 1.1042969970703125, 0.5312860107421875, 0.5315072021484375, 0.5316034545898437, 0.5322137451171876, 0.5317847290039063, 0.5316690063476562, 0.5311959228515625, 0.531631103515625, 0.5317171020507813, 0.531820556640625, 0.5314273071289063, 0.5316249389648438, 0.5310904541015625, 0.531863525390625, 0.5312501831054688, 0.53151025390625, 0.531252197265625, 0.5320970458984375, 0.5310699462890625, 0.5312184448242188, 0.5309183959960937, 0.5319987182617187, 0.531577880859375, 0.5319649047851562, 0.5313546142578125, 0.53127783203125, 0.5312000122070313, 0.5314058227539062, 0.5309552612304688, 0.5314765014648437, 0.5311477661132813, 0.5315399780273438, 0.5311395874023438, 0.5313935546875, 0.5317590942382813, 0.531430419921875, 0.5314949340820313, 0.53096240234375, 0.5309450073242188, 0.5312819213867187, 0.5316137084960938, 0.5311641845703124, 0.5311508178710938, 0.5317099609375, 0.531009521484375, 0.5310576782226563, 0.530977783203125, 0.5313966064453125, 0.531272705078125, 0.531178466796875, 0.53085693359375, 0.5311764526367188, 0.5310115966796874, 0.531294189453125, 0.5309757690429687, 0.5316322021484375, 0.5310668334960937, 0.531282958984375, 0.531072021484375, 0.5319813232421875, 0.5315072021484375, 0.5314345092773437, 1.1038515625, 0.5315747680664062, 0.5315563354492188, 0.5314283447265625, 0.5311314086914063, 0.5308887329101563, 0.5313218383789062, 0.5309593505859375, 0.5309398803710937, 0.5308262329101563, 0.5310136108398438, 0.5309296875, 0.5314590454101562, 0.5310699462890625, 0.5316587524414063, 0.5311928100585938, 0.5312696533203125, 0.5310955810546875, 0.5310802001953125, 0.5312481079101562, 0.5316751098632813, 0.5312921752929688, 0.5321625366210937, 0.531294189453125, 0.5315594482421875, 0.5310873413085937, 0.5322465209960937, 0.5310218505859375, 0.5312512817382813, 0.5308609619140625, 0.531178466796875, 0.5309429931640625, 0.5310105590820312, 0.5310392456054688, 0.5313372192382813, 0.5314283447265625, 0.5313710327148438, 0.5309235229492187, 0.5312706298828125, 0.5308815307617187, 0.5310955810546875, 0.5310003051757812, 0.5311764526367188, 0.5315164184570312, 0.53134130859375, 0.5313402709960937, 0.5311385498046876, 0.5311682739257813, 0.53187890625, 0.5310679321289062, 0.5314652099609375, 0.5309286499023438, 0.5317437744140625, 0.5314334716796875, 0.5315911865234375, 0.5309706420898438, 0.5311815795898438, 0.5312235717773437, 0.5312030639648437, 0.5309204711914063, 0.5312102661132813, 0.5308764038085938, 0.5311703491210937]",tokens/s,1.853538952869569,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neox-20b,EleutherAI/gpt-neox-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -556,34 +859,30 @@ ChildProcessError: Traceback (most recent call last): self.load_model_from_pretrained() File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained self.pretrained_model = self.automodel_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 558, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained - hf_quantizer.postprocess_model(model) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model - return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading - model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 432, in post_init_awq_exllama_modules - model = exllama_post_init(model) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init - submodule.post_init() - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init - self.q4 = exl_ext.make_q4( -NameError: name 'exl_ext' is not defined + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 -4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-j-6b,EleutherAI/gpt-j-6b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-4B,Qwen/Qwen1.5-4B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,3149.217792,5128.060928,0.0,4481.613824,4276.256768,s,10,3.149311706542969,0.3149311706542969,0.0015717060185626956,0.31493319702148437,0.3165461700439453,0.3169147476196289,0.31720960968017575,"[0.3172833251953125, 0.3164642639160156, 0.3132740478515625, 0.3144434814453125, 0.31399490356445314, 0.3119559020996094, 0.3140972900390625, 0.31542291259765626, 0.3162637023925781, 0.31611187744140623]",tokens/s,812.8760308741042,kWh,3.694201881686847e-06,2.024221956560268e-06,1.6665230346062446e-05,2.238365418430956e-05,tokens/kWh,11436917.220578322,MB,3149.217792,5128.060928,0.0,4481.613824,4465.661952,s,10,184.204947265625,18.420494726562502,0.008042192397515377,18.4165078125,18.4298185546875,18.432060644531248,18.433854316406247,"[18.41758203125, 18.4293203125, 18.415013671875, 18.434302734375, 18.41543359375, 18.427755859375, 18.427966796875, 18.41258203125, 18.409763671875, 18.4152265625]",tokens/s,3.420103582188457,kWh,0.0002180770655886995,0.00011952181134027568,0.0009648250270677349,0.00130242390399671,tokens/kWh,48371.3480739057,,s,629,186.7401061401367,0.2968841115105512,0.03741416455160097,0.29225473022460935,0.29295676879882815,0.29340855712890623,0.6063264038085938,"[0.29236837768554685, 0.29334426879882813, 0.29221682739257815, 0.29205914306640623, 0.2920222778320313, 0.2922946472167969, 0.2924492797851562, 0.2926929931640625, 0.29193624877929686, 0.2920509338378906, 0.29204376220703127, 0.2921123962402344, 0.291873779296875, 0.2915205078125, 0.29172222900390626, 0.29199871826171875, 0.2923089904785156, 0.2930401306152344, 0.2921902160644531, 0.29266842651367186, 0.29187994384765625, 0.2924564514160156, 0.2924277648925781, 0.2920263671875, 0.29210009765625, 0.2922608642578125, 0.2919096374511719, 0.29193624877929686, 0.29192294311523437, 0.29198541259765626, 0.29212261962890623, 0.29261724853515625, 0.2921973571777344, 0.29344256591796875, 0.2920785827636719, 0.2922250366210937, 0.29242059326171876, 0.2925332336425781, 0.2925537414550781, 0.29240524291992187, 0.2921902160644531, 0.2919741516113281, 0.29227825927734374, 0.2922506103515625, 0.2918768615722656, 0.29220248413085936, 0.29238067626953124, 0.29187481689453126, 0.29204788208007815, 0.2923376770019531, 0.2920396728515625, 0.29325619506835937, 0.2925700988769531, 0.29266226196289064, 0.29220452880859377, 0.2922823791503906, 0.2919710693359375, 0.29259878540039064, 0.2924472351074219, 0.29214208984375, 0.2923765869140625, 0.2920519714355469, 0.6092022705078125, 0.29223724365234377, 0.2937958374023438, 0.29247589111328126, 0.29319781494140623, 0.2925875244140625, 0.2924288024902344, 0.29230181884765627, 0.29243902587890624, 0.2923919372558594, 0.2927564697265625, 0.29259161376953124, 0.29243084716796874, 0.29296743774414064, 0.2923427734375, 0.29186456298828123, 0.2922250366210937, 0.29213082885742186, 0.29205810546875, 0.2923417663574219, 0.2929213562011719, 0.2921850891113281, 0.2936012878417969, 0.29257931518554686, 0.29359002685546876, 0.29242059326171876, 0.2921574401855469, 0.2920642700195312, 0.2926080017089844, 0.29214617919921876, 0.2923816833496094, 0.29216973876953123, 0.29204684448242185, 0.2921922607421875, 0.2923284606933594, 0.2922137451171875, 0.2923929748535156, 0.292546630859375, 0.2926417236328125, 0.2925137939453125, 0.2924062805175781, 0.2922659912109375, 0.29193319702148435, 0.29348370361328124, 0.29205792236328126, 0.2921850891113281, 0.2963548278808594, 0.292305908203125, 0.2925137939453125, 0.2923642883300781, 0.2921707458496094, 0.2921430969238281, 0.2926448669433594, 0.2927411193847656, 0.29220761108398435, 0.292232177734375, 0.29215130615234375, 0.2934067077636719, 0.29257626342773435, 0.29231924438476564, 0.2922465209960938, 0.2922557373046875, 0.2923213195800781, 0.6058434448242187, 0.2923243408203125, 0.29236737060546875, 0.2920919189453125, 0.2922966918945312, 0.2922536926269531, 0.2924298095703125, 0.29252813720703125, 0.29234893798828127, 0.2925455322265625, 0.2922403869628906, 0.29219635009765627, 0.2920130615234375, 0.29205810546875, 0.2921932678222656, 0.2922680358886719, 0.2919772033691406, 0.29201919555664063, 0.29235302734375, 0.2921318359375, 0.29199154663085936, 0.2918656005859375, 0.29201611328125, 0.294361083984375, 0.2921062316894531, 0.29225473022460935, 0.2920929260253906, 0.2933616638183594, 0.29293466186523437, 0.29222708129882813, 0.292073486328125, 0.29259878540039064, 0.2922178649902344, 0.29183078002929685, 0.29233355712890624, 0.2924349365234375, 0.291989501953125, 0.29248818969726564, 0.29223526000976563, 0.29176217651367187, 0.2918082580566406, 0.2931097717285156, 0.29253836059570315, 0.29301043701171875, 0.2920130615234375, 0.2920407409667969, 0.2919710388183594, 0.29457406616210935, 0.2923519897460938, 0.2923028564453125, 0.2920929260253906, 0.2918901672363281, 0.29224755859375, 0.2919526672363281, 0.291937255859375, 0.29211032104492185, 0.29193011474609376, 0.29155224609375, 0.2918410339355469, 0.2920980529785156, 0.292168701171875, 0.29260595703125, 0.292453369140625, 0.6075504760742187, 0.2925127563476563, 0.2923765869140625, 0.29210931396484374, 0.29227825927734374, 0.2921656188964844, 0.29187890625, 0.29214105224609377, 0.29254144287109374, 0.29202023315429687, 0.29206732177734374, 0.293718017578125, 0.2936227722167969, 0.29316915893554685, 0.29292236328125, 0.29260595703125, 0.292431884765625, 0.2927585144042969, 0.2929725341796875, 0.29310772705078125, 0.29271142578125, 0.2924615783691406, 0.29305242919921876, 0.29276776123046877, 0.2925915832519531, 0.29245440673828127, 0.29250765991210936, 0.2925107116699219, 0.29247589111328126, 0.29257830810546875, 0.29249331665039063, 0.2926243896484375, 0.2935541687011719, 0.2926243896484375, 0.29202532958984373, 0.2919198608398437, 0.292021240234375, 0.29210418701171875, 0.2919884948730469, 0.2921758728027344, 0.2921983947753906, 0.292052978515625, 0.2918523254394531, 0.2925229797363281, 0.2951219177246094, 0.2927698059082031, 0.2923653259277344, 0.29271551513671873, 0.2925813903808594, 0.29277389526367187, 0.2923816833496094, 0.2925250549316406, 0.2931272277832031, 0.2930205993652344, 0.29253631591796875, 0.2923786315917969, 0.2930882568359375, 0.2928087158203125, 0.29209906005859376, 0.2928845520019531, 0.29307998657226564, 0.2928609313964844, 0.2925823974609375, 0.6065142211914063, 0.2926080322265625, 0.29234066772460937, 0.29249432373046874, 0.2920980529785156, 0.2918707275390625, 0.2923858032226562, 0.2918133850097656, 0.2919024963378906, 0.2923048706054687, 0.29231924438476564, 0.29207757568359377, 0.29218612670898436, 0.29211032104492185, 0.29206936645507814, 0.29310873413085936, 0.2925721740722656, 0.2921850891113281, 0.292738037109375, 0.2921891784667969, 0.2920550537109375, 0.2918143920898438, 0.2921983947753906, 0.29208370971679687, 0.29210214233398435, 0.29210214233398435, 0.2925189208984375, 0.2919208984375, 0.29203353881835936, 0.2927615966796875, 0.291736572265625, 0.29209701538085936, 0.2920550537109375, 0.2918656005859375, 0.2918370056152344, 0.2924051818847656, 0.2920263671875, 0.2931199951171875, 0.29166488647460936, 0.29208370971679687, 0.2922158203125, 0.2921277465820312, 0.2917294006347656, 0.29210012817382813, 0.29211541748046876, 0.29188607788085935, 0.29205810546875, 0.29206732177734374, 0.29190142822265625, 0.2926776428222656, 0.29242471313476565, 0.2921656188964844, 0.29246054077148437, 0.2923315124511719, 0.29213287353515627, 0.2920325012207031, 0.2924195861816406, 0.29780581665039063, 0.29234072875976563, 0.2921574401855469, 0.29292031860351564, 0.2927329406738281, 0.2922823791503906, 0.6048440551757812, 0.29242266845703124, 0.29194342041015625, 0.2922823791503906, 0.29246875, 0.2921922607421875, 0.2924472351074219, 0.2927779846191406, 0.29216461181640624, 0.29211953735351565, 0.29237759399414065, 0.29223526000976563, 0.2919372863769531, 0.2918973388671875, 0.2923294982910156, 0.2918020935058594, 0.29216152954101565, 0.2923212890625, 0.29266021728515623, 0.2927503356933594, 0.2927216491699219, 0.29210418701171875, 0.2921973876953125, 0.29242471313476565, 0.29217279052734374, 0.29210931396484374, 0.2921912231445313, 0.292021240234375, 0.29214617919921876, 0.29203866577148435, 0.29192190551757813, 0.2918973388671875, 0.29224346923828126, 0.291884033203125, 0.291857421875, 0.29202532958984373, 0.2921379699707031, 0.2925025329589844, 0.2959308776855469, 0.2924974060058594, 0.2935582580566406, 0.2933534851074219, 0.2929541015625, 0.29257931518554686, 0.2927984619140625, 0.29274725341796876, 0.29252197265625, 0.29275750732421874, 0.2934097900390625, 0.29243289184570315, 0.2923284606933594, 0.2925189208984375, 0.2929377136230469, 0.29240524291992187, 0.2938306579589844, 0.29293670654296877, 0.29282302856445314, 0.29310772705078125, 0.2923100280761719, 0.2923991088867188, 0.29276568603515624, 0.2923294677734375, 0.2927698059082031, 0.6069278564453126, 0.2922486267089844, 0.29226181030273435, 0.29200384521484374, 0.29218099975585937, 0.2922465209960938, 0.29204376220703127, 0.2927329406738281, 0.2925977478027344, 0.292210693359375, 0.29327462768554685, 0.2930093994140625, 0.29313638305664064, 0.29303192138671874, 0.292389892578125, 0.29288653564453127, 0.29291519165039065, 0.29289266967773436, 0.2928721923828125, 0.292701171875, 0.29249331665039063, 0.29188607788085935, 0.29253631591796875, 0.29483929443359375, 0.2933084106445312, 0.29292849731445314, 0.2923530883789062, 0.29386029052734375, 0.29232742309570314, 0.2925465698242187, 0.29262130737304687, 0.292274169921875, 0.2924564514160156, 0.292242431640625, 0.2929407958984375, 0.2920704040527344, 0.2924810791015625, 0.292454345703125, 0.2922188720703125, 0.29238885498046874, 0.29204888916015626, 0.2919598083496094, 0.2920232849121094, 0.2920048522949219, 0.29257220458984373, 0.29197512817382815, 0.2930882568359375, 0.29200384521484374, 0.292384765625, 0.2927626342773437, 0.2920519714355469, 0.2919045104980469, 0.2923561096191406, 0.29254348754882814, 0.2925537414550781, 0.2922148132324219, 0.292211669921875, 0.2919024658203125, 0.29274008178710936, 0.29211032104492185, 0.29219430541992186, 0.2925619201660156, 0.2921574401855469, 0.6078638305664062, 0.2920642700195312, 0.29204171752929686, 0.29206219482421875, 0.2920550537109375, 0.2921185302734375, 0.29250149536132813, 0.29204888916015626, 0.29186456298828123, 0.2920867919921875, 0.29199871826171875, 0.29200689697265625, 0.29195059204101564, 0.2919956359863281, 0.2926612548828125, 0.291999755859375, 0.29500723266601564, 0.2922413940429687, 0.2919342041015625, 0.2922650146484375, 0.2921286926269531, 0.29177548217773436, 0.2920263671875, 0.29223629760742187, 0.2919342041015625, 0.2920151062011719, 0.29193624877929686, 0.29171505737304687, 0.29255679321289063, 0.29226290893554685, 0.29204888916015626, 0.29261932373046873, 0.2921389465332031, 0.2924472351074219, 0.292569091796875, 0.29272988891601565, 0.2924666442871094, 0.29255474853515623, 0.2924288024902344, 0.29235711669921877, 0.29266842651367186, 0.2926305236816406, 0.292279296875, 0.29281585693359374, 0.29269195556640626, 0.2922127380371094, 0.2923765869140625, 0.29226190185546874, 0.29332275390625, 0.29225677490234375, 0.2925455322265625, 0.29204684448242185, 0.2919465026855469, 0.29239501953125, 0.292094970703125, 0.2916546630859375, 0.29185330200195314, 0.2919178161621094, 0.29201202392578124, 0.2919045104980469, 0.2917683715820312, 0.29191879272460936, 0.29268173217773436, 0.609132568359375, 0.2924984436035156, 0.2927001647949219, 0.29203353881835936, 0.2919598083496094, 0.29245440673828127, 0.29231411743164065, 0.2920304565429688, 0.29216973876953123, 0.2918604736328125, 0.2920816650390625, 0.29198541259765626, 0.29182769775390627, 0.29221478271484375, 0.29226290893554685, 0.29183078002929685, 0.2922137451171875, 0.2939207763671875, 0.29266226196289064, 0.292632568359375, 0.2929541015625, 0.29197210693359377, 0.2920867919921875, 0.29224346923828126, 0.2920704040527344, 0.292126708984375, 0.2921769104003906, 0.2920243225097656, 0.292173828125, 0.2919126892089844, 0.2919823303222656, 0.2920867919921875, 0.2925608825683594, 0.29213287353515627, 0.29228546142578127, 0.29201608276367186, 0.2929305725097656, 0.2924963989257813, 0.2920181884765625, 0.2923519897460938, 0.29213082885742186, 0.2921359252929687, 0.29183795166015625, 0.29199258422851565, 0.2920724487304687, 0.29192807006835936, 0.292031494140625, 0.2920294494628906, 0.292173828125, 0.2920765380859375, 0.29200689697265625, 0.2918758544921875, 0.2921349182128906, 0.29268896484375, 0.2934025573730469, 0.292505615234375, 0.2919270324707031, 0.29225164794921876, 0.292105224609375, 0.29204071044921875, 0.2915840148925781, 0.2917908630371094, 0.2920570983886719, 0.6085621948242188, 0.2921451416015625, 0.29239706420898437, 0.29239501953125, 0.29228955078125, 0.2920243225097656, 0.2924892272949219, 0.2920345458984375, 0.2921123962402344, 0.291962890625, 0.2920796203613281, 0.29241549682617185, 0.29221682739257815, 0.2918553466796875, 0.291915771484375, 0.29186663818359376, 0.2920427551269531, 0.2920980529785156, 0.29218817138671876, 0.29189529418945315, 0.29189837646484373, 0.29198028564453127, 0.29211953735351565, 0.29269195556640626, 0.2930841674804687, 0.2925148010253906, 0.29509222412109376, 0.29249127197265623, 0.2924369812011719, 0.29232537841796874, 0.2920162048339844, 0.29203448486328126, 0.29213900756835937, 0.2920243225097656, 0.29217279052734374, 0.2922342529296875, 0.2917130126953125, 0.2923991088867188, 0.29214718627929687, 0.2926458740234375, 0.2934947814941406, 0.2932357177734375, 0.2925066223144531, 0.2929162292480469, 0.29218817138671876, 0.2923735046386719, 0.2922332153320312, 0.2920325012207031, 0.2919761962890625, 0.29228543090820314, 0.29216461181640624, 0.29214718627929687, 0.29235302734375, 0.292274169921875, 0.2920376281738281, 0.29199154663085936, 0.2923724670410156, 0.2924646301269531, 0.29217279052734374, 0.29284454345703126, 0.29187277221679686, 0.2919321594238281, 0.29226495361328125]",tokens/s,3.368317674233167,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-1.8B,Qwen/Qwen1.5-1.8B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,1869.656064,3326.60736,0.0,2682.257408,2578.238464,s,10,1.3870304412841796,0.13870304412841797,0.00244373759901306,0.13811017608642578,0.14131588592529296,0.14323313064575197,0.14476692642211916,"[0.14515037536621095, 0.13752809143066405, 0.1378865966796875, 0.13688485717773438, 0.13833375549316407, 0.13642874145507813, 0.13711347961425782, 0.13835682678222655, 0.1384578857421875, 0.14088983154296875]",tokens/s,1845.6696578554026,kWh,1.6179650182774805e-06,8.864855210497944e-07,6.8487079142743225e-06,9.353158453601597e-06,tokens/kWh,27370433.342912387,MB,1869.656064,3328.704512,0.0,2682.257408,2667.0976,s,10,82.21512695312498,8.2215126953125,0.02726168309176718,8.22408544921875,8.254994140625,8.26155078125,8.26679609375,"[8.218158203125, 8.2300126953125, 8.231037109375, 8.253537109375, 8.2426494140625, 8.18561865234375, 8.1915498046875, 8.18607568359375, 8.268107421875, 8.208380859375]",tokens/s,7.662823416415751,kWh,9.660354430590225e-05,5.294595236465346e-05,0.0003949548913061273,0.000544504387976683,tokens/kWh,115701.54693169857,,s,629,83.32182740783702,0.1324671341936994,0.01648122008836376,0.13012069702148438,0.13181768188476561,0.13225840454101562,0.2677710559082031,"[0.1353492431640625, 0.13449932861328126, 0.13354396057128906, 0.13218812561035156, 0.1321062469482422, 0.132279296875, 0.1320765380859375, 0.1297838134765625, 0.12942950439453124, 0.13008486938476563, 0.1297357482910156, 0.12950111389160157, 0.12954214477539064, 0.12996812438964844, 0.12972236633300782, 0.129586181640625, 0.12967730712890624, 0.12944589233398437, 0.12974386596679688, 0.12966400146484375, 0.12964044189453125, 0.12967730712890624, 0.13006745910644532, 0.13016166687011718, 0.12972647094726564, 0.13020672607421874, 0.1301012420654297, 0.13006541442871095, 0.1300264892578125, 0.13026918029785156, 0.1298155517578125, 0.12965580749511718, 0.12980120849609375, 0.13100137329101563, 0.12973667907714845, 0.12997222900390626, 0.12998655700683595, 0.13009100341796875, 0.13086618041992187, 0.13041253662109376, 0.13003570556640626, 0.13003366088867188, 0.12998655700683595, 0.13012069702148438, 0.12974899291992187, 0.1296711730957031, 0.13009510803222657, 0.13012582397460937, 0.13009408569335937, 0.129723388671875, 0.12967532348632813, 0.13028140258789062, 0.1303726043701172, 0.13008793640136718, 0.13204173278808592, 0.13000090026855468, 0.1302650909423828, 0.12970188903808594, 0.13084979248046874, 0.12998963928222657, 0.12973773193359375, 0.13070130920410156, 0.26767672729492187, 0.13097874450683594, 0.12999679565429687, 0.1298524169921875, 0.12990463256835938, 0.12967730712890624, 0.130735107421875, 0.12993740844726562, 0.12993843078613282, 0.12966195678710937, 0.13075456237792968, 0.1301739501953125, 0.1298462677001953, 0.12979507446289062, 0.12965171813964843, 0.12960255432128906, 0.12983296203613282, 0.1297029113769531, 0.12987289428710938, 0.13024870300292968, 0.12961074829101563, 0.12961587524414062, 0.1304627227783203, 0.13114883422851562, 0.13138124084472655, 0.13123989868164063, 0.1312419891357422, 0.1315799102783203, 0.13145703125, 0.1315246124267578, 0.13152359008789063, 0.13154917907714844, 0.13147648620605468, 0.13150822448730468, 0.1314273223876953, 0.1317058563232422, 0.1314283447265625, 0.13160345458984374, 0.13148159790039063, 0.13152767944335939, 0.13164851379394532, 0.13085594177246093, 0.13025791931152345, 0.12963226318359375, 0.13061222839355469, 0.12971110534667968, 0.1318850860595703, 0.13038383483886717, 0.12956982421875, 0.13069512939453126, 0.12964761352539061, 0.1297664337158203, 0.1301851806640625, 0.13052313232421875, 0.1300756530761719, 0.131198974609375, 0.1302650909423828, 0.13156045532226562, 0.13132595825195312, 0.13137612915039062, 0.13114163208007812, 0.13150822448730468, 0.13148774719238282, 0.26988442993164063, 0.1297592315673828, 0.1303961639404297, 0.13102079772949218, 0.13180621337890625, 0.1311805419921875, 0.131198974609375, 0.1320120391845703, 0.13153382873535155, 0.13055078125, 0.13116621398925782, 0.13224858093261718, 0.1315379180908203, 0.13149183654785157, 0.13085594177246093, 0.1307740173339844, 0.13012069702148438, 0.13034701538085938, 0.13029273986816406, 0.12967219543457031, 0.12965785217285156, 0.13052621459960936, 0.1309951934814453, 0.13132908630371093, 0.12980528259277344, 0.12976431274414063, 0.130735107421875, 0.1307904052734375, 0.13091941833496093, 0.13070028686523438, 0.13097164916992188, 0.13049139404296875, 0.13055282592773437, 0.13094093322753905, 0.1304883270263672, 0.13196185302734376, 0.13015142822265624, 0.13110272216796875, 0.1304780731201172, 0.13022618103027345, 0.13066342163085937, 0.1300070343017578, 0.1298851776123047, 0.12999679565429687, 0.13136691284179688, 0.13021286010742186, 0.1300305938720703, 0.13127577209472657, 0.12978688049316406, 0.12988313293457032, 0.12990975952148437, 0.13098086547851562, 0.12965785217285156, 0.13045555114746094, 0.1315246124267578, 0.1313638458251953, 0.12958309936523438, 0.12952677917480468, 0.12959642028808593, 0.13055795288085936, 0.13012889099121094, 0.13161677551269532, 0.13049856567382812, 0.2679295959472656, 0.13007772827148437, 0.1300623016357422, 0.1296732177734375, 0.1302405090332031, 0.13101568603515626, 0.13053543090820313, 0.12958309936523438, 0.12989543151855468, 0.13040640258789063, 0.12972032165527345, 0.12959744262695314, 0.12965682983398438, 0.12967628479003906, 0.1296046142578125, 0.12968960571289062, 0.1309102020263672, 0.13147955322265625, 0.13136282348632813, 0.13167820739746094, 0.1325506591796875, 0.1314693145751953, 0.1313454132080078, 0.1313280029296875, 0.13255577087402343, 0.1314273223876953, 0.13156965637207033, 0.13132698059082032, 0.1305374755859375, 0.1306746826171875, 0.1309071350097656, 0.13065113830566405, 0.13108326721191407, 0.1307310333251953, 0.13129315185546875, 0.13122866821289061, 0.13088768005371093, 0.13070541381835937, 0.1310146484375, 0.13133619689941406, 0.1298104248046875, 0.13102490234375, 0.13102694702148437, 0.13097779846191407, 0.13345689392089843, 0.13114060974121095, 0.13085081481933594, 0.13208985900878906, 0.13153689575195313, 0.13162701416015626, 0.13186355590820312, 0.13157069396972657, 0.13132492065429688, 0.13146829223632814, 0.13144781494140625, 0.13143244934082032, 0.13133209228515624, 0.13199154663085938, 0.13162086486816407, 0.13136589050292968, 0.13134745788574217, 0.13149900817871094, 0.13125018310546874, 0.27049984741210936, 0.13173965454101563, 0.13157376098632811, 0.13129933166503907, 0.13149183654785157, 0.13152255249023437, 0.13144268798828124, 0.13151744079589844, 0.13144781494140625, 0.1312665557861328, 0.1313638458251953, 0.1312542724609375, 0.13116621398925782, 0.1316812744140625, 0.1304279022216797, 0.13107609558105468, 0.13173452758789062, 0.13157273864746094, 0.13156761169433595, 0.131114013671875, 0.13155325317382813, 0.13144883728027343, 0.13135667419433594, 0.1312972869873047, 0.13135154724121093, 0.13139456176757813, 0.13143653869628907, 0.13130137634277345, 0.13196800231933595, 0.1318707275390625, 0.131631103515625, 0.1325455322265625, 0.13174578857421876, 0.13033779907226561, 0.12979814147949217, 0.129902587890625, 0.13098188781738282, 0.13004287719726562, 0.13019442749023438, 0.12989439392089844, 0.13051187133789063, 0.12981350708007813, 0.12995890808105467, 0.13021900939941405, 0.12963839721679687, 0.12967628479003906, 0.13002546691894531, 0.13009510803222657, 0.1296711730957031, 0.13236019897460938, 0.12973362731933594, 0.13032858276367187, 0.13070643615722657, 0.1302425537109375, 0.1299220428466797, 0.12974490356445312, 0.12966812133789063, 0.12980630493164064, 0.129623046875, 0.13014938354492187, 0.13049958801269532, 0.12993023681640625, 0.1298841552734375, 0.26678680419921874, 0.1296680908203125, 0.12964556884765624, 0.12968447875976563, 0.12973260498046876, 0.129765380859375, 0.1300142059326172, 0.1296855010986328, 0.12954725646972656, 0.1295247344970703, 0.12960870361328125, 0.1295667266845703, 0.12963941955566408, 0.12973158264160156, 0.13008895874023438, 0.13011558532714843, 0.12971827697753907, 0.12998963928222657, 0.12966706848144532, 0.12973568725585938, 0.12970803833007813, 0.1298350067138672, 0.1297090606689453, 0.12965274047851563, 0.12967730712890624, 0.13019648742675782, 0.12975820922851564, 0.12972854614257812, 0.12970799255371093, 0.1302794189453125, 0.12972647094726564, 0.12966400146484375, 0.12970086669921874, 0.12959333801269532, 0.1299220428466797, 0.1300633544921875, 0.1296537628173828, 0.12968038940429688, 0.1319403839111328, 0.12973974609375, 0.13016064453125, 0.13014527893066405, 0.1306234893798828, 0.1300377655029297, 0.13088461303710938, 0.13168946838378906, 0.1299251251220703, 0.12965580749511718, 0.12978688049316406, 0.13135565185546874, 0.13036749267578124, 0.12981248474121093, 0.1297510986328125, 0.1297816925048828, 0.12961383056640624, 0.1298667449951172, 0.12978175354003907, 0.12962611389160156, 0.13000294494628906, 0.12986572265625, 0.1298462677001953, 0.12970803833007813, 0.12991897583007814, 0.2678077392578125, 0.13065216064453125, 0.12982272338867187, 0.13018418884277344, 0.13069625854492187, 0.12963424682617186, 0.12971827697753907, 0.12956063842773438, 0.13137094116210937, 0.13036647033691406, 0.12986265563964844, 0.12970803833007813, 0.1298462677001953, 0.12968447875976563, 0.1297029113769531, 0.12973670959472655, 0.12944793701171875, 0.12968345642089843, 0.1307125701904297, 0.12972134399414062, 0.12976742553710938, 0.12971929931640624, 0.12958309936523438, 0.13087026977539062, 0.12964659118652344, 0.1298032684326172, 0.12998757934570312, 0.13012275695800782, 0.12993536376953124, 0.1304698944091797, 0.1299988555908203, 0.12983807373046874, 0.13000192260742188, 0.13018623352050782, 0.1299199981689453, 0.13019648742675782, 0.13000090026855468, 0.13050265502929687, 0.12989645385742188, 0.12952677917480468, 0.12986778259277343, 0.13002546691894531, 0.13098597717285157, 0.129828857421875, 0.12967730712890624, 0.13003981018066407, 0.1295636444091797, 0.12984115600585938, 0.12994764709472656, 0.13031321716308594, 0.12971315002441405, 0.12961485290527344, 0.12971417236328125, 0.12996202087402345, 0.13037257385253906, 0.12986061096191406, 0.12995584106445313, 0.13010226440429687, 0.1315707550048828, 0.1297571258544922, 0.12957183837890626, 0.12970188903808594, 0.13077810668945314, 0.2692198486328125, 0.13007154846191407, 0.12988313293457032, 0.12971212768554688, 0.12965481567382814, 0.12962608337402343, 0.12967219543457031, 0.12959231567382812, 0.12979200744628908, 0.1296680908203125, 0.13002957153320313, 0.12997938537597656, 0.12960153198242189, 0.13012786865234374, 0.13016986083984375, 0.12967219543457031, 0.1310576629638672, 0.12998348999023437, 0.12976332092285156, 0.1300869140625, 0.1296824951171875, 0.13126751708984374, 0.12993536376953124, 0.13007359313964845, 0.12969984436035156, 0.12999679565429687, 0.1298698272705078, 0.1300858917236328, 0.1298913269042969, 0.1299404754638672, 0.13012275695800782, 0.13052313232421875, 0.1302425537109375, 0.13017805480957031, 0.12996812438964844, 0.12995071411132814, 0.13004083251953125, 0.130302978515625, 0.1300623321533203, 0.13056716918945313, 0.1300695037841797, 0.1299261474609375, 0.12973260498046876, 0.13010330200195314, 0.1296855010986328, 0.13009613037109374, 0.12976742553710938, 0.12981350708007813, 0.12980429077148437, 0.12988723754882814, 0.1295564727783203, 0.12995277404785155, 0.12949913024902343, 0.12956877136230469, 0.129870849609375, 0.13016371154785156, 0.12968754577636718, 0.1295667266845703, 0.12944383239746093, 0.1296282196044922, 0.1295543670654297, 0.12969778442382812, 0.12969062805175782, 0.26870681762695314, 0.1297827911376953, 0.1297592315673828, 0.131557373046875, 0.12989439392089844, 0.13009408569335937, 0.1305999298095703, 0.1311446990966797, 0.13035110473632813, 0.13046885681152343, 0.1303521270751953, 0.13139045715332032, 0.1301749725341797, 0.1304627227783203, 0.13010841369628906, 0.1311856689453125, 0.1305753936767578, 0.1308395233154297, 0.12992515563964843, 0.13041970825195312, 0.13009100341796875, 0.1309265594482422, 0.130155517578125, 0.13011354064941405, 0.12959437561035156, 0.1299148864746094, 0.13058047485351562, 0.13014527893066405, 0.13060198974609374, 0.12966400146484375, 0.12963533020019533, 0.1295380554199219, 0.12953395080566407, 0.1297786865234375, 0.13530213928222656, 0.13242477416992188, 0.13271443176269532, 0.1324779510498047, 0.13233970642089843, 0.13220352172851563, 0.1323294677734375, 0.13220556640625, 0.1319833526611328, 0.1321994171142578, 0.13245542907714844, 0.132347900390625, 0.13226495361328125, 0.13233561706542968, 0.13197415161132814, 0.13208883666992188, 0.13216461181640626, 0.1321881561279297, 0.13220249938964843, 0.13227622985839843, 0.1321246795654297, 0.13223423767089842, 0.13205606079101562, 0.13206629943847656, 0.13218611145019532, 0.13227622985839843, 0.13237759399414062, 0.13250457763671875, 0.13218917846679687, 0.27291647338867187, 0.13057228088378905, 0.12985139465332032, 0.12977766418457032, 0.12981146240234376, 0.1296363525390625, 0.12980120849609375, 0.12959642028808593, 0.12995993041992188, 0.129691650390625, 0.12974592590332032, 0.12970086669921874, 0.12972647094726564, 0.1295984649658203, 0.12972032165527345, 0.13100953674316407, 0.13002957153320313, 0.13050367736816407, 0.13074021911621095, 0.13157478332519532, 0.12986880493164063, 0.12991693115234376, 0.12960050964355468, 0.12965682983398438, 0.13195468139648436, 0.13054464721679687, 0.1317058563232422, 0.13119078063964842, 0.13114982604980469, 0.1297049560546875, 0.12969573974609375, 0.12971315002441405, 0.12958412170410155, 0.1308958740234375, 0.1309696044921875, 0.12978994750976564, 0.12954112243652344, 0.12962713623046876, 0.12967832946777344, 0.12949913024902343, 0.12971827697753907, 0.12982989501953124, 0.1297622985839844, 0.12973464965820314, 0.12974490356445312, 0.1297244110107422, 0.13159628295898437, 0.13082418823242187, 0.1297407989501953, 0.12985548400878907, 0.1299814453125, 0.13154815673828124, 0.13089791870117187, 0.12983602905273436, 0.1315010528564453, 0.13150003051757814, 0.132068359375, 0.13167718505859374, 0.13150309753417969, 0.13090509033203124, 0.12976332092285156, 0.1298053741455078, 0.1295963592529297]",tokens/s,7.549042304619914,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-14B,Qwen/Qwen-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -591,34 +890,28 @@ ChildProcessError: Traceback (most recent call last): self.load_model_from_pretrained() File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained self.pretrained_model = self.automodel_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained - return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained - hf_quantizer.postprocess_model(model) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model - return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading - model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 432, in post_init_awq_exllama_modules - model = exllama_post_init(model) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init - submodule.post_init() - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init - self.q4 = exl_ext.make_q4( -NameError: name 'exl_ext' is not defined + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 550, in from_pretrained + model_class = get_class_from_dynamic_module( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 498, in get_class_from_dynamic_module + final_module = get_cached_module_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 323, in get_cached_module_file + modules_needed = check_imports(resolved_module_file) + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 181, in check_imports + raise ImportError( +ImportError: This modeling file requires the following packages that were not found in your environment: transformers_stream_generator. Run `pip install transformers_stream_generator` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 -4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-1.3B,EleutherAI/gpt-neo-1.3B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-7.5B,facebook/xglm-7.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -628,32 +921,26 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained - hf_quantizer.postprocess_model(model) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model - return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading - model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 432, in post_init_awq_exllama_modules - model = exllama_post_init(model) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init - submodule.post_init() - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init - self.q4 = exl_ext.make_q4( -NameError: name 'exl_ext' is not defined + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: XGLMForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 -4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-34B,01-ai/Yi-34B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-12b,EleutherAI/pythia-12b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -663,32 +950,103 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained - hf_quantizer.postprocess_model(model) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model - return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading - model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 432, in post_init_awq_exllama_modules - model = exllama_post_init(model) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init - submodule.post_init() - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init - self.q4 = exl_ext.make_q4( -NameError: name 'exl_ext' is not defined + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 -4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-125m,EleutherAI/gpt-neo-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-7B,Qwen/Qwen1.5-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,2805.69856,8389.132288,0.0,7742.685184,7007.0144,s,10,5.757144287109376,0.5757144287109376,0.0011485729711129637,0.5757064819335938,0.5769704406738282,0.5771491607666016,0.5772921368408204,"[0.5760953979492187, 0.5769307250976563, 0.574836669921875, 0.5745257568359375, 0.574697998046875, 0.5738778076171875, 0.5753175659179688, 0.5767816162109375, 0.577327880859375, 0.5767528686523438]",tokens/s,444.66490196050984,kWh,6.783310471125591e-06,3.7157116531489013e-06,3.30976653669996e-05,4.3596687491274095e-05,tokens/kWh,5872005.758493431,MB,2805.69856,8389.132288,0.0,7742.685184,7283.984384,s,10,336.2841484375,33.62841484375,0.0041773861216769514,33.62973046875,33.63291328125,33.63334921875,33.63369796875,"[33.62066796875, 33.6314140625, 33.6304453125, 33.62453125, 33.627359375, 33.629015625, 33.63378515625, 33.63111328125, 33.623, 33.63281640625]",tokens/s,1.873415690056198,kWh,0.0003969375698617947,0.00021755745967517217,0.0019112099178555991,0.0025257049473925665,tokens/kWh,24943.531137727943,,s,629,340.93043640136693,0.5420197717032864,0.06847421137969394,0.5337415771484375,0.53422822265625,0.5344366455078124,1.1091660595703123,"[0.5336361083984374, 0.5341531982421875, 0.533359619140625, 0.5337293090820312, 0.53304931640625, 0.5336708984375, 0.5331292114257813, 0.5338419189453125, 0.5331640625, 0.5335398559570312, 0.53309033203125, 0.5335951538085938, 0.5331834716796875, 0.5338798217773437, 0.5333298950195312, 0.5336842041015625, 0.5331845092773437, 0.5336678466796875, 0.533411865234375, 0.53406103515625, 0.5336094970703125, 0.5338388671875, 0.5332828369140625, 0.5337896728515625, 0.5332971801757812, 0.533770263671875, 0.5334599609375, 0.5337487182617188, 0.5331834716796875, 0.5339166870117188, 0.5336627197265625, 0.5344092407226563, 0.5335090942382813, 0.5339432983398438, 0.533738525390625, 0.5342853393554687, 0.5333934326171875, 0.5340252075195312, 0.53340771484375, 0.534096923828125, 0.5334374389648437, 0.5337191162109375, 0.5333012084960937, 0.5337835693359375, 0.5332633666992187, 0.53411328125, 0.533443603515625, 0.5338327026367188, 0.5333759765625, 0.5339893798828125, 0.5333831787109375, 0.5338306274414063, 0.5334395141601562, 0.5340466918945312, 0.5333514404296875, 0.5337937622070312, 0.5338665161132813, 0.5344153442382813, 0.5337047119140625, 0.5338112182617187, 0.5332122192382812, 0.5337702026367187, 1.1116871337890626, 0.5333892822265625, 0.53395556640625, 0.5335818481445312, 0.534329345703125, 0.53378662109375, 0.533939208984375, 0.533485595703125, 0.5340098266601563, 0.533265380859375, 0.5337507934570312, 0.5338316650390625, 0.53416552734375, 0.533770263671875, 0.5342669067382813, 0.5334896850585937, 0.5339904174804687, 0.5334537963867187, 0.5338736572265625, 0.5339002685546875, 0.5341122436523438, 0.5333739624023438, 0.5337057495117188, 0.5331763305664062, 0.5336084594726562, 0.5334323120117187, 0.5338009643554688, 0.5331425170898437, 0.5337620239257812, 0.5331527709960937, 0.5337323608398438, 0.5339801635742187, 0.5345730590820312, 0.53325927734375, 0.5336207275390625, 0.53326953125, 0.5340211791992188, 0.5333349609375, 0.5338777465820312, 0.53340673828125, 0.5337753295898438, 0.533338134765625, 0.533880859375, 0.533570556640625, 0.5340579833984375, 0.5336616821289063, 0.5373572998046875, 0.5339514770507813, 0.53390234375, 0.5336719360351563, 0.5342648315429688, 0.5339279174804688, 0.5340426025390625, 0.5336504516601562, 0.5338839721679688, 0.53328173828125, 0.534540283203125, 0.5336729736328125, 0.5342125854492188, 0.5336329956054687, 0.53383984375, 0.5335726318359375, 0.5343897705078124, 1.10948046875, 0.5335797729492188, 0.5341194458007813, 0.5339064331054687, 0.5345751342773437, 0.5335787353515625, 0.533949462890625, 0.5334251708984376, 0.5337794799804687, 0.5332059936523438, 0.5337763671875, 0.5334661254882812, 0.533986328125, 0.5332305908203125, 0.5337221069335938, 0.5334609985351563, 0.53401904296875, 0.5335316772460937, 0.5339371337890625, 0.53347021484375, 0.5337108764648437, 0.5331456298828126, 0.5338245239257813, 0.5349140625, 0.5346262817382812, 0.5334579467773437, 0.5338726196289062, 0.533796875, 0.5340897216796875, 0.5336565551757813, 0.5343047485351563, 0.5332715454101562, 0.534076416015625, 0.533580810546875, 0.5337415771484375, 0.5334824829101562, 0.53382861328125, 0.53319677734375, 0.5336668090820312, 0.5332664184570313, 0.5337743530273438, 0.5335418701171875, 0.5340856323242188, 0.5335429077148437, 0.53378662109375, 0.5334508056640626, 0.5341675415039062, 0.5349437255859375, 0.5338644409179687, 0.5339566040039062, 0.5342269287109375, 0.5335900268554687, 0.5341214599609375, 0.53355419921875, 0.5342976684570313, 0.5334230346679687, 0.5340139770507812, 0.5335992431640625, 0.5341634521484375, 0.5334210815429687, 0.5340436401367188, 0.5337569580078125, 0.5342617797851562, 1.1091947021484374, 0.5332183227539062, 0.5337579345703125, 0.533496826171875, 0.5339699096679688, 0.53334326171875, 0.5339320068359374, 0.5333401489257813, 0.533749755859375, 0.533454833984375, 0.533738525390625, 0.5335623779296875, 0.5348157348632813, 0.5332367553710937, 0.5340784301757813, 0.5337159423828125, 0.5338306274414063, 0.5333944091796875, 0.5338828735351563, 0.533201904296875, 0.5338593139648438, 0.53323681640625, 0.5340261840820313, 0.5336555786132813, 0.5336821899414063, 0.5331660766601563, 0.5336535034179688, 0.5331763305664062, 0.5336637573242188, 0.5330974731445313, 0.5336381225585938, 0.5332664184570313, 0.5338818359375, 0.533327880859375, 0.53427099609375, 0.5336790771484375, 0.5338746948242188, 0.533159912109375, 0.5337006225585937, 0.5332623291015625, 0.5337293090820312, 0.5337589721679687, 0.533981201171875, 0.533370849609375, 0.5360977783203125, 0.5334261474609375, 0.5341010131835937, 0.5335623779296875, 0.5340078125, 0.5334896850585937, 0.5340006103515625, 0.5336053466796875, 0.5344645385742187, 0.53421875, 0.5342853393554687, 0.5334138793945312, 0.5337979125976563, 0.53326953125, 0.533792724609375, 0.5334814453125, 0.5338665161132813, 0.5334456176757812, 0.53382861328125, 1.1090924072265624, 0.5336094970703125, 0.5341419677734375, 0.5334814453125, 0.5340355224609376, 0.53357666015625, 0.53401904296875, 0.5336771240234375, 0.5337230834960938, 0.5334425048828125, 0.5336268920898437, 0.5331640625, 0.5335726318359375, 0.533123046875, 0.5340631103515625, 0.5331763305664062, 0.5340078125, 0.5347676391601562, 0.5339586791992188, 0.5332183227539062, 0.5338541870117187, 0.5333883056640625, 0.53372314453125, 0.5331834716796875, 0.5336309814453125, 0.5334537963867187, 0.534108154296875, 0.5333411865234375, 0.5341143188476563, 0.5333524780273438, 0.5335510864257812, 0.5337293090820312, 0.5341480712890625, 0.5335869140625, 0.5337671508789062, 0.533243896484375, 0.5338193969726562, 0.5337169799804687, 0.5342074584960937, 0.533265380859375, 0.5335838623046875, 0.533375, 0.5341593627929687, 0.5334948120117188, 0.5341030883789063, 0.5336299438476563, 0.5339105224609375, 0.5335091552734375, 0.5336759643554687, 0.5331793823242188, 0.5338460083007812, 0.5332551879882812, 0.5341522216796875, 0.5335869140625, 0.53411328125, 0.5335480346679687, 0.53422802734375, 0.5365042724609375, 0.5340682373046876, 0.533644287109375, 0.5340825805664062, 0.533528564453125, 0.5341306762695313, 1.1086192626953124, 0.5332972412109375, 0.534067138671875, 0.5337210693359375, 0.5342228393554688, 0.53347021484375, 0.53422900390625, 0.5334702758789063, 0.5339790649414062, 0.5337579345703125, 0.5341173706054687, 0.5335818481445312, 0.534245361328125, 0.5336678466796875, 0.5341265869140625, 0.5333944091796875, 0.5336135864257813, 0.5332551879882812, 0.5341911010742187, 0.533359619140625, 0.5341552734375, 0.5331834716796875, 0.53369140625, 0.5332572021484375, 0.53432421875, 0.533712890625, 0.5336708984375, 0.533106689453125, 0.533802001953125, 0.5334364013671875, 0.53359716796875, 0.5331845092773437, 0.53380810546875, 0.5332346801757812, 0.5336627197265625, 0.5332254638671875, 0.5337774047851562, 0.5336309814453125, 0.537302001953125, 0.5335510864257812, 0.5340999755859375, 0.5334763793945313, 0.5337241821289063, 0.5331824340820313, 0.5339197387695312, 0.53344970703125, 0.533875732421875, 0.53338623046875, 0.5339381713867187, 0.5334395141601562, 0.5338777465820312, 0.5334159545898437, 0.534012939453125, 0.5338378295898437, 0.5343323974609375, 0.534024169921875, 0.534361083984375, 0.5335521240234375, 0.5342996826171875, 0.5335305786132812, 0.533734375, 0.5334640502929687, 0.5342750854492188, 1.1099852294921875, 0.533232666015625, 0.5340067749023437, 0.5332183227539062, 0.5340108642578125, 0.5337364501953125, 0.534192138671875, 0.533570556640625, 0.5341460571289063, 0.533876708984375, 0.53406005859375, 0.5334681396484375, 0.5340528564453125, 0.5333984985351562, 0.53374462890625, 0.5331937255859375, 0.5337579345703125, 0.533475341796875, 0.5339535522460938, 0.533396484375, 0.53382861328125, 0.5333401489257813, 0.5337302856445313, 0.535394287109375, 0.5337927856445313, 0.5334507446289063, 0.5339443359375, 0.5337825317382813, 0.5344010009765625, 0.533728271484375, 0.5337845458984375, 0.5336329956054687, 0.5338839111328125, 0.5333759765625, 0.5338009643554688, 0.533201904296875, 0.5340794677734375, 0.5336320190429688, 0.5339146118164062, 0.53344873046875, 0.5339638061523437, 0.5336893310546875, 0.5338931274414063, 0.5337426147460937, 0.5341430053710937, 0.5335347290039063, 0.5341624145507813, 0.5345700073242188, 0.5347133178710938, 0.5338163452148438, 0.53465087890625, 0.5338040161132812, 0.5341941528320312, 0.533674072265625, 0.5342606811523437, 0.5343539428710937, 0.5342125854492188, 0.5335050048828125, 0.5339566040039062, 0.533570556640625, 0.5341276245117188, 0.5338695678710937, 0.5341880493164063, 1.111352294921875, 0.533518310546875, 0.534761474609375, 0.5335090942382813, 0.5338511352539063, 0.5336331176757813, 0.5341572265625, 0.5336279296875, 0.5341102294921874, 0.5339033813476562, 0.53374462890625, 0.5332838134765625, 0.5336801147460938, 0.5332142333984375, 0.5337333984375, 0.5331947631835937, 0.534908935546875, 0.533190673828125, 0.5336279296875, 0.5330892944335938, 0.5340794677734375, 0.5336063842773437, 0.5338890380859375, 0.5336463623046875, 0.5338880004882812, 0.5336104736328126, 0.533712890625, 0.5332838134765625, 0.5341624145507813, 0.5335418701171875, 0.5341460571289063, 0.53351220703125, 0.5338890380859375, 0.533243896484375, 0.5342177124023437, 0.5337088012695312, 0.5343447265625, 0.533712890625, 0.5340774536132813, 0.5339156494140626, 0.5344481201171875, 0.5336882934570313, 0.5341992797851562, 0.5336084594726562, 0.533676025390625, 0.5333197021484375, 0.5338849487304688, 0.533917724609375, 0.53441943359375, 0.5336135864257813, 0.5341010131835937, 0.5335132446289063, 0.5340897216796875, 0.5336616821289063, 0.5339934692382813, 0.5337794799804687, 0.5340200805664063, 0.5335142211914062, 0.5342648315429688, 0.533865478515625, 0.5339801635742187, 0.5338695678710937, 0.534255615234375, 1.112015869140625, 0.5335879516601563, 0.534097900390625, 0.5332183227539062, 0.5336187133789062, 0.5332059936523438, 0.5337405395507813, 0.5333309326171874, 0.5338091430664063, 0.533423095703125, 0.5338306274414063, 0.533191650390625, 0.5339586791992188, 0.5338961791992187, 0.5337569580078125, 0.5333165893554688, 0.534349853515625, 0.5343836059570313, 0.5336043701171875, 0.5334517822265625, 0.5338685913085938, 0.5332326049804688, 0.5336350708007812, 0.5330892944335938, 0.5336309814453125, 0.5331548461914063, 0.5336053466796875, 0.5331824340820313, 0.5337999267578125, 0.5332408447265625, 0.533622802734375, 0.5338716430664062, 0.5345494995117187, 0.53351220703125, 0.5343057861328125, 0.5334180297851563, 0.5338480224609375, 0.53349169921875, 0.5338357543945312, 0.533622802734375, 0.533875732421875, 0.5334579467773437, 0.5337610473632812, 0.5334579467773437, 0.5340170288085937, 0.5332745971679688, 0.534091796875, 0.5333995361328125, 0.5338532104492187, 0.5333872680664062, 0.5340620727539063, 0.533886962890625, 0.5340108642578125, 0.5335654296875, 0.533928955078125, 0.5333514404296875, 0.533950439453125, 0.5334404907226562, 0.5343201293945312, 0.5335582885742187, 0.5337221069335938, 0.5335408935546875, 0.5341378784179688, 1.111615478515625, 0.5335418701171875, 0.534171630859375, 0.5334824829101562, 0.533771240234375, 0.5340958862304688, 0.5341583251953125, 0.5335234375, 0.5342811889648438, 0.5338716430664062, 0.53395556640625, 0.533296142578125, 0.5340877075195313, 0.53382861328125, 0.5342125854492188, 0.5335961303710938, 0.5341245727539062, 0.5336575927734375, 0.5344921875, 0.5339248657226563, 0.5340242309570312, 0.533560302734375, 0.5341531982421875, 0.5337302856445313, 0.5339627685546875, 0.5335162963867187, 0.535699462890625, 0.5336187133789062, 0.5340529174804688, 0.5334834594726563, 0.5338634033203125, 0.5333811645507812, 0.5340415649414062, 0.5334630126953125, 0.53372314453125, 0.5333606567382813, 0.5338726196289062, 0.5338736572265625, 0.5340108642578125, 0.5332828369140625, 0.5338992919921876, 0.5332766723632812, 0.5340753784179687, 0.5338142700195313, 0.5339904174804687, 0.5335613403320313, 0.5342269287109375, 0.533306396484375, 0.5340200805664063, 0.5333984985351562, 0.5338890380859375, 0.5335675048828125, 0.5339054565429687, 0.5333565063476563, 0.5340200805664063, 0.5336309814453125, 0.5339617309570313, 0.5336258544921875, 0.5341859741210937, 0.5334886474609375, 0.5339893798828125, 0.53378662109375, 0.534213623046875]",tokens/s,1.8449511479212646,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-34B,,cuda,0,42,,,,,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,,MB,1721.061376,22129.672192,0.0,21483.225088,20799.036928,s,10,27.921223388671876,2.7921223388671876,0.002873024218980136,2.791891357421875,2.7956688476562497,2.7963155029296876,2.7968328271484375,"[2.793120361328125, 2.796962158203125, 2.790662353515625, 2.790180419921875, 2.7901220703125, 2.787002685546875, 2.79004345703125, 2.793321533203125, 2.795525146484375, 2.794283203125]",tokens/s,91.68652692484228,kWh,3.2928969429598913e-05,1.8046407099463975e-05,0.00015823184880759956,0.00020920722533666246,tokens/kWh,1223667.1060860218,MB,1726.52544,22129.672192,0.0,21483.225088,20902.142976,s,10,1662.0463125,166.20463124999998,0.013890741352160448,166.199171875,166.2239140625,166.22626953125,166.22815390625,"[166.211921875, 166.223390625, 166.191390625, 166.191359375, 166.191640625, 166.203515625, 166.194828125, 166.2176875, 166.191953125, 166.228625]",tokens/s,0.3790508093919315,kWh,0.0019620591764814327,0.001075380349219922,0.009375421333664404,0.012412860859365755,tokens/kWh,5075.381148131151,,s,629,1684.5351452636726,2.678116288177539,0.33164961442268687,2.637888427734375,2.63990244140625,2.640667578125,5.42840712890625,"[2.638138427734375, 2.638636962890625, 2.638865478515625, 2.6390283203125, 2.6389248046875, 2.63697509765625, 2.6374482421875, 2.637588623046875, 2.6383994140625, 2.637781982421875, 2.63811181640625, 2.639296630859375, 2.6380380859375, 2.637068359375, 2.638099365234375, 2.637255615234375, 2.637476806640625, 2.63743798828125, 2.637467529296875, 2.63781787109375, 2.637115478515625, 2.637315185546875, 2.63943896484375, 2.639993896484375, 2.6398291015625, 2.63933349609375, 2.63897802734375, 2.639859619140625, 2.63980126953125, 2.639171630859375, 2.637111328125, 2.63720556640625, 2.636434326171875, 2.63687158203125, 2.63707958984375, 2.639266845703125, 2.63845263671875, 2.637464599609375, 2.638060546875, 2.64020263671875, 2.638950439453125, 2.63784033203125, 2.63769091796875, 2.6381884765625, 2.639057861328125, 2.637621337890625, 2.638141357421875, 2.63823876953125, 2.639244384765625, 2.63790283203125, 2.6387138671875, 2.6390537109375, 2.638361572265625, 2.637244384765625, 2.639206298828125, 2.638740478515625, 2.639088623046875, 2.638509033203125, 2.637013916015625, 2.637980712890625, 2.637846435546875, 2.638864501953125, 5.43535302734375, 2.63853662109375, 2.63802978515625, 2.63697607421875, 2.637854736328125, 2.638664794921875, 2.63739794921875, 2.637233154296875, 2.637539306640625, 2.637358154296875, 2.637158447265625, 2.637360107421875, 2.63921875, 2.63846923828125, 2.638044189453125, 2.639015869140625, 2.636918701171875, 2.637613037109375, 2.637201416015625, 2.63785888671875, 2.6367763671875, 2.637592529296875, 2.63693310546875, 2.6386318359375, 2.63771337890625, 2.638298095703125, 2.637001708984375, 2.6378701171875, 2.6421064453125, 2.6414111328125, 2.641197021484375, 2.64100146484375, 2.63889111328125, 2.6406533203125, 2.6406552734375, 2.637020263671875, 2.637322265625, 2.63722900390625, 2.63769091796875, 2.6369853515625, 2.63765185546875, 2.63745751953125, 2.638650390625, 2.638825439453125, 2.637306884765625, 2.64330859375, 2.641152099609375, 2.6409931640625, 2.63775244140625, 2.639099853515625, 2.638685302734375, 2.6388388671875, 2.637993896484375, 2.639129638671875, 2.638023681640625, 2.638602294921875, 2.63790283203125, 2.63815380859375, 2.637550537109375, 2.63849169921875, 2.639201171875, 2.6400400390625, 2.638790771484375, 5.4299453125, 2.63796826171875, 2.638594970703125, 2.637592529296875, 2.6385439453125, 2.637388916015625, 2.638464111328125, 2.637737060546875, 2.63819775390625, 2.63681005859375, 2.637154296875, 2.63805224609375, 2.638739501953125, 2.63828173828125, 2.638024658203125, 2.637736083984375, 2.637816650390625, 2.63781884765625, 2.63653466796875, 2.637263916015625, 2.637474853515625, 2.63747998046875, 2.636569580078125, 2.64140478515625, 2.637125732421875, 2.638499755859375, 2.6369833984375, 2.637737060546875, 2.636338134765625, 2.63744921875, 2.637201416015625, 2.638740478515625, 2.638299072265625, 2.63929150390625, 2.637087646484375, 2.637667236328125, 2.6377021484375, 2.636683349609375, 2.637559814453125, 2.637581298828125, 2.6383564453125, 2.636505126953125, 2.63706005859375, 2.636032958984375, 2.6379765625, 2.637094970703125, 2.63819775390625, 2.64007568359375, 2.6376611328125, 2.6375966796875, 2.6389228515625, 2.63778515625, 2.639754150390625, 2.6381884765625, 2.63853466796875, 2.63933642578125, 2.639509521484375, 2.639793212890625, 2.640154541015625, 2.63959033203125, 2.636541015625, 2.6371533203125, 2.63670068359375, 5.43085791015625, 2.64038818359375, 2.640405517578125, 2.63921875, 2.638604248046875, 2.638392333984375, 2.63703857421875, 2.638612548828125, 2.638138427734375, 2.637631591796875, 2.637518798828125, 2.636412841796875, 2.639814697265625, 2.63709912109375, 2.638194580078125, 2.636851318359375, 2.637906982421875, 2.636904541015625, 2.63757421875, 2.637557861328125, 2.637801513671875, 2.637253662109375, 2.638193603515625, 2.63731396484375, 2.63754345703125, 2.637656005859375, 2.638063720703125, 2.6375537109375, 2.63655517578125, 2.636990478515625, 2.6373427734375, 2.638109619140625, 2.637427734375, 2.637330322265625, 2.63708154296875, 2.63813232421875, 2.636632080078125, 2.638185546875, 2.636789794921875, 2.6382294921875, 2.636559326171875, 2.63756494140625, 2.6364814453125, 2.637990966796875, 2.64270947265625, 2.63874755859375, 2.640064453125, 2.638127197265625, 2.637509521484375, 2.637388916015625, 2.637737060546875, 2.63737353515625, 2.639585205078125, 2.637530029296875, 2.638297119140625, 2.63809423828125, 2.637974609375, 2.63773583984375, 2.637834228515625, 2.63741552734375, 2.637762451171875, 2.637958251953125, 2.63785791015625, 5.428443359375, 2.63807275390625, 2.6370693359375, 2.638107666015625, 2.637665283203125, 2.63819677734375, 2.6379326171875, 2.63821923828125, 2.6370078125, 2.63802685546875, 2.638180419921875, 2.63954443359375, 2.6387373046875, 2.638023681640625, 2.636853271484375, 2.637129638671875, 2.637197265625, 2.636950439453125, 2.63758544921875, 2.638212158203125, 2.63828076171875, 2.6365849609375, 2.6370908203125, 2.638017578125, 2.63769384765625, 2.636833740234375, 2.638193603515625, 2.638341064453125, 2.640734130859375, 2.63990576171875, 2.639550537109375, 2.638340087890625, 2.63885107421875, 2.63923095703125, 2.6383955078125, 2.636854248046875, 2.636768310546875, 2.6369443359375, 2.637570068359375, 2.637836181640625, 2.63730078125, 2.63779541015625, 2.63750244140625, 2.637592529296875, 2.6366474609375, 2.638162841796875, 2.636866455078125, 2.6375146484375, 2.636231689453125, 2.6370693359375, 2.637305908203125, 2.63849169921875, 2.636894287109375, 2.636900390625, 2.637894775390625, 2.637490234375, 2.6377666015625, 2.64785107421875, 2.6376591796875, 2.637158447265625, 2.637832275390625, 2.63720849609375, 2.639467529296875, 5.425724609375, 2.637199462890625, 2.63809130859375, 2.638032958984375, 2.637854736328125, 2.638718017578125, 2.63857861328125, 2.637035400390625, 2.63773583984375, 2.638855224609375, 2.63739892578125, 2.638341064453125, 2.63885205078125, 2.639034423828125, 2.63807177734375, 2.637761474609375, 2.639326171875, 2.639801513671875, 2.639014892578125, 2.639296630859375, 2.637675537109375, 2.63769287109375, 2.637137939453125, 2.637119384765625, 2.6383955078125, 2.63736328125, 2.637094970703125, 2.637177734375, 2.6371953125, 2.637066162109375, 2.637592529296875, 2.639177734375, 2.637533203125, 2.6364169921875, 2.636937255859375, 2.6390283203125, 2.63927294921875, 2.63655126953125, 2.64357373046875, 2.637197265625, 2.637576171875, 2.638668701171875, 2.639214599609375, 2.639289306640625, 2.639532958984375, 2.638011474609375, 2.63690234375, 2.63817626953125, 2.636898193359375, 2.637287353515625, 2.637656982421875, 2.637981689453125, 2.636938232421875, 2.63751171875, 2.6373232421875, 2.63774609375, 2.637768798828125, 2.63883056640625, 2.639134765625, 2.641203125, 2.639278076171875, 2.638201904296875, 2.637488037109375, 5.42831396484375, 2.637304931640625, 2.638138427734375, 2.636865478515625, 2.639958984375, 2.63714404296875, 2.638241943359375, 2.63680810546875, 2.638630859375, 2.638310302734375, 2.640628662109375, 2.639406982421875, 2.63910302734375, 2.6390712890625, 2.640678955078125, 2.640310302734375, 2.63718701171875, 2.637646728515625, 2.637350830078125, 2.638884765625, 2.6375947265625, 2.637400146484375, 2.6368818359375, 2.642212890625, 2.637052978515625, 2.637252685546875, 2.636802001953125, 2.64029296875, 2.6377216796875, 2.6375556640625, 2.6366923828125, 2.63757421875, 2.638635009765625, 2.637696044921875, 2.637308837890625, 2.637477783203125, 2.637402099609375, 2.6365869140625, 2.637488037109375, 2.638253173828125, 2.6386328125, 2.637402099609375, 2.63834716796875, 2.6377451171875, 2.637392822265625, 2.636634033203125, 2.637593505859375, 2.637182861328125, 2.637476806640625, 2.6378896484375, 2.637464599609375, 2.639731689453125, 2.637717529296875, 2.63756591796875, 2.638674072265625, 2.63798974609375, 2.638200927734375, 2.63733251953125, 2.63752197265625, 2.63793359375, 2.636739501953125, 2.63832470703125, 2.637581298828125, 5.43222705078125, 2.638201904296875, 2.63801025390625, 2.639353759765625, 2.63851513671875, 2.637308837890625, 2.637641845703125, 2.6370263671875, 2.636823486328125, 2.638310302734375, 2.637739990234375, 2.636812255859375, 2.637005859375, 2.63739794921875, 2.637341796875, 2.636630126953125, 2.64057861328125, 2.638487548828125, 2.6377880859375, 2.63784033203125, 2.637433837890625, 2.638781494140625, 2.638268310546875, 2.63773486328125, 2.63734375, 2.64085302734375, 2.637772705078125, 2.63678369140625, 2.636781494140625, 2.637518798828125, 2.637106201171875, 2.636708984375, 2.63695166015625, 2.63906103515625, 2.6390087890625, 2.639414306640625, 2.63884912109375, 2.638530517578125, 2.639602783203125, 2.639301513671875, 2.638671875, 2.6394306640625, 2.640407470703125, 2.63750244140625, 2.640359375, 2.640257080078125, 2.641056884765625, 2.639901611328125, 2.64020263671875, 2.640825439453125, 2.637892578125, 2.63866357421875, 2.63870166015625, 2.638103515625, 2.641052734375, 2.639697998046875, 2.639138916015625, 2.63690234375, 2.63733447265625, 2.637642822265625, 2.63709912109375, 2.6370068359375, 2.638327880859375, 5.4337607421875, 2.63835546875, 2.638138427734375, 2.637888427734375, 2.63785986328125, 2.63816796875, 2.637724609375, 2.63817431640625, 2.63734375, 2.637358154296875, 2.637413330078125, 2.63850390625, 2.63788134765625, 2.638455810546875, 2.63848046875, 2.637498291015625, 2.637077392578125, 2.640005126953125, 2.638171142578125, 2.63721875, 2.638088134765625, 2.637340576171875, 2.637073486328125, 2.637824951171875, 2.63834619140625, 2.63819580078125, 2.63836474609375, 2.63822021484375, 2.637216796875, 2.63757421875, 2.63648046875, 2.6365439453125, 2.63712255859375, 2.63768994140625, 2.636919921875, 2.637600830078125, 2.637790283203125, 2.63785986328125, 2.636936279296875, 2.637821044921875, 2.637937744140625, 2.638752685546875, 2.63861767578125, 2.637978515625, 2.63923291015625, 2.639403076171875, 2.63928515625, 2.63895751953125, 2.63758447265625, 2.637350830078125, 2.63724853515625, 2.6378095703125, 2.638350341796875, 2.637401123046875, 2.636908447265625, 2.63790478515625, 2.638477294921875, 2.638087158203125, 2.639475830078125, 2.639469482421875, 2.640037841796875, 2.637907958984375, 2.636857421875, 5.4322666015625, 2.640575439453125, 2.641816650390625, 2.640384033203125, 2.639365234375, 2.639097900390625, 2.6402724609375, 2.638215087890625, 2.638160888671875, 2.637603759765625, 2.637083740234375, 2.6377421875, 2.637223876953125, 2.6389013671875, 2.638928955078125, 2.637473876953125, 2.637214599609375, 2.636859375, 2.637595703125, 2.63667822265625, 2.6370498046875, 2.638342041015625, 2.63901904296875, 2.637927490234375, 2.63882958984375, 2.640554931640625, 2.64306884765625, 2.639297607421875, 2.6393896484375, 2.64045166015625, 2.63986376953125, 2.639677490234375, 2.64067578125, 2.64060009765625, 2.640331787109375, 2.640183349609375, 2.639626220703125, 2.640194580078125, 2.63760498046875, 2.639280029296875, 2.6397255859375, 2.641039306640625, 2.6397880859375, 2.6377646484375, 2.639326171875, 2.63863916015625, 2.636462158203125, 2.637134765625, 2.6374892578125, 2.637487060546875, 2.63678466796875, 2.63695361328125, 2.636927978515625, 2.636856201171875, 2.63660546875, 2.637761474609375, 2.637675537109375, 2.6367119140625, 2.637622314453125, 2.6384404296875, 2.637370361328125, 2.63602880859375, 2.636900390625]",tokens/s,0.37339678057090686,,,main,False,False,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/gemma-2b,google/gemma-2b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/gemma-2b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aa9ab-3155b8f1179cb3026a9ec5ec;4c4213ae-cbb7-4a21-ad9d-a28df34cd1dd) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/gemma-2b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/gemma-2b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/polyglot-ko-12.8b,EleutherAI/polyglot-ko-12.8b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -698,32 +1056,55 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained - hf_quantizer.postprocess_model(model) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model - return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading - model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 432, in post_init_awq_exllama_modules - model = exllama_post_init(model) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init - submodule.post_init() - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init - self.q4 = exl_ext.make_q4( -NameError: name 'exl_ext' is not defined + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-125m,EleutherAI/gpt-neo-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 -4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-6B,01-ai/Yi-6B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-13b,facebook/opt-13b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -733,32 +1114,96 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained - hf_quantizer.postprocess_model(model) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model - return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading - model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 432, in post_init_awq_exllama_modules - model = exllama_post_init(model) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init - submodule.post_init() - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init - self.q4 = exl_ext.make_q4( -NameError: name 'exl_ext' is not defined + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: OPTForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/recurrentgemma-7b,google/recurrentgemma-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 404 Client Error: Not Found for url: https://huggingface.co/google/recurrentgemma-7b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1823, in _raise_on_head_call_error + raise head_call_error + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 352, in hf_raise_for_status + raise RepositoryNotFoundError(message, response) from e +huggingface_hub.utils._errors.RepositoryNotFoundError: 404 Client Error. (Request ID: Root=1-664aaaab-2ed688ac64f8d5a558c8c046;1a301e85-3907-41e9-9cc1-e2dd23e63ab9) + +Repository Not Found for url: https://huggingface.co/google/recurrentgemma-7b/resolve/main/config.json. +Please make sure you specified the correct `repo_id` and `repo_type`. +If you are trying to access a private or gated repo, make sure you are authenticated. -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 -4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-2.7B,EleutherAI/gpt-neo-2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 422, in cached_file + raise EnvironmentError( +OSError: google/recurrentgemma-7b is not a local folder and is not a valid model identifier listed on 'https://huggingface.co/models' +If this is a private repository, make sure to pass a token having permission to this repo either by logging in with `huggingface-cli login` or by passing `token=` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-410m,EleutherAI/pythia-410m,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -768,635 +1213,394 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained - hf_quantizer.postprocess_model(model) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model - return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading - model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 432, in post_init_awq_exllama_modules - model = exllama_post_init(model) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init - submodule.post_init() - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init - self.q4 = exl_ext.make_q4( -NameError: name 'exl_ext' is not defined + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 -4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciCoder-1b,Deci/DeciCoder-1b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-32B,Qwen/Qwen1.5-32B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,4461.907968,24111.480832,0.0,23465.033728,21690.932224,s,10,26.134937744140625,2.6134937744140623,0.002110944060506743,2.6131138916015626,2.6165289794921875,2.616893395996094,2.617184929199219,"[2.61349365234375, 2.615531005859375, 2.611662353515625, 2.612973388671875, 2.612295654296875, 2.611193115234375, 2.61325439453125, 2.616447998046875, 2.6172578125, 2.610828369140625]",tokens/s,97.95317000798843,kWh,3.0828719950384564e-05,1.6895240036155882e-05,0.0001496905919745939,0.00019741455196113433,tokens/kWh,1296763.5741989252,MB,4461.907968,24111.480832,0.0,23465.033728,21890.217984,s,10,1550.2514375,155.02514375,0.01758597574684522,155.027953125,155.04441406249998,155.04866796875,155.05207109375002,"[155.040484375, 155.052921875, 155.004375, 155.03190625, 155.025, 155.03090625, 155.00465625, 154.996828125, 155.020890625, 155.04346875]",tokens/s,0.4063856899342497,kWh,0.0018302277258535225,0.001003126846851137,0.008909165071770808,0.01174251964447547,tokens/kWh,5365.117701092351,,s,629,1571.3079760742191,2.498104890420062,0.3103664945461323,2.460632080078125,2.46187294921875,2.4623572265625002,5.0725159374999995,"[2.460918701171875, 2.46082666015625, 2.461980712890625, 2.4610048828125, 2.460906494140625, 2.46040576171875, 2.46205126953125, 2.4613447265625, 2.461484130859375, 2.4613662109375, 2.461151123046875, 2.461115478515625, 2.460440673828125, 2.460030029296875, 2.461190185546875, 2.460876708984375, 2.460451904296875, 2.460170166015625, 2.463140869140625, 2.46112060546875, 2.46163671875, 2.46084814453125, 2.463774658203125, 2.461010986328125, 2.4610498046875, 2.46070263671875, 2.461833251953125, 2.460528564453125, 2.4617880859375, 2.46072021484375, 2.461929443359375, 2.4613662109375, 2.46120849609375, 2.46112060546875, 2.4614892578125, 2.459990966796875, 2.461517822265625, 2.4609638671875, 2.46169189453125, 2.460205078125, 2.460928955078125, 2.461189208984375, 2.459634765625, 2.45965625, 2.45981396484375, 2.459525146484375, 2.459683837890625, 2.46053369140625, 2.46042529296875, 2.459734130859375, 2.461192138671875, 2.46074072265625, 2.46057470703125, 2.46108154296875, 2.46078564453125, 2.46036474609375, 2.461035400390625, 2.4617646484375, 2.460655517578125, 2.461253662109375, 2.461116455078125, 2.46118505859375, 5.0779052734375, 2.461738037109375, 2.461781982421875, 2.461154296875, 2.463372314453125, 2.462037109375, 2.462613525390625, 2.462189453125, 2.46213427734375, 2.46221923828125, 2.462265380859375, 2.462834716796875, 2.461442138671875, 2.460442626953125, 2.460788818359375, 2.460884033203125, 2.46097509765625, 2.461346923828125, 2.46034130859375, 2.460333984375, 2.46076318359375, 2.461075439453125, 2.460369873046875, 2.459978759765625, 2.4613203125, 2.461739013671875, 2.460712890625, 2.461740966796875, 2.46230322265625, 2.462064697265625, 2.4611767578125, 2.4604365234375, 2.46226318359375, 2.46073046875, 2.461908935546875, 2.4604140625, 2.4613857421875, 2.460633056640625, 2.46066796875, 2.459797607421875, 2.460780517578125, 2.46072119140625, 2.46026025390625, 2.4598271484375, 2.46049169921875, 2.46009765625, 2.460492919921875, 2.46042529296875, 2.461200439453125, 2.461295654296875, 2.460458984375, 2.461582275390625, 2.46042822265625, 2.460832763671875, 2.461412353515625, 2.460978271484375, 2.460579833984375, 2.4606044921875, 2.460789794921875, 2.460420166015625, 2.4605654296875, 2.461767578125, 2.46121875, 5.0727412109375, 2.4607958984375, 2.46120751953125, 2.4612392578125, 2.460706787109375, 2.46135595703125, 2.46187109375, 2.461371337890625, 2.4605634765625, 2.46158642578125, 2.46133251953125, 2.46209326171875, 2.46131103515625, 2.461284423828125, 2.46093212890625, 2.46143798828125, 2.460158935546875, 2.460937255859375, 2.459987060546875, 2.4602919921875, 2.460261474609375, 2.460375, 2.460521484375, 2.461273193359375, 2.46073046875, 2.46131396484375, 2.460286865234375, 2.46070263671875, 2.459401123046875, 2.460560302734375, 2.460210205078125, 2.4605859375, 2.459334716796875, 2.46042822265625, 2.460444580078125, 2.460652587890625, 2.4594482421875, 2.459885498046875, 2.459210693359375, 2.459854736328125, 2.45916162109375, 2.459714599609375, 2.45982421875, 2.45997265625, 2.460665771484375, 2.461526123046875, 2.459255859375, 2.4603525390625, 2.460240966796875, 2.460114990234375, 2.459881591796875, 2.46009765625, 2.459779052734375, 2.45985791015625, 2.459809814453125, 2.459896728515625, 2.459599853515625, 2.459850830078125, 2.459740234375, 2.459671630859375, 2.459470947265625, 2.45933984375, 2.4605419921875, 5.07296044921875, 2.459979736328125, 2.460074951171875, 2.460875732421875, 2.46111328125, 2.461024169921875, 2.4617041015625, 2.45998583984375, 2.461939697265625, 2.4617861328125, 2.46111328125, 2.460988525390625, 2.460284912109375, 2.460103759765625, 2.4607939453125, 2.461890625, 2.46074072265625, 2.4606279296875, 2.46205224609375, 2.46086865234375, 2.461393798828125, 2.461231201171875, 2.462035888671875, 2.4613251953125, 2.462630859375, 2.460718017578125, 2.46099853515625, 2.460675048828125, 2.4612158203125, 2.460883056640625, 2.461149169921875, 2.461590576171875, 2.4607939453125, 2.460654541015625, 2.460872802734375, 2.461231201171875, 2.46080712890625, 2.461241455078125, 2.46087158203125, 2.459632568359375, 2.460600341796875, 2.460632080078125, 2.46099365234375, 2.460303466796875, 2.46082568359375, 2.462000244140625, 2.461948974609375, 2.4596806640625, 2.460571533203125, 2.460125244140625, 2.46019677734375, 2.45987841796875, 2.46080810546875, 2.460212158203125, 2.460813232421875, 2.45957421875, 2.46019384765625, 2.460409912109375, 2.459958251953125, 2.45973095703125, 2.460412841796875, 2.46169384765625, 2.46048046875, 5.07272607421875, 2.461116455078125, 2.460166259765625, 2.461013916015625, 2.460505126953125, 2.46078662109375, 2.461305908203125, 2.4609658203125, 2.460851318359375, 2.461484130859375, 2.4605869140625, 2.4610498046875, 2.4601630859375, 2.46051318359375, 2.4606064453125, 2.461643798828125, 2.461190185546875, 2.461765625, 2.46075390625, 2.4614091796875, 2.460251220703125, 2.4604609375, 2.460229736328125, 2.460904541015625, 2.46035546875, 2.460232666015625, 2.46175439453125, 2.462803955078125, 2.4607958984375, 2.461580322265625, 2.461107177734375, 2.461393798828125, 2.460929931640625, 2.460676025390625, 2.461404052734375, 2.460982177734375, 2.460367919921875, 2.4600791015625, 2.4596552734375, 2.460010498046875, 2.45901318359375, 2.4592802734375, 2.46055224609375, 2.460517333984375, 2.459428955078125, 2.46019384765625, 2.460880859375, 2.4609208984375, 2.459928466796875, 2.46034423828125, 2.460252197265625, 2.460788818359375, 2.460273681640625, 2.45994287109375, 2.460907470703125, 2.4596796875, 2.460264404296875, 2.460949462890625, 2.460282958984375, 2.4613427734375, 2.462763916015625, 2.461613037109375, 2.460632080078125, 5.0719755859375, 2.462064697265625, 2.46147998046875, 2.46153515625, 2.461073486328125, 2.460905517578125, 2.461240234375, 2.460853271484375, 2.461421630859375, 2.462041015625, 2.46237890625, 2.46096484375, 2.461697998046875, 2.4613037109375, 2.461698974609375, 2.460219482421875, 2.462074951171875, 2.460283935546875, 2.459872314453125, 2.46007080078125, 2.45981689453125, 2.460726318359375, 2.459724853515625, 2.45985888671875, 2.460906494140625, 2.46042822265625, 2.46017626953125, 2.460051513671875, 2.46018359375, 2.46103759765625, 2.460602294921875, 2.46068212890625, 2.461318115234375, 2.460810302734375, 2.4603740234375, 2.46188037109375, 2.462437255859375, 2.460948486328125, 2.460538818359375, 2.460599365234375, 2.461042724609375, 2.46105712890625, 2.45985595703125, 2.46139697265625, 2.46086865234375, 2.460706787109375, 2.46046826171875, 2.4603740234375, 2.461024169921875, 2.4599951171875, 2.460008544921875, 2.460411865234375, 2.4604580078125, 2.460127197265625, 2.4596591796875, 2.460818359375, 2.461365234375, 2.460373046875, 2.46135302734375, 2.461212646484375, 2.4607734375, 2.460324951171875, 2.46021728515625, 5.07445458984375, 2.4605234375, 2.461729736328125, 2.461869140625, 2.461232177734375, 2.461845458984375, 2.463498291015625, 2.460968994140625, 2.46013232421875, 2.460211181640625, 2.45943603515625, 2.460124267578125, 2.46013037109375, 2.460103759765625, 2.461614013671875, 2.46322900390625, 2.461684814453125, 2.4606064453125, 2.4600556640625, 2.45979345703125, 2.459621337890625, 2.460230712890625, 2.459693115234375, 2.460508056640625, 2.460240966796875, 2.460556396484375, 2.459707275390625, 2.459610107421875, 2.459715576171875, 2.460527587890625, 2.459706298828125, 2.45965625, 2.45981591796875, 2.45966845703125, 2.4598037109375, 2.459366455078125, 2.4594267578125, 2.45954052734375, 2.46019580078125, 2.459505615234375, 2.4598701171875, 2.4605830078125, 2.46051953125, 2.4611103515625, 2.460297119140625, 2.45998388671875, 2.459361328125, 2.460695556640625, 2.46107958984375, 2.46091162109375, 2.45956494140625, 2.460255126953125, 2.460651611328125, 2.460200927734375, 2.45922314453125, 2.460324951171875, 2.460303466796875, 2.46060546875, 2.460147705078125, 2.459874267578125, 2.459558837890625, 2.460209228515625, 2.46225927734375, 5.07727880859375, 2.46051953125, 2.459748291015625, 2.460139404296875, 2.459675537109375, 2.460464111328125, 2.45935400390625, 2.460265380859375, 2.460979248046875, 2.46025927734375, 2.460180419921875, 2.45945654296875, 2.45956396484375, 2.4599091796875, 2.459864990234375, 2.459249755859375, 2.459658203125, 2.45960693359375, 2.4602666015625, 2.463476806640625, 2.46247314453125, 2.461254638671875, 2.4606064453125, 2.46013134765625, 2.4603095703125, 2.46048046875, 2.460453857421875, 2.460747802734375, 2.4610087890625, 2.45964501953125, 2.460734375, 2.460478515625, 2.460600341796875, 2.460285888671875, 2.460894287109375, 2.460316650390625, 2.460169189453125, 2.460146728515625, 2.459640869140625, 2.459826171875, 2.459534423828125, 2.46048876953125, 2.459629638671875, 2.459568115234375, 2.459454345703125, 2.46009765625, 2.46054296875, 2.46160693359375, 2.459303955078125, 2.459570068359375, 2.460251220703125, 2.459474853515625, 2.45965625, 2.460379150390625, 2.45972900390625, 2.459229248046875, 2.459660400390625, 2.45932958984375, 2.464058349609375, 2.461569091796875, 2.45967041015625, 2.459884521484375, 2.460506103515625, 5.07673388671875, 2.459622314453125, 2.462644287109375, 2.4620205078125, 2.461388916015625, 2.460607421875, 2.461044677734375, 2.461013916015625, 2.461294677734375, 2.460347412109375, 2.46035546875, 2.4605205078125, 2.460373046875, 2.459558837890625, 2.46158740234375, 2.462236572265625, 2.462750732421875, 2.46039453125, 2.459875244140625, 2.460180419921875, 2.459242431640625, 2.461361083984375, 2.461865966796875, 2.4626064453125, 2.46232470703125, 2.462096435546875, 2.460041259765625, 2.4603125, 2.460031005859375, 2.460012451171875, 2.460483642578125, 2.46249365234375, 2.460971923828125, 2.460064697265625, 2.460525634765625, 2.4618515625, 2.461947998046875, 2.46048046875, 2.460482666015625, 2.459716552734375, 2.46003515625, 2.45914111328125, 2.459892822265625, 2.4607099609375, 2.461216796875, 2.45960693359375, 2.459496337890625, 2.459989013671875, 2.46011181640625, 2.461148193359375, 2.459788330078125, 2.459845703125, 2.45964501953125, 2.45973291015625, 2.459387939453125, 2.460958740234375, 2.4599326171875, 2.459988037109375, 2.459134033203125, 2.4599716796875, 2.459010009765625, 2.4597216796875, 2.466093017578125, 5.07026025390625, 2.459715576171875, 2.459428955078125, 2.460008544921875, 2.45975244140625, 2.460271728515625, 2.4604365234375, 2.46042724609375, 2.46135595703125, 2.460927001953125, 2.460580810546875, 2.46221826171875, 2.461213623046875, 2.461199462890625, 2.460501953125, 2.461664306640625, 2.46034423828125, 2.460695556640625, 2.460241943359375, 2.46086865234375, 2.4613837890625, 2.4614892578125, 2.460771240234375, 2.461020263671875, 2.462573486328125, 2.461728759765625, 2.461054931640625, 2.46047021484375, 2.46080712890625, 2.46276806640625, 2.46089306640625, 2.461909912109375, 2.46152392578125, 2.4612607421875, 2.460707763671875, 2.461158447265625, 2.46102734375, 2.46084716796875, 2.460749755859375, 2.460251220703125, 2.460735595703125, 2.46051220703125, 2.461053955078125, 2.4607548828125, 2.46175244140625, 2.460388427734375, 2.46027880859375, 2.4614501953125, 2.461758544921875, 2.461529052734375, 2.460019775390625, 2.4599306640625, 2.460873779296875, 2.460814453125, 2.460541015625, 2.460760009765625, 2.46168994140625, 2.460673095703125, 2.46126806640625, 2.460421142578125, 2.462738525390625, 2.46156689453125, 2.461497314453125]",tokens/s,0.4003034475593408,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-4.5B,facebook/xglm-4.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 64, in run - report = scenario.run(backend) - File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run - _ = backend.generate(self.inputs, self.config.generate_kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate - return self.pretrained_model.generate(**inputs, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1576, in generate - result = self._greedy_search( - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2494, in _greedy_search - outputs = self( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1211, in forward - outputs = self.model( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1018, in forward - layer_outputs = decoder_layer( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 741, in forward - hidden_states, self_attn_weights, present_key_value = self.self_attn( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) -TypeError: DeciCoderAttention.forward() got an unexpected keyword argument 'cache_position' + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: XGLMForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciLM-7B,,cuda,0,42,,,,,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,main,False,False,MB,1674.596352,5516.034048,0.0,4869.586944,4743.593472,s,10,6.137113647460938,0.6137113647460938,0.0032895717697753717,0.6129547424316406,0.6153997436523437,0.6190885131835938,0.6220395288085938,"[0.6227772827148438, 0.6144583129882812, 0.61261669921875, 0.61060546875, 0.612577392578125, 0.6105811157226563, 0.6121093139648438, 0.6145800170898438, 0.6132927856445313, 0.6135152587890625]",tokens/s,417.13420136176387,kWh,7.2172448039054885e-06,3.954746732982612e-06,3.355963469082604e-05,4.473162622771414e-05,tokens/kWh,5723020.189268045,MB,1674.596352,5516.034048,0.0,4869.586944,4769.651712,s,10,361.12618359375006,36.112618359375006,0.009999043540991457,36.112980468749996,36.125964453125,36.127720507812505,36.1291253515625,"[36.1294765625, 36.0964296875, 36.12557421875, 36.1174140625, 36.11404296875, 36.1114296875, 36.11191796875, 36.09769140625, 36.10714453125, 36.1150625]",tokens/s,1.7445425688343892,kWh,0.00042614284929302004,0.0002335616249413365,0.0019425026683799653,0.0026022071426143214,tokens/kWh,24210.217153084406,,s,629,366.0647033081055,0.5819788605852233,0.07282065333686359,0.5730969848632812,0.5741784912109374,0.5745491943359375,1.1854891552734375,"[0.5736980590820312, 0.573138916015625, 0.5735465087890625, 0.574129150390625, 0.5737092895507813, 0.57447216796875, 0.5745264892578125, 0.5731635131835937, 0.5737205810546875, 0.5735956420898437, 0.5749248046875, 0.5725880126953125, 0.5736376342773437, 0.57371337890625, 0.573749267578125, 0.5733990478515625, 0.5727344360351563, 0.57394482421875, 0.57497705078125, 0.5740851440429687, 0.5741229858398438, 0.5742008056640625, 0.5734738159179688, 0.574509033203125, 0.5738157958984375, 0.573897705078125, 0.5732402954101562, 0.5745704956054688, 0.5727620849609375, 0.5734061889648437, 0.5746319580078125, 0.5745643310546875, 0.5739735107421875, 0.5731901245117188, 0.5725511474609375, 0.5742970581054687, 0.5729157104492187, 0.572837890625, 0.57303857421875, 0.5725552368164063, 0.5740175170898437, 0.5737943115234375, 0.57314404296875, 0.5743441772460938, 0.5747333374023438, 0.5747230834960938, 0.5726239013671875, 0.5742520141601563, 0.5729638671875, 0.5725716552734375, 0.572348388671875, 0.572031005859375, 0.5730672607421875, 0.5724190673828125, 0.572590087890625, 0.5723648071289062, 0.5724487915039063, 0.572526611328125, 0.5727549438476562, 0.57301708984375, 0.5722869873046875, 0.5721016235351563, 1.18763623046875, 0.5728788452148438, 0.5735679931640625, 0.5728031005859375, 0.5727078247070313, 0.5732781982421875, 0.5730969848632812, 0.5728460693359375, 0.57259521484375, 0.5722409057617187, 0.57280615234375, 0.57242724609375, 0.57225830078125, 0.5726760864257813, 0.572142578125, 0.5723013305664062, 0.5727958984375, 0.5730846557617187, 0.572316650390625, 0.5732341918945313, 0.57354443359375, 0.5724467163085938, 0.5723648071289062, 0.5720924072265625, 0.5729249267578125, 0.5723299560546875, 0.5724815063476563, 0.572590087890625, 0.5723525390625, 0.5726453857421875, 0.572458984375, 0.5725296630859374, 0.5720729370117188, 0.572706787109375, 0.5740257568359375, 0.572758056640625, 0.5724251708984375, 0.5737236328125, 0.57303857421875, 0.5752218017578125, 0.5730856323242187, 0.57284814453125, 0.5734194946289063, 0.5729658813476562, 0.5726095581054688, 0.5726730346679687, 0.5732310791015625, 0.5731287231445312, 0.5729985961914063, 0.572822509765625, 0.5728573608398437, 0.5731840209960938, 0.5744937133789062, 0.5729187622070312, 0.5740841064453125, 0.5739561157226563, 0.5734850463867187, 0.5729812622070313, 0.5727211303710937, 0.5733560180664062, 0.5732577514648437, 0.5737369384765625, 0.5736038208007812, 1.18655078125, 0.5729157104492187, 0.5731143798828126, 0.5728522338867188, 0.5734451293945313, 0.5739530029296875, 0.572958740234375, 0.572759033203125, 0.5748234252929687, 0.57375537109375, 0.573359130859375, 0.5733130493164063, 0.5738352661132813, 0.57318603515625, 0.5733314819335937, 0.5734747924804687, 0.5750180053710937, 0.5730693359375, 0.5736345825195313, 0.573718505859375, 0.57335498046875, 0.5729044189453125, 0.5738772583007813, 0.5738352661132813, 0.5730682983398437, 0.5731051635742187, 0.5734297485351563, 0.5730928344726562, 0.5731317749023438, 0.5730283813476562, 0.573849609375, 0.57297509765625, 0.5731461181640625, 0.5727354736328125, 0.5730130004882813, 0.5731461181640625, 0.5728123168945313, 0.5729003295898437, 0.5728778076171875, 0.5738741455078125, 0.5740830688476563, 0.572737548828125, 0.5726536254882812, 0.5742734985351563, 0.5730897827148438, 0.5729822998046875, 0.5732136840820312, 0.5756242065429688, 0.5729976196289063, 0.5738700561523438, 0.573106201171875, 0.573149169921875, 0.5729013671875, 0.573991943359375, 0.573201416015625, 0.5733877563476563, 0.5732608032226563, 0.5743902587890625, 0.5742141723632812, 0.5728829345703125, 0.5743646850585937, 0.5734010620117187, 0.573259765625, 1.1840142822265625, 0.5737728271484375, 0.5731215209960937, 0.5732106323242188, 0.5732505493164063, 0.572969970703125, 0.573033447265625, 0.57314306640625, 0.5731942138671875, 0.5728460693359375, 0.5744957275390625, 0.5742643432617187, 0.5740144653320313, 0.5731768798828125, 0.5741864624023437, 0.5734174194335937, 0.573322265625, 0.5733375854492188, 0.5739192504882813, 0.5734788818359375, 0.5737677001953125, 0.5734850463867187, 0.5734430541992187, 0.5732044677734375, 0.5726986083984374, 0.572821533203125, 0.5730499267578125, 0.5735310668945313, 0.574656494140625, 0.5733201904296875, 0.5725225219726563, 0.5739570922851562, 0.5737728271484375, 0.57303857421875, 0.5733232421875, 0.5734522705078124, 0.5738311767578125, 0.575151123046875, 0.5733201904296875, 0.5735167846679687, 0.5731133422851562, 0.5729003295898437, 0.5728235473632812, 0.573318115234375, 0.5730816040039063, 0.5742418212890625, 0.573048828125, 0.5728880615234375, 0.572990478515625, 0.5729679565429687, 0.5725675659179688, 0.5726566162109376, 0.5728235473632812, 0.5745910034179688, 0.5722357788085938, 0.5722327270507812, 0.572821533203125, 0.5726668701171875, 0.5724385375976563, 0.5728890991210938, 0.5728051147460937, 0.5724405517578125, 0.5733775634765625, 1.1864791259765626, 0.5730631713867187, 0.5741793212890625, 0.574244873046875, 0.5729197998046875, 0.5730785522460937, 0.5742643432617187, 0.5736028442382812, 0.5729208374023438, 0.5731768188476563, 0.573612060546875, 0.5728307495117188, 0.5726996459960938, 0.57314306640625, 0.5741906127929688, 0.572669921875, 0.5738731689453125, 0.5743124389648437, 0.5728778076171875, 0.5729352416992187, 0.574119873046875, 0.5733079223632812, 0.5724866333007812, 0.572231689453125, 0.5732946166992188, 0.5724682006835937, 0.5730529174804687, 0.5733519287109375, 0.573106201171875, 0.5727999877929687, 0.57236376953125, 0.5724620971679687, 0.5725665893554688, 0.5737697143554688, 0.5733406372070312, 0.573254638671875, 0.5726945190429688, 0.5743114013671875, 0.5735679931640625, 0.5726617431640625, 0.5724610595703125, 0.57396533203125, 0.5732372436523437, 0.5741332397460938, 0.5733570556640625, 0.5733375854492188, 0.572717041015625, 0.572859375, 0.5731901245117188, 0.572558349609375, 0.5723750610351562, 0.5745213623046875, 0.573633544921875, 0.5729638671875, 0.573412353515625, 0.5735505981445312, 0.5726494750976563, 0.5753558959960937, 0.5728399658203125, 0.573432861328125, 0.5723391723632812, 0.5729290161132813, 0.572626953125, 1.1842232666015624, 0.5728092041015626, 0.572779541015625, 0.572632080078125, 0.5737963256835937, 0.5736775512695312, 0.5726033935546875, 0.572564453125, 0.5733416748046875, 0.5725654907226563, 0.5723832397460937, 0.5723812255859375, 0.5731624755859375, 0.5723832397460937, 0.5729197998046875, 0.572788818359375, 0.5723237915039062, 0.5723801879882813, 0.5726771850585938, 0.5727251586914063, 0.5726986083984374, 0.57333349609375, 0.5739939575195312, 0.5731481323242188, 0.5727344360351563, 0.5737195434570312, 0.5727918090820312, 0.5725234985351563, 0.5729740600585937, 0.5731522827148438, 0.5736734619140625, 0.5729310913085938, 0.5734471435546875, 0.573127685546875, 0.5729720458984375, 0.5730723876953125, 0.5730529174804687, 0.5730682983398437, 0.5760235595703125, 0.5742151489257813, 0.573750244140625, 0.5731143798828126, 0.573844482421875, 0.5748480224609375, 0.5746339721679687, 0.5739223022460938, 0.5732044677734375, 0.5739735107421875, 0.5730263061523437, 0.5728583984375, 0.5729924926757812, 0.5730785522460937, 0.5734502563476562, 0.5732157592773437, 0.573191162109375, 0.5728983154296875, 0.5738905639648437, 0.5733457641601563, 0.57269873046875, 0.572650390625, 0.5734747924804687, 0.5728338012695312, 0.5733191528320313, 1.1859814453125, 0.57253271484375, 0.5722654418945312, 0.5727242431640625, 0.5725839233398438, 0.5722265625, 0.572601318359375, 0.5729863891601562, 0.572706787109375, 0.5726239013671875, 0.5739888916015625, 0.5729249267578125, 0.573317138671875, 0.5733673095703125, 0.5743882446289063, 0.5732106323242188, 0.5734819946289063, 0.5727047729492187, 0.5730140380859375, 0.572675048828125, 0.5729136352539063, 0.5729924926757812, 0.5723709716796875, 0.5744793701171875, 0.57339599609375, 0.5728604125976563, 0.5722327270507812, 0.5737615356445313, 0.5753026733398438, 0.5738475341796875, 0.5734512939453125, 0.5752913818359375, 0.5733673095703125, 0.5729085693359375, 0.5732177734375, 0.573886474609375, 0.5727651977539062, 0.57282763671875, 0.57284814453125, 0.5731399536132813, 0.5727467651367187, 0.5733314819335937, 0.5736110229492187, 0.5734830322265625, 0.5729003295898437, 0.574867431640625, 0.5740697631835937, 0.572969970703125, 0.5728942260742188, 0.5735751342773437, 0.5731563720703124, 0.5731113891601562, 0.5729085083007812, 0.5745828247070313, 0.5727620849609375, 0.5729269409179687, 0.5724876708984376, 0.573665283203125, 0.5732260131835938, 0.5726268920898437, 0.57246923828125, 0.572621826171875, 0.5729013671875, 1.1888896484375, 0.57364990234375, 0.5738916015625, 0.5726515502929688, 0.5726239013671875, 0.5722838745117188, 0.5730723876953125, 0.573032470703125, 0.5727662353515625, 0.5732413330078125, 0.5728880615234375, 0.5727396240234375, 0.5731368408203125, 0.573233154296875, 0.5731963500976562, 0.5735249633789062, 0.5744219970703125, 0.5727999877929687, 0.5726064453125, 0.5729290771484375, 0.5737103271484375, 0.572811279296875, 0.5732679443359375, 0.5726730346679687, 0.5734287109375, 0.5722675170898438, 0.5725828857421875, 0.5725849609375, 0.5723627319335938, 0.5727416381835938, 0.5728348388671874, 0.5723678588867187, 0.5728604125976563, 0.573675537109375, 0.57333349609375, 0.5725634765625, 0.5726546020507812, 0.5731399536132813, 0.5728031005859375, 0.5724630737304688, 0.57310205078125, 0.5734573974609375, 0.5732689819335938, 0.5729290161132813, 0.5729782104492187, 0.5726064453125, 0.572527587890625, 0.5728818969726562, 0.5733007202148438, 0.5730426635742187, 0.5736365966796875, 0.5730549926757813, 0.573111328125, 0.5725552368164063, 0.5726505126953125, 0.5732802734375, 0.5726494750976563, 0.5724876708984376, 0.5733058471679687, 0.5727211303710937, 0.5725931396484375, 0.5726648559570312, 0.5728655395507812, 1.1863818359375, 0.5727938842773438, 0.5733508911132813, 0.5740676879882812, 0.574234619140625, 0.5727047729492187, 0.5733447875976563, 0.5732515869140625, 0.5743012084960938, 0.5728123168945313, 0.5730969848632812, 0.5729976196289063, 0.5739694213867188, 0.5728798828125, 0.5730140380859375, 0.5724334106445312, 0.5727744140625, 0.5740390625, 0.5736099853515625, 0.5726300048828125, 0.572969970703125, 0.5736683349609375, 0.572948486328125, 0.5725542602539062, 0.5725962524414062, 0.5727139892578125, 0.573559814453125, 0.572568603515625, 0.5725419311523438, 0.5730426635742187, 0.5727252197265625, 0.5726781616210938, 0.5734676513671875, 0.573179931640625, 0.5726351318359375, 0.573233154296875, 0.5737062377929687, 0.5729423217773437, 0.5729710083007813, 0.5735885009765626, 0.5732741088867187, 0.5730344848632812, 0.5727262573242188, 0.5732567138671875, 0.5732689819335938, 0.5727232055664062, 0.5727559814453125, 0.5730703125, 0.5724334716796875, 0.5731378784179687, 0.5730191650390625, 0.572896240234375, 0.5732177734375, 0.5728727416992188, 0.5729361572265625, 0.5730979614257813, 0.574118896484375, 0.573497314453125, 0.5732976684570312, 0.5737840576171875, 0.5730130004882813, 0.5730549926757813, 0.572958740234375, 1.187092529296875, 0.5724129028320313, 0.5726473999023437, 0.57280615234375, 0.572416015625, 0.5723361206054688, 0.572821533203125, 0.5732761840820313, 0.5727733764648437, 0.573095947265625, 0.5734052124023438, 0.573053955078125, 0.572674072265625, 0.5729341430664062, 0.5734830322265625, 0.573849609375, 0.5724999389648437, 0.5727334594726563, 0.5741107177734375, 0.5736663208007813, 0.57402978515625, 0.5733611450195313, 0.572958740234375, 0.5730099487304687, 0.5733345336914063, 0.5739038696289063, 0.575910888671875, 0.574392333984375, 0.5739898681640625, 0.5732730712890625, 0.5733109741210938, 0.5730928344726562, 0.5741782836914062, 0.5731993408203125, 0.57302734375, 0.5732771606445313, 0.572506103515625, 0.5729556274414063, 0.5733088989257813, 0.572968994140625, 0.5736283569335937, 0.5738291015625, 0.5733499145507812, 0.5729085693359375, 0.573539306640625, 0.5743226928710937, 0.5736959838867187, 0.5733365478515625, 0.57354443359375, 0.5733949584960938, 0.5731727294921874, 0.5731215209960937, 0.5734061889648437, 0.5725133056640626, 0.5726607055664062, 0.5729464111328125, 0.5725101928710937, 0.57212109375, 0.5734563598632813, 0.5726597290039063, 0.5727313842773437, 0.572416015625, 0.5735731201171875]",tokens/s,1.7182754696526694,,,,, -4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-j-6b,,cuda,0,42,,,,,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,main,False,False,MB,2082.578432,5566.365696,0.0,4919.918592,4635.53792,s,10,5.0974686889648435,0.5097468688964844,0.002523113409812242,0.5091204681396484,0.5118887878417968,0.5140431793212891,0.5157666925048828,"[0.5161975708007812, 0.508827880859375, 0.5094269104003907, 0.508350830078125, 0.5081865539550782, 0.5073575439453125, 0.5094130554199219, 0.5110750427246094, 0.5072232666015625, 0.5114100341796874]",tokens/s,502.2100489880333,kWh,5.994651640454928e-06,3.28480666631549e-06,2.7987258500894452e-05,3.726671680766487e-05,tokens/kWh,6869400.417569034,MB,2082.578432,5566.365696,0.0,4919.918592,4794.464768,s,10,295.3985,29.53985,0.004666201106872227,29.538681640625,29.5469361328125,29.54772978515625,29.548364707031247,"[29.534837890625, 29.546759765625, 29.542654296875, 29.540826171875, 29.5363359375, 29.535224609375, 29.54083984375, 29.5359609375, 29.5485234375, 29.536537109375]",tokens/s,2.132712251416307,kWh,0.0003485870901164082,0.0001910528620806053,0.0016025825737317144,0.002142222525928728,tokens/kWh,29408.709523623045,,s,629,299.5096441345215,0.47616795569876236,0.06052059742525611,0.46878207397460936,0.46948126220703124,0.4697759826660156,0.977483603515625,"[0.4688783264160156, 0.46847384643554685, 0.46878515625, 0.4694343566894531, 0.4695541687011719, 0.46879034423828125, 0.4688465270996094, 0.46848306274414064, 0.46828851318359377, 0.4686806945800781, 0.46836428833007815, 0.4685475769042969, 0.4686540832519531, 0.46948455810546874, 0.4689162292480469, 0.46845745849609377, 0.4684011535644531, 0.4683714599609375, 0.4684134521484375, 0.46870220947265623, 0.468790283203125, 0.4682076110839844, 0.4684062805175781, 0.46837042236328125, 0.469607421875, 0.46908721923828123, 0.4688885803222656, 0.4687656860351562, 0.4693012390136719, 0.4686929931640625, 0.46864382934570314, 0.468430908203125, 0.4682413330078125, 0.4685291442871094, 0.46904525756835935, 0.4706498413085938, 0.46837042236328125, 0.4685875244140625, 0.4687718505859375, 0.4685537414550781, 0.46882406616210937, 0.46856805419921876, 0.4686806945800781, 0.46866943359375, 0.4686581726074219, 0.4682403869628906, 0.4695777893066406, 0.4685884704589844, 0.4686663818359375, 0.4685537414550781, 0.46855987548828126, 0.4685066223144531, 0.4687831115722656, 0.46864794921875, 0.4684666748046875, 0.4691046447753906, 0.46919476318359377, 0.4684984436035156, 0.4691128234863281, 0.4691221008300781, 0.46927557373046874, 0.46959002685546875, 0.977623046875, 0.46814208984375, 0.4683407287597656, 0.46863873291015623, 0.46877490234375, 0.46880459594726565, 0.4687513732910156, 0.46857113647460935, 0.468421630859375, 0.46894284057617186, 0.4688281555175781, 0.46850765991210935, 0.4687431640625, 0.4691404724121094, 0.4699504699707031, 0.46978561401367186, 0.46937701416015626, 0.4690975341796875, 0.4691311950683594, 0.47030169677734374, 0.4690933837890625, 0.4689039306640625, 0.4690298767089844, 0.46860595703125, 0.4687359924316406, 0.4692275085449219, 0.46936166381835936, 0.46901248168945314, 0.46843902587890623, 0.46878106689453125, 0.46938113403320314, 0.46973849487304686, 0.46935861206054685, 0.4687390441894531, 0.4685845031738281, 0.46890591430664064, 0.46899813842773436, 0.4688281555175781, 0.4687083435058594, 0.4687564697265625, 0.4687308654785156, 0.4687615966796875, 0.4689858703613281, 0.4688762817382813, 0.46893264770507814, 0.46929302978515625, 0.46915994262695315, 0.4696995849609375, 0.4688701477050781, 0.4687912902832031, 0.4691885986328125, 0.46969558715820314, 0.4696114196777344, 0.4692490234375, 0.46899301147460937, 0.46896435546875, 0.4694558715820312, 0.46925927734375, 0.4685393981933594, 0.4686530456542969, 0.46877902221679685, 0.4687155151367188, 0.46934527587890623, 0.9787330322265625, 0.468917236328125, 0.46886502075195313, 0.46889984130859375, 0.4691875915527344, 0.4684267578125, 0.46861004638671877, 0.46856500244140625, 0.46833560180664063, 0.4684933166503906, 0.4683735046386719, 0.4683263854980469, 0.4683735046386719, 0.468236328125, 0.46872265625, 0.4684236755371094, 0.46842572021484374, 0.46854452514648437, 0.46859982299804687, 0.4687155151367188, 0.46844732666015626, 0.4685188293457031, 0.4686991577148438, 0.46980194091796873, 0.46855987548828126, 0.4690032653808594, 0.46884454345703125, 0.4687575073242187, 0.4688609313964844, 0.4688558044433594, 0.4689459228515625, 0.46866329956054686, 0.4685557861328125, 0.46875238037109374, 0.4686827392578125, 0.46909747314453126, 0.46889984130859375, 0.4689756164550781, 0.46924798583984373, 0.46901144409179685, 0.4689469299316406, 0.46897048950195314, 0.46984909057617186, 0.4700190734863281, 0.4696781005859375, 0.469550048828125, 0.4697763977050781, 0.47159500122070314, 0.46917837524414063, 0.4684482421875, 0.4692193298339844, 0.46906060791015625, 0.4689141845703125, 0.46894284057617186, 0.468885498046875, 0.469064697265625, 0.46948043823242186, 0.4687861633300781, 0.4690513916015625, 0.46873907470703124, 0.46875341796875, 0.4690708618164062, 0.46917938232421874, 0.9775615844726563, 0.46874624633789064, 0.46858853149414065, 0.4686796875, 0.46987161254882814, 0.46889370727539065, 0.4687656860351562, 0.468790283203125, 0.4687278137207031, 0.46867864990234376, 0.4685322265625, 0.46837454223632813, 0.4686530456542969, 0.46875955200195313, 0.46893466186523436, 0.468706298828125, 0.4689858703613281, 0.46853018188476564, 0.46881585693359373, 0.46894796752929685, 0.4691353454589844, 0.46871038818359373, 0.46851071166992186, 0.468632568359375, 0.46843902587890623, 0.46874725341796875, 0.468890625, 0.4688670654296875, 0.46844723510742187, 0.468490234375, 0.4689090576171875, 0.46941900634765626, 0.46909030151367187, 0.4689100952148437, 0.4687175598144531, 0.46938113403320314, 0.4690616455078125, 0.4687718505859375, 0.46904730224609376, 0.4689336242675781, 0.4687933349609375, 0.46920501708984375, 0.46855471801757814, 0.4687718505859375, 0.47131646728515625, 0.4688424987792969, 0.46912005615234376, 0.4692090148925781, 0.4688189392089844, 0.4685926513671875, 0.4688619384765625, 0.46930743408203124, 0.4693247680664063, 0.4690636901855469, 0.46892135620117187, 0.4687145690917969, 0.4691373291015625, 0.4687145080566406, 0.4686673889160156, 0.46867352294921877, 0.46872677612304686, 0.4689254455566406, 0.46927871704101565, 0.9772830810546875, 0.46864077758789063, 0.4684922790527344, 0.46857421875, 0.46906878662109375, 0.4690033264160156, 0.46841543579101563, 0.4685895690917969, 0.46849639892578127, 0.468279296875, 0.46869912719726564, 0.4685619201660156, 0.4687575073242187, 0.46905242919921875, 0.46849432373046873, 0.46846875, 0.46872576904296875, 0.46877490234375, 0.4686592102050781, 0.4684892272949219, 0.4687974548339844, 0.46863565063476564, 0.4685844421386719, 0.46846157836914065, 0.46825778198242185, 0.4684994506835938, 0.4686448669433594, 0.46830081176757815, 0.46856805419921876, 0.4685209655761719, 0.46874929809570315, 0.4689264526367187, 0.46845541381835937, 0.4686520385742188, 0.46867770385742186, 0.46937490844726565, 0.46933709716796873, 0.4687554626464844, 0.4690616455078125, 0.46915789794921875, 0.46891213989257813, 0.46888754272460936, 0.4687278137207031, 0.46915277099609376, 0.4688332824707031, 0.4693790588378906, 0.46926437377929686, 0.46878515625, 0.4690831298828125, 0.46895718383789065, 0.4689776611328125, 0.4690616455078125, 0.46907391357421874, 0.46851788330078126, 0.46869195556640625, 0.4686376953125, 0.46900634765625, 0.4722012023925781, 0.4687718505859375, 0.4688670654296875, 0.46880459594726565, 0.468642822265625, 0.4693155822753906, 0.9763594360351563, 0.46846771240234375, 0.46886502075195313, 0.4687083740234375, 0.4689837646484375, 0.4684646301269531, 0.4687247314453125, 0.46836224365234375, 0.4683929748535156, 0.4685619201660156, 0.46836737060546874, 0.4687503356933594, 0.46857830810546874, 0.46878411865234376, 0.4683345947265625, 0.468358154296875, 0.46845849609375, 0.4687032470703125, 0.4686520385742188, 0.4686458740234375, 0.46885784912109374, 0.46872677612304686, 0.4685823974609375, 0.46873095703125, 0.4685475158691406, 0.468790283203125, 0.46927053833007815, 0.46861822509765627, 0.46847589111328125, 0.46836224365234375, 0.470181884765625, 0.4690575256347656, 0.4688353271484375, 0.4692449340820313, 0.4685137939453125, 0.46857217407226565, 0.46875238037109374, 0.46873190307617185, 0.4689664001464844, 0.4691271667480469, 0.4686090087890625, 0.4692101135253906, 0.4687974548339844, 0.469317626953125, 0.4686315612792969, 0.46899917602539065, 0.4692490234375, 0.4690309143066406, 0.4687646789550781, 0.46850253295898436, 0.4687083435058594, 0.46883941650390626, 0.46927871704101565, 0.4686152038574219, 0.4694292297363281, 0.4691435546875, 0.46974566650390626, 0.4686940307617187, 0.468969482421875, 0.46890188598632815, 0.46895001220703125, 0.4687564697265625, 0.46904730224609376, 0.9786060791015625, 0.46841650390625, 0.4683786315917969, 0.46915994262695315, 0.46917938232421874, 0.4692777099609375, 0.46883636474609375, 0.468716552734375, 0.46857318115234375, 0.4691128234863281, 0.468600830078125, 0.4694783935546875, 0.46952960205078126, 0.4696739807128906, 0.4687196044921875, 0.46852505493164065, 0.46841854858398435, 0.4683591613769531, 0.4683100280761719, 0.4684431457519531, 0.4690411376953125, 0.4687923278808594, 0.4687216491699219, 0.46836224365234375, 0.4686090087890625, 0.4691363830566406, 0.46857113647460935, 0.46862335205078126, 0.46869094848632814, 0.4686581726074219, 0.4694435729980469, 0.4694640502929687, 0.46918450927734373, 0.46924288940429687, 0.4694343566894531, 0.4690462646484375, 0.469317626953125, 0.46890188598632815, 0.46938323974609375, 0.4693267822265625, 0.46866329956054686, 0.4690370483398438, 0.468864013671875, 0.4687575073242187, 0.4686438598632813, 0.46876776123046876, 0.46984698486328125, 0.4690380859375, 0.46935552978515627, 0.46853836059570314, 0.46839910888671876, 0.46874008178710935, 0.46875955200195313, 0.4687667236328125, 0.4688209838867187, 0.4691302490234375, 0.46921624755859376, 0.4689674377441406, 0.46906777954101564, 0.46867864990234376, 0.46859161376953123, 0.46864794921875, 0.4688332824707031, 0.980326416015625, 0.468701171875, 0.46863360595703124, 0.468738037109375, 0.46885989379882814, 0.4684267578125, 0.468389892578125, 0.46853839111328127, 0.46828131103515624, 0.4683601989746094, 0.46820352172851565, 0.467962890625, 0.4681553955078125, 0.4686499938964844, 0.4686612548828125, 0.4687503356933594, 0.46850253295898436, 0.4680570983886719, 0.4683458557128906, 0.46834994506835936, 0.46889471435546876, 0.46862130737304686, 0.4687196044921875, 0.46831512451171875, 0.4685035400390625, 0.4694077453613281, 0.4687421569824219, 0.4688087158203125, 0.4701829528808594, 0.4684830322265625, 0.4688619384765625, 0.4692244567871094, 0.46869094848632814, 0.4689715270996094, 0.46905548095703126, 0.4686315612792969, 0.468864013671875, 0.4689029235839844, 0.46895001220703125, 0.46892340087890627, 0.46898687744140627, 0.469212158203125, 0.46885272216796875, 0.4688230285644531, 0.46921624755859376, 0.46866943359375, 0.46926849365234374, 0.46906777954101564, 0.469248046875, 0.4691763000488281, 0.46952960205078126, 0.4694057006835938, 0.46893157958984377, 0.4688035888671875, 0.4690667419433594, 0.4687615966796875, 0.4690288696289063, 0.46915890502929686, 0.46848818969726563, 0.4687585144042969, 0.4686315612792969, 0.4686049194335937, 0.46867352294921877, 0.9793843383789063, 0.4697733154296875, 0.46955825805664064, 0.46915890502929686, 0.46897869873046877, 0.4683345947265625, 0.46859982299804687, 0.468316162109375, 0.46864694213867186, 0.4686110534667969, 0.46845745849609377, 0.46816357421875, 0.4684892272949219, 0.46949169921875, 0.46866842651367185, 0.46861312866210936, 0.4686090087890625, 0.4685424499511719, 0.46886605834960937, 0.4688752746582031, 0.46871353149414063, 0.4691127624511719, 0.46852615356445315, 0.46842361450195313, 0.46897665405273437, 0.4689141845703125, 0.4685926513671875, 0.46885989379882814, 0.468421630859375, 0.4683888549804687, 0.46856600952148436, 0.4689182739257812, 0.46858547973632814, 0.46870220947265623, 0.4691896362304688, 0.46975283813476565, 0.46981631469726565, 0.4697753601074219, 0.470096923828125, 0.469834716796875, 0.46959820556640625, 0.4698542175292969, 0.46952243041992187, 0.4697272338867188, 0.4697907104492188, 0.4687656860351562, 0.4691937255859375, 0.46966680908203123, 0.46944461059570314, 0.46914764404296877, 0.469760009765625, 0.4693544921875, 0.46949993896484377, 0.4696063537597656, 0.46877694702148437, 0.4688281555175781, 0.4692244567871094, 0.468706298828125, 0.4684646301269531, 0.46868582153320315, 0.4686725158691406, 0.46858547973632814, 0.4686612548828125, 0.9803509521484375, 0.46829159545898436, 0.4681697387695313, 0.4690083923339844, 0.46880972290039064, 0.46866329956054686, 0.4685547485351563, 0.46837042236328125, 0.468105224609375, 0.4683243713378906, 0.46842364501953127, 0.4681779174804688, 0.4683816833496094, 0.4685599365234375, 0.4682454528808594, 0.46819635009765626, 0.46841036987304685, 0.46834994506835936, 0.468537353515625, 0.46845849609375, 0.46887115478515623, 0.46867352294921877, 0.46852197265625, 0.46868377685546875, 0.46900018310546876, 0.46878207397460936, 0.4686499938964844, 0.4685537414550781, 0.46879437255859374, 0.4684031982421875, 0.4687779846191406, 0.4690032653808594, 0.4684912719726563, 0.46856298828125, 0.46899917602539065, 0.4696033020019531, 0.4696708984375, 0.469855224609375, 0.4690411376953125, 0.468674560546875, 0.46875442504882814, 0.469359619140625, 0.4701552734375, 0.469064697265625, 0.46861312866210936, 0.4686315612792969, 0.4697047119140625, 0.4690083923339844, 0.46888754272460936, 0.46868582153320315, 0.469073974609375, 0.4693411254882813, 0.4695132141113281, 0.46920089721679686, 0.46941696166992186, 0.46919271850585936, 0.46985626220703125, 0.46913946533203127, 0.46915789794921875, 0.4686253967285156, 0.46875955200195313, 0.468642822265625, 0.4691517333984375]",tokens/s,2.1000993200656053,,,,, -4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-1.3B,,cuda,0,42,,,,,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,main,False,False,MB,1562.181632,1957.167104,0.0,1310.72,1163.955712,s,10,1.3245970916748049,0.13245970916748048,0.001266998397567483,0.13200260162353517,0.1341699493408203,0.13443305816650392,0.1346435452270508,"[0.1346961669921875, 0.1341114807128906, 0.1312022705078125, 0.13115029907226564, 0.13112367248535156, 0.1315875244140625, 0.13163551330566406, 0.13236968994140624, 0.13365353393554688, 0.13306694030761718]",tokens/s,1932.6631593031557,kWh,1.5487117942793545e-06,8.486256414471426e-07,6.508118482104666e-06,8.905455917831163e-06,tokens/kWh,28746422.683134936,MB,1562.181632,1959.264256,0.0,1312.817152,1232.77568,s,10,76.92137255859376,7.692137255859374,0.017487914910414803,7.693824951171875,7.702772509765625,7.711682641601563,7.718810747070313,"[7.69504345703125, 7.69082470703125, 7.6926064453125, 7.6866689453125, 7.7205927734375, 7.69639892578125, 7.6467568359375, 7.70079248046875, 7.69208056640625, 7.699607421875]",tokens/s,8.190181467707257,kWh,9.084797721340027e-05,4.979003748630931e-05,0.0003722497123017,0.0005128877270014096,tokens/kWh,122833.9004489902,,s,629,77.98684770965575,0.12398544945891218,0.01575529620495015,0.121744384765625,0.12330782623291016,0.12368689422607422,0.2534359008789062,"[0.12507750701904297, 0.12424604797363281, 0.1234441909790039, 0.12341862487792969, 0.12180786895751954, 0.1224120330810547, 0.12155903625488282, 0.12088832092285157, 0.12079001617431641, 0.12080332946777343, 0.1209518051147461, 0.12134912109375, 0.12099378967285156, 0.12220416259765625, 0.12147408294677735, 0.12154876708984375, 0.12117708587646485, 0.12111360168457032, 0.12086067199707032, 0.12105318450927735, 0.12099174499511718, 0.12216934204101562, 0.12297727966308594, 0.12166143798828125, 0.12153446197509765, 0.12155801391601563, 0.12161023712158203, 0.12131839752197265, 0.12115455627441406, 0.12147609710693359, 0.12320972442626953, 0.12328857421875, 0.12294041442871094, 0.1213460464477539, 0.12115360260009765, 0.12151289367675781, 0.12153343963623046, 0.12244172668457032, 0.12324454498291015, 0.12321791839599609, 0.1221396484375, 0.123504638671875, 0.12190930938720704, 0.12093536376953125, 0.12353228759765625, 0.1231247329711914, 0.12405248260498047, 0.12286566162109375, 0.12216831970214843, 0.1221580810546875, 0.12220928192138672, 0.12256153869628907, 0.12320460510253907, 0.12468940734863282, 0.12308889770507812, 0.1220997085571289, 0.12128972625732422, 0.12146380615234376, 0.12163587188720704, 0.12112380981445313, 0.12155289459228516, 0.12149759674072266, 0.2558975982666016, 0.12277452850341797, 0.12251545715332031, 0.12256665802001954, 0.12305101013183593, 0.12229837036132812, 0.1212968978881836, 0.1239582748413086, 0.12347596740722656, 0.12176076507568359, 0.12155289459228516, 0.12141977691650391, 0.1213685760498047, 0.1208463363647461, 0.12073062133789063, 0.1210040283203125, 0.12351385498046875, 0.12159487915039062, 0.12144640350341797, 0.12170649719238281, 0.12249292755126953, 0.12232396697998046, 0.12123442840576172, 0.12141567993164062, 0.12308889770507812, 0.12184268951416016, 0.1215098876953125, 0.12254208374023437, 0.12142387390136719, 0.12173619079589844, 0.12151910400390625, 0.12153446197509765, 0.12306739044189453, 0.12236492919921875, 0.12201881408691406, 0.12173824310302735, 0.12155699157714844, 0.12209356689453126, 0.12139315032958985, 0.12139520263671875, 0.12346572875976562, 0.12301107025146485, 0.12293836975097656, 0.12192870330810547, 0.12156313323974609, 0.12141056060791015, 0.12382415771484374, 0.12120060729980468, 0.12172902679443359, 0.12317593383789062, 0.12274995422363282, 0.12246323394775391, 0.12149657440185548, 0.12139110565185547, 0.12291788482666016, 0.12123340606689453, 0.12196150207519531, 0.12387631988525391, 0.12276838684082031, 0.12148223876953125, 0.12265267181396484, 0.1214044189453125, 0.12125389099121094, 0.25443635559082034, 0.12313497924804688, 0.12153446197509765, 0.12169830322265625, 0.12147097778320312, 0.12127334594726563, 0.12249088287353516, 0.1221396484375, 0.12196044921875, 0.12192460632324219, 0.12144742584228516, 0.1228226547241211, 0.12275917053222657, 0.12112697601318359, 0.12176787567138672, 0.1217228775024414, 0.12139424133300782, 0.12269356536865235, 0.1223720932006836, 0.12153343963623046, 0.1215478057861328, 0.12162351989746094, 0.12261682891845703, 0.12271308898925781, 0.1215098876953125, 0.12233932495117188, 0.12170342254638672, 0.121238525390625, 0.12113817596435547, 0.12111769866943359, 0.12136653137207032, 0.12165529632568359, 0.1218897933959961, 0.121744384765625, 0.12428594970703125, 0.12338790130615235, 0.12270489501953125, 0.1215129623413086, 0.12190310668945313, 0.12194918060302734, 0.12152114868164063, 0.12212838745117187, 0.12310630035400391, 0.12293427276611328, 0.12179046630859375, 0.12117298889160157, 0.12185906982421875, 0.12301107025146485, 0.1216522216796875, 0.120953857421875, 0.12247654724121093, 0.12268236541748047, 0.12308684539794922, 0.12277657318115234, 0.12169420623779297, 0.12158668518066407, 0.12113510131835938, 0.1211361312866211, 0.1227540512084961, 0.12328550720214844, 0.12320972442626953, 0.12324147033691406, 0.12298242950439453, 0.25302217102050784, 0.1215467529296875, 0.12238642883300781, 0.12331622314453125, 0.12284416198730469, 0.12259225463867188, 0.1225902099609375, 0.12293222045898437, 0.12212531280517579, 0.12088832092285157, 0.12144127655029296, 0.12321894073486328, 0.12279398345947265, 0.12161740875244141, 0.12199628448486328, 0.12130611419677734, 0.12155699157714844, 0.12218367767333985, 0.12152627563476562, 0.12282061004638672, 0.12332236480712891, 0.12172697448730468, 0.1214730224609375, 0.12131743621826171, 0.12206380462646485, 0.12272434997558594, 0.12205158233642578, 0.12277043151855468, 0.12318822479248047, 0.12324352264404297, 0.12150886535644531, 0.12119551849365234, 0.12129280090332031, 0.12111567687988281, 0.1214658203125, 0.12191129302978515, 0.12111974334716796, 0.1216358413696289, 0.12116377258300781, 0.12117708587646485, 0.12161539459228515, 0.1231072998046875, 0.12339097595214844, 0.12407295989990234, 0.12407295989990234, 0.12135116577148437, 0.12155903625488282, 0.12125183868408203, 0.12111154937744141, 0.1215272979736328, 0.12139315032958985, 0.12157440185546875, 0.12285030364990235, 0.12157644653320313, 0.1217976303100586, 0.12137881469726562, 0.12175360107421875, 0.12123750305175782, 0.12148838043212891, 0.12119859313964844, 0.12210585784912109, 0.12211199951171875, 0.12212940979003906, 0.2530303955078125, 0.12125593566894531, 0.12254003143310546, 0.1212252197265625, 0.12286156463623046, 0.12248985290527344, 0.12266598510742187, 0.12236185455322265, 0.12159385681152343, 0.1212590103149414, 0.12165631866455077, 0.12115660858154297, 0.12235878753662109, 0.12386713409423829, 0.1217802276611328, 0.12304691314697265, 0.12239974212646484, 0.12245503997802734, 0.12413849639892578, 0.12299366760253906, 0.12299263763427734, 0.12301107025146485, 0.12142489624023438, 0.12230963134765625, 0.12332236480712891, 0.12175052642822265, 0.12143718719482421, 0.12149247741699219, 0.1212938232421875, 0.12226150512695312, 0.12151193237304687, 0.12162457275390624, 0.12326092529296875, 0.12302438354492187, 0.12452352142333985, 0.12339405059814453, 0.12212735748291016, 0.12276131439208984, 0.12299766540527343, 0.1216911392211914, 0.12096409606933593, 0.12123033905029297, 0.12140953826904297, 0.12123238372802735, 0.12380467224121093, 0.12290969848632813, 0.12300806427001953, 0.12272940826416015, 0.12320665740966796, 0.12299878692626953, 0.12275917053222657, 0.12293119812011719, 0.12306432342529297, 0.12317798614501953, 0.12322406768798828, 0.12217958068847656, 0.1228267822265625, 0.12593353271484375, 0.1234872283935547, 0.12296601867675781, 0.12340940856933594, 0.12347187042236328, 0.12346470642089843, 0.25439436340332033, 0.12301414489746093, 0.1228062744140625, 0.1231431655883789, 0.12312268829345703, 0.12464742279052735, 0.12372991943359375, 0.1234708480834961, 0.12335411071777344, 0.12233523559570313, 0.12217958068847656, 0.12264959716796875, 0.12329881286621094, 0.12281549072265625, 0.12341248321533203, 0.123109375, 0.12312268829345703, 0.12331110382080078, 0.12178943634033203, 0.12153343963623046, 0.12128153228759765, 0.12127027130126954, 0.12139008331298828, 0.12159487915039062, 0.12141875457763672, 0.12109516906738281, 0.12095590209960938, 0.12107469177246094, 0.12224205017089844, 0.1224263687133789, 0.12177005004882813, 0.12150163269042968, 0.12146585845947265, 0.12140953826904297, 0.12113005065917969, 0.12179347229003906, 0.12241817474365234, 0.12146585845947265, 0.1236316146850586, 0.12164096069335938, 0.12152934265136718, 0.12139622497558594, 0.12146688079833984, 0.12121600341796875, 0.12141567993164062, 0.12133478546142579, 0.12168498992919922, 0.1214167709350586, 0.12150675201416015, 0.12333977508544922, 0.12334899139404297, 0.12139826965332032, 0.12266291046142579, 0.12162457275390624, 0.1213675537109375, 0.12148838043212891, 0.123219970703125, 0.12320358276367188, 0.12155494689941407, 0.12110848236083985, 0.12138294219970704, 0.12153238677978516, 0.12358963012695312, 0.25359359741210935, 0.12145458984375, 0.12112588500976562, 0.12119039916992187, 0.12124671936035156, 0.12149964904785156, 0.12119961547851563, 0.12119245147705078, 0.12146688079833984, 0.12120269012451172, 0.12097740936279297, 0.1208616943359375, 0.12124671936035156, 0.12146176147460938, 0.12102349090576171, 0.1214505615234375, 0.12145657348632813, 0.12126924896240235, 0.12115869140625, 0.12115657806396485, 0.12137574768066406, 0.12156928253173828, 0.12129894256591797, 0.1242234878540039, 0.12195740509033202, 0.12128765106201173, 0.12127436828613282, 0.12119142150878906, 0.12085968017578125, 0.12119548797607421, 0.12113203430175781, 0.12153139495849609, 0.12129587554931641, 0.12157234954833984, 0.12127641296386718, 0.12110438537597656, 0.12128870391845703, 0.12145970916748047, 0.12127846527099609, 0.12160921478271484, 0.12154879760742188, 0.1213306884765625, 0.12123238372802735, 0.12116070556640625, 0.12110643005371094, 0.12110031890869141, 0.12115042877197266, 0.12125593566894531, 0.12130406188964844, 0.12108185577392579, 0.12092108917236329, 0.12128562927246093, 0.12140748596191406, 0.12148838043212891, 0.12150784301757812, 0.12131737518310547, 0.12162969970703125, 0.12153958129882812, 0.122281982421875, 0.12196249389648438, 0.12154879760742188, 0.1211883544921875, 0.12114022064208985, 0.25422848510742185, 0.12144435119628906, 0.12121600341796875, 0.12331622314453125, 0.12173926544189453, 0.12114329528808594, 0.1231800308227539, 0.1233438720703125, 0.1232721939086914, 0.12304691314697265, 0.12367155456542969, 0.12309913635253907, 0.12304998779296875, 0.12310630035400391, 0.12295680236816406, 0.12294348907470704, 0.12386918640136718, 0.12152320098876954, 0.12160205078125, 0.121059326171875, 0.12094258880615234, 0.12096717071533203, 0.12145868682861329, 0.12323331451416016, 0.12289020538330078, 0.12314832305908203, 0.1236971206665039, 0.12315955352783203, 0.12311039733886718, 0.12137471771240234, 0.12235366058349609, 0.1222451171875, 0.1222973403930664, 0.12167884826660157, 0.12279296112060546, 0.12148735809326172, 0.12146995544433593, 0.12277760314941406, 0.12156723022460937, 0.1216358413696289, 0.12169728088378906, 0.12312166595458984, 0.12145664215087891, 0.12275507354736329, 0.1232353286743164, 0.12311347198486328, 0.123072509765625, 0.1227171859741211, 0.12111154937744141, 0.12276019287109376, 0.12255232238769531, 0.12235059356689452, 0.12264141082763672, 0.1211514892578125, 0.12123545837402344, 0.12166246032714843, 0.12119347381591797, 0.12130201721191407, 0.12149964904785156, 0.12166246032714843, 0.12141875457763672, 0.12106034851074218, 0.12132147216796875, 0.25745306396484374, 0.12312064361572266, 0.12307865905761718, 0.12312166595458984, 0.12121913909912109, 0.12116063690185547, 0.1212774429321289, 0.12236185455322265, 0.1215498275756836, 0.12132454681396485, 0.12093132781982421, 0.12260454559326171, 0.12201676940917969, 0.12116480255126953, 0.12119449615478516, 0.121206787109375, 0.12122112274169922, 0.12207103729248046, 0.12143001556396485, 0.12134400177001953, 0.12148531341552735, 0.12378009796142578, 0.12321279907226562, 0.12303257751464844, 0.12391117095947266, 0.12308889770507812, 0.12307762908935548, 0.12230758666992188, 0.12212636566162109, 0.12145769500732422, 0.12122207641601562, 0.12131123352050781, 0.12108799743652343, 0.1211156463623047, 0.12251033782958984, 0.12306432342529297, 0.12318310546875, 0.12159795379638672, 0.12140850830078125, 0.12139520263671875, 0.12131839752197265, 0.12117196655273438, 0.12120371246337891, 0.12156723022460937, 0.12148941040039063, 0.12145664215087891, 0.12146892547607421, 0.1215447006225586, 0.12250316619873047, 0.12295782470703125, 0.12217036437988281, 0.12270387268066406, 0.12305919647216797, 0.12331520080566406, 0.12305203247070312, 0.12250521850585938, 0.12303667449951172, 0.12276838684082031, 0.12293427276611328, 0.12293427276611328, 0.12159283447265624, 0.12137062072753907, 0.12141670227050781, 0.2548654022216797, 0.1212416000366211, 0.12151193237304687, 0.12308480072021484, 0.12140338897705077, 0.12150377655029297, 0.12124156951904297, 0.121275390625, 0.12150374603271484, 0.12148838043212891, 0.12146482849121094, 0.12227788543701172, 0.12289740753173828, 0.12292819213867187, 0.12136339569091797, 0.1211822052001953, 0.12195027160644531, 0.12174432373046876, 0.12153241729736328, 0.12099174499511718, 0.12128463745117188, 0.12138390350341798, 0.121385986328125, 0.12258918762207031, 0.1230161895751953, 0.12225638580322265, 0.12275917053222657, 0.12151602935791016, 0.12140850830078125, 0.12134809875488281, 0.12176076507568359, 0.1217791976928711, 0.12272946929931641, 0.12169420623779297, 0.12219187164306641, 0.12145561981201172, 0.12164096069335938, 0.12145254516601563, 0.12175360107421875, 0.12436787414550782, 0.1233602523803711, 0.12332339477539063, 0.1232701416015625, 0.12218675231933594, 0.1233039321899414, 0.12280217742919922, 0.12283596801757812, 0.1221048355102539, 0.12304793548583984, 0.1227509765625, 0.12293023681640625, 0.12202182769775391, 0.12274483489990234, 0.1233070068359375, 0.12290560150146485, 0.12302130889892578, 0.1229496307373047, 0.12324864196777344, 0.12319129943847656, 0.12199935913085938, 0.12175363159179688, 0.12325577545166015, 0.12313600158691407]",tokens/s,8.06546255519598,,,,, -4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neox-20b,,cuda,0,42,,,,,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,main,False,False,MB,1888.280576,15194.390528,0.0,14547.943424,13898.252288,s,10,16.972713500976564,1.6972713500976564,0.0011802396473078822,1.6969771728515626,1.6988580200195311,1.699245782470703,1.6995559924316406,"[1.6969825439453126, 1.6969718017578126, 1.6960919189453125, 1.6958106689453125, 1.6964619140625, 1.69635986328125, 1.6974122314453124, 1.6982171630859375, 1.699633544921875, 1.6987718505859375]",tokens/s,150.83033127570937,kWh,2.0026979611979592e-05,1.097496669508473e-05,9.603446571639917e-05,0.0001270364120234635,tokens/kWh,2015170.2643547354,MB,1888.280576,15194.390528,0.0,14547.943424,14315.97312,s,10,986.8612499999999,98.68612499999999,0.014850841233799409,98.68501953125,98.70599609374999,98.707103515625,98.707989453125,"[98.6797109375, 98.687671875, 98.6747890625, 98.6555546875, 98.6919140625, 98.6783203125, 98.6969609375, 98.7082109375, 98.70575, 98.6823671875]",tokens/s,0.6383876152802637,kWh,0.0011649224273860455,0.0006384806384544572,0.00558826860950019,0.007391671675340693,tokens/kWh,8523.105836826315,,s,629,1000.5712542724599,1.5907333136287136,0.2015788363996436,1.5663739013671876,1.5674868164062499,1.5676927978515625,3.262457841796875,"[1.5671644287109374, 1.5661424560546875, 1.56664013671875, 1.5672913818359375, 1.566525390625, 1.5666636962890625, 1.56687158203125, 1.566798828125, 1.565517822265625, 1.56556494140625, 1.56592333984375, 1.56533251953125, 1.5655465087890625, 1.5658004150390625, 1.565998046875, 1.5670538330078125, 1.5664508056640625, 1.56584033203125, 1.5667855224609375, 1.5674542236328124, 1.567055908203125, 1.5672022705078126, 1.5673907470703126, 1.56657666015625, 1.5663001708984374, 1.566665771484375, 1.5659376220703125, 1.5659520263671876, 1.5661055908203125, 1.566899169921875, 1.5660892333984375, 1.566604248046875, 1.5656611328125, 1.5667138671875, 1.565654052734375, 1.566857177734375, 1.5661802978515624, 1.5664854736328124, 1.5667210693359375, 1.5670538330078125, 1.56666162109375, 1.566857177734375, 1.5659581298828125, 1.565955078125, 1.5662294921875, 1.5661065673828125, 1.566191650390625, 1.5659642333984376, 1.5664527587890624, 1.5659898681640625, 1.5656510009765625, 1.5664803466796875, 1.566614501953125, 1.566317626953125, 1.5665059814453124, 1.565919189453125, 1.5657093505859374, 1.5657308349609376, 1.5665244140625, 1.565981689453125, 1.5660799560546874, 1.56632373046875, 3.263318115234375, 1.5650416259765625, 1.5667425537109374, 1.5671868896484376, 1.5674766845703125, 1.566899169921875, 1.5670477294921874, 1.567247314453125, 1.5669462890625, 1.566482421875, 1.56721044921875, 1.567352783203125, 1.56718603515625, 1.5668541259765625, 1.565632568359375, 1.5655035400390624, 1.565454345703125, 1.566171142578125, 1.5655423583984376, 1.5656632080078126, 1.565739013671875, 1.5655546875, 1.56594384765625, 1.565739990234375, 1.5656090087890624, 1.5654093017578126, 1.5656141357421876, 1.5657379150390625, 1.565550537109375, 1.5656468505859376, 1.566182373046875, 1.56573388671875, 1.5660994873046874, 1.566614501953125, 1.5671705322265626, 1.5668858642578125, 1.5673907470703126, 1.5655372314453124, 1.5656827392578125, 1.567224853515625, 1.5674173583984374, 1.567141845703125, 1.567562744140625, 1.56704150390625, 1.566593994140625, 1.5660902099609375, 1.56824267578125, 1.56689404296875, 1.5673763427734375, 1.5663564453125, 1.566992431640625, 1.567836181640625, 1.5679334716796876, 1.5667803955078126, 1.5659478759765626, 1.566908447265625, 1.566277587890625, 1.5658690185546875, 1.5664803466796875, 1.5664271240234375, 1.5662049560546876, 1.5662171630859374, 1.5660595703125, 3.26247021484375, 1.5656990966796875, 1.5657799072265626, 1.566017578125, 1.5667598876953126, 1.565744140625, 1.566393310546875, 1.56580859375, 1.56586083984375, 1.5654072265625, 1.56617529296875, 1.5659202880859375, 1.5659725341796875, 1.5665029296875, 1.5664486083984375, 1.5655260009765626, 1.5660780029296875, 1.5663062744140626, 1.5667117919921876, 1.566983154296875, 1.5657728271484375, 1.565895751953125, 1.566373779296875, 1.5677890625, 1.566457763671875, 1.5663277587890625, 1.566962646484375, 1.566697509765625, 1.5665489501953125, 1.566899169921875, 1.56623046875, 1.566688232421875, 1.5666124267578125, 1.5657093505859374, 1.566341064453125, 1.5660462646484374, 1.5663482666015625, 1.566123046875, 1.5662838134765624, 1.5664619140625, 1.56621826171875, 1.566017578125, 1.5659622802734374, 1.565760498046875, 1.5661669921875, 1.566396484375, 1.5663912353515625, 1.5682620849609374, 1.5664476318359375, 1.5661475830078124, 1.56608203125, 1.56626123046875, 1.5659765625, 1.5659632568359374, 1.565811767578125, 1.56657763671875, 1.5658956298828124, 1.5657799072265626, 1.5660042724609375, 1.566275634765625, 1.5663533935546874, 1.56682958984375, 1.566255126953125, 3.26236865234375, 1.5657420654296874, 1.5653990478515625, 1.5662427978515625, 1.565697021484375, 1.5651962890625, 1.56592333984375, 1.5652525634765626, 1.5654676513671875, 1.5651666259765624, 1.56598583984375, 1.5653150634765625, 1.5663011474609374, 1.5655894775390624, 1.5655516357421875, 1.565338623046875, 1.5655577392578126, 1.5656806640625, 1.56554345703125, 1.56598779296875, 1.5655843505859375, 1.5654718017578124, 1.5656263427734376, 1.5660400390625, 1.5658824462890626, 1.5657850341796875, 1.5657625732421876, 1.5657196044921875, 1.56554443359375, 1.566350341796875, 1.5656375732421874, 1.5655997314453125, 1.566192626953125, 1.565496337890625, 1.5660390625, 1.566271484375, 1.5658936767578124, 1.565685791015625, 1.5660216064453125, 1.5656141357421876, 1.565828125, 1.566434326171875, 1.5664906005859376, 1.5654307861328125, 1.56860107421875, 1.56617529296875, 1.566096435546875, 1.5658311767578126, 1.5665806884765625, 1.5662591552734375, 1.5666104736328126, 1.56617724609375, 1.566076904296875, 1.5665244140625, 1.566587890625, 1.56634521484375, 1.5658741455078125, 1.5661434326171875, 1.566371826171875, 1.5663533935546874, 1.566328857421875, 1.5666011962890625, 1.5661240234375, 3.263153076171875, 1.56556494140625, 1.565706298828125, 1.5666165771484375, 1.566034912109375, 1.566002197265625, 1.5660831298828124, 1.565319091796875, 1.5665797119140625, 1.5655751953125, 1.5664332275390624, 1.5658896484375, 1.5659744873046875, 1.5658916015625, 1.566130126953125, 1.5668695068359375, 1.5666226806640624, 1.5660933837890625, 1.566509033203125, 1.5661363525390626, 1.5657738037109374, 1.5658199462890625, 1.5663206787109376, 1.5657840576171875, 1.566213134765625, 1.5667742919921874, 1.5661240234375, 1.5659315185546876, 1.565498291015625, 1.5660472412109374, 1.5656407470703124, 1.56573486328125, 1.566254150390625, 1.56568359375, 1.5663575439453126, 1.566123046875, 1.5667579345703124, 1.5663963623046875, 1.5667947998046876, 1.5661865234375, 1.5662919921875, 1.5661158447265624, 1.5670897216796875, 1.56710302734375, 1.567573974609375, 1.5670128173828124, 1.5670006103515626, 1.5675535888671874, 1.568489501953125, 1.5677030029296875, 1.5673804931640625, 1.5675074462890626, 1.5675924072265626, 1.56819970703125, 1.56786181640625, 1.5679241943359374, 1.5680113525390624, 1.5678955078125, 1.56760888671875, 1.5664681396484375, 1.5662049560546876, 1.5662213134765626, 1.5661905517578125, 3.262426025390625, 1.5655628662109375, 1.56493212890625, 1.5661669921875, 1.5657431640625, 1.5655537109375, 1.5660482177734374, 1.5657359619140625, 1.5666063232421874, 1.5661004638671876, 1.5656785888671876, 1.5670230712890625, 1.5668284912109376, 1.5666944580078126, 1.566755859375, 1.566581787109375, 1.5667763671875, 1.5658946533203124, 1.5655372314453124, 1.5659530029296875, 1.565708251953125, 1.5669320068359376, 1.56744091796875, 1.566341064453125, 1.5660360107421876, 1.5659837646484376, 1.5658629150390626, 1.5660892333984375, 1.565887451171875, 1.56593359375, 1.5656898193359374, 1.5661455078125, 1.566086181640625, 1.5658076171875, 1.5659263916015624, 1.5658218994140625, 1.56609033203125, 1.56613525390625, 1.5673641357421875, 1.5665848388671875, 1.5666175537109375, 1.5664681396484375, 1.5664619140625, 1.566224365234375, 1.5664588623046876, 1.5661363525390626, 1.5667579345703124, 1.5667978515625, 1.5667178955078125, 1.5658916015625, 1.5658680419921875, 1.5657728271484375, 1.5670528564453126, 1.567161376953125, 1.5674490966796875, 1.567394775390625, 1.567477783203125, 1.56653369140625, 1.5658792724609376, 1.56665966796875, 1.5675699462890624, 1.56628076171875, 1.5665899658203124, 3.26359765625, 1.5656744384765624, 1.565328369140625, 1.566392333984375, 1.565917236328125, 1.566066650390625, 1.567247314453125, 1.56674560546875, 1.567057861328125, 1.566623779296875, 1.5667916259765624, 1.5669647216796876, 1.56680810546875, 1.5664117431640625, 1.566813232421875, 1.5673231201171876, 1.566376953125, 1.5661844482421876, 1.567056884765625, 1.567025146484375, 1.566773193359375, 1.5668214111328125, 1.5673548583984376, 1.568668701171875, 1.567552490234375, 1.566614501953125, 1.566899169921875, 1.567447021484375, 1.5675135498046875, 1.5669381103515625, 1.5667794189453126, 1.567139892578125, 1.5667916259765624, 1.566656494140625, 1.566287841796875, 1.5665806884765625, 1.5664046630859374, 1.5664854736328124, 1.56649169921875, 1.56577587890625, 1.5662652587890624, 1.5666688232421875, 1.5665521240234375, 1.56607080078125, 1.5668214111328125, 1.5663809814453125, 1.5665684814453125, 1.5666226806640624, 1.566630859375, 1.5663360595703124, 1.566734375, 1.5659100341796874, 1.566339111328125, 1.5664271240234375, 1.5667056884765624, 1.5661854248046876, 1.5660308837890624, 1.56668310546875, 1.5665264892578126, 1.5664742431640626, 1.5667752685546874, 1.5658568115234375, 1.566477294921875, 3.26474853515625, 1.5660062255859375, 1.5659560546875, 1.5659263916015624, 1.5657471923828126, 1.5661905517578125, 1.5662623291015625, 1.5657779541015624, 1.5659345703125, 1.5662991943359375, 1.5659427490234374, 1.565487060546875, 1.566613525390625, 1.56559765625, 1.56596533203125, 1.566318603515625, 1.566983154296875, 1.5657728271484375, 1.5661322021484374, 1.566572509765625, 1.5666114501953126, 1.5659100341796874, 1.56704150390625, 1.56611376953125, 1.5677716064453124, 1.5673876953125, 1.5676068115234374, 1.5671654052734374, 1.5670262451171875, 1.5657738037109374, 1.5667547607421874, 1.5669217529296875, 1.5676558837890624, 1.56761083984375, 1.567972412109375, 1.567614990234375, 1.5675115966796875, 1.5676702880859374, 1.56752587890625, 1.567363037109375, 1.567826904296875, 1.567635498046875, 1.5676488037109375, 1.5662724609375, 1.5680296630859376, 1.567309814453125, 1.567677490234375, 1.567458251953125, 1.5668489990234375, 1.565865966796875, 1.566993408203125, 1.5673487548828124, 1.567581298828125, 1.566005126953125, 1.5679825439453126, 1.567009765625, 1.566161865234375, 1.5676138916015625, 1.56758935546875, 1.5667568359375, 1.5663258056640625, 1.566224365234375, 1.5670302734375, 3.26736279296875, 1.5667579345703124, 1.5664261474609376, 1.5665531005859374, 1.5668193359375, 1.566329833984375, 1.5663739013671876, 1.566214111328125, 1.5671173095703126, 1.565612060546875, 1.5653099365234375, 1.565854736328125, 1.5658486328125, 1.5660902099609375, 1.566256103515625, 1.5667547607421874, 1.5674736328125, 1.56900244140625, 1.5678065185546874, 1.567677490234375, 1.5673057861328126, 1.5676702880859374, 1.5678648681640626, 1.566625732421875, 1.566234619140625, 1.5665244140625, 1.565811767578125, 1.566467041015625, 1.5663380126953126, 1.5663995361328125, 1.56670458984375, 1.566562255859375, 1.56580859375, 1.5658076171875, 1.567499267578125, 1.566665771484375, 1.5675023193359374, 1.5676190185546874, 1.5666585693359374, 1.5654676513671875, 1.5674849853515624, 1.5671900634765625, 1.567220703125, 1.5676558837890624, 1.5666514892578125, 1.5663369140625, 1.5666441650390626, 1.566841796875, 1.566866455078125, 1.56685107421875, 1.567130615234375, 1.5664896240234376, 1.5665152587890625, 1.5667691650390625, 1.56735693359375, 1.5668797607421876, 1.5669544677734375, 1.565843505859375, 1.5657564697265625, 1.5662509765625, 1.5679661865234376, 1.567494140625, 1.5670753173828125, 3.26488671875, 1.56588232421875, 1.5666165771484375, 1.56659912109375, 1.5660400390625, 1.565919189453125, 1.5653775634765625, 1.5651685791015626, 1.5665255126953126, 1.5655526123046875, 1.56554443359375, 1.565875244140625, 1.565812744140625, 1.56569091796875, 1.56594384765625, 1.566340087890625, 1.565750244140625, 1.5661793212890625, 1.5670374755859375, 1.5664691162109374, 1.5666708984375, 1.5669124755859376, 1.5670743408203125, 1.565885498046875, 1.567392822265625, 1.56645068359375, 1.567561767578125, 1.5667598876953126, 1.566843994140625, 1.566246826171875, 1.56635546875, 1.5664219970703126, 1.565961181640625, 1.5663524169921874, 1.5675914306640626, 1.56609326171875, 1.5661405029296875, 1.5667711181640624, 1.566587890625, 1.5662694091796876, 1.566982177734375, 1.566482421875, 1.566496826171875, 1.5665531005859374, 1.5663370361328126, 1.5661158447265624, 1.566561279296875, 1.566202880859375, 1.566286865234375, 1.5659674072265626, 1.5661629638671875, 1.5660155029296876, 1.566634033203125, 1.5668284912109376, 1.5664476318359375, 1.5665438232421875, 1.5667864990234375, 1.5668736572265625, 1.56661865234375, 1.5671204833984376, 1.566720947265625, 1.5667332763671875, 1.566159912109375]",tokens/s,0.6286408862079101,,,,, -4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-34B,01-ai/Yi-34B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,,,,,MB,1767.067648,22129.672192,0.0,21483.225088,20799.168,s,10,28.05056616210938,2.805056616210938,0.0027095714336187707,2.804907958984375,2.8082408447265625,2.808313903808594,2.808372351074219,"[2.805282470703125, 2.808224609375, 2.803196533203125, 2.80234619140625, 2.80277294921875, 2.800419677734375, 2.804533447265625, 2.80722705078125, 2.80817626953125, 2.808386962890625]",tokens/s,91.26375507735885,kWh,3.3081144508388305e-05,1.8129723654910774e-05,0.00015725384802519894,0.000208464716188498,tokens/kWh,1228025.5607789264,MB,1772.261376,22129.672192,0.0,21483.225088,20902.144,s,10,1667.162484375,166.71624843749996,0.015085835868376713,166.71732031250002,166.73279531249997,166.73349140624998,166.73404828124998,"[166.730125, 166.715734375, 166.68440625, 166.724703125, 166.7341875, 166.71721875, 166.708265625, 166.717421875, 166.69778125, 166.732640625]",tokens/s,0.377887581987055,kWh,0.0019678901955319777,0.0010785770549249355,0.009270038554913802,0.012316505805370715,tokens/kWh,5115.087103074991,,s,629,1689.7617126464822,2.686425616290118,0.33327430639158717,2.64620947265625,2.6474256835937497,2.647894970703125,5.4507228125000005,"[2.646453369140625, 2.646373291015625, 2.64631591796875, 2.6454580078125, 2.646531982421875, 2.647617431640625, 2.64826171875, 2.647658447265625, 2.647793701171875, 2.64637646484375, 2.64711376953125, 2.64620947265625, 2.646602783203125, 2.646426513671875, 2.64635498046875, 2.64557568359375, 2.644958251953125, 2.647402587890625, 2.64601904296875, 2.646287353515625, 2.647202880859375, 2.64688525390625, 2.646246337890625, 2.646867919921875, 2.64715478515625, 2.647743408203125, 2.646212646484375, 2.647287841796875, 2.648404052734375, 2.64638671875, 2.64559912109375, 2.647835693359375, 2.648161376953125, 2.6474189453125, 2.647456787109375, 2.6468076171875, 2.646687744140625, 2.64502880859375, 2.646285400390625, 2.646274169921875, 2.64652490234375, 2.64616455078125, 2.64710546875, 2.6460908203125, 2.646740966796875, 2.645139404296875, 2.646036376953125, 2.646411376953125, 2.64627099609375, 2.64452197265625, 2.64525830078125, 2.647901123046875, 2.64601806640625, 2.645980224609375, 2.64719677734375, 2.64639990234375, 2.64538720703125, 2.645042236328125, 2.64540576171875, 2.646120361328125, 2.646118408203125, 2.64612353515625, 5.45506494140625, 2.64526025390625, 2.6462626953125, 2.64690576171875, 2.646042724609375, 2.64546826171875, 2.646393798828125, 2.647192626953125, 2.644919189453125, 2.644935791015625, 2.645579833984375, 2.647086181640625, 2.646035400390625, 2.644955078125, 2.645644287109375, 2.645537841796875, 2.644991943359375, 2.645560302734375, 2.646414306640625, 2.645199951171875, 2.646506591796875, 2.646548583984375, 2.64702880859375, 2.64570166015625, 2.644788330078125, 2.64700732421875, 2.646330322265625, 2.64543017578125, 2.645243896484375, 2.646043701171875, 2.645831787109375, 2.6459453125, 2.64614697265625, 2.64631494140625, 2.64669091796875, 2.646116455078125, 2.647741455078125, 2.6478857421875, 2.645671875, 2.647362548828125, 2.646822998046875, 2.648393798828125, 2.647374755859375, 2.64627294921875, 2.646519775390625, 2.645937255859375, 2.6523720703125, 2.647458740234375, 2.646916015625, 2.6467646484375, 2.644612060546875, 2.645245849609375, 2.64698779296875, 2.646232177734375, 2.645567626953125, 2.645909423828125, 2.64679541015625, 2.64620947265625, 2.645465087890625, 2.646067138671875, 2.6467060546875, 2.6458798828125, 2.6457109375, 5.4513837890625, 2.647235595703125, 2.645803955078125, 2.64654443359375, 2.64496533203125, 2.64464892578125, 2.645158935546875, 2.646266845703125, 2.64560546875, 2.646096923828125, 2.64475537109375, 2.64629052734375, 2.644760498046875, 2.64480859375, 2.6463427734375, 2.64760009765625, 2.645761962890625, 2.645760986328125, 2.646381591796875, 2.646036376953125, 2.64513134765625, 2.646679443359375, 2.6458369140625, 2.6495498046875, 2.646917236328125, 2.644770751953125, 2.645843017578125, 2.644592529296875, 2.6452919921875, 2.6457333984375, 2.64617578125, 2.64542822265625, 2.645474365234375, 2.644977783203125, 2.6475703125, 2.645327880859375, 2.64601611328125, 2.64604052734375, 2.644948974609375, 2.645347412109375, 2.64539453125, 2.64572314453125, 2.645024658203125, 2.64540771484375, 2.64523681640625, 2.64541796875, 2.644529052734375, 2.649079833984375, 2.645666748046875, 2.646816650390625, 2.645528564453125, 2.645729248046875, 2.6472724609375, 2.646464599609375, 2.644356201171875, 2.644589599609375, 2.64574267578125, 2.645088134765625, 2.6443017578125, 2.644094970703125, 2.64595556640625, 2.645572509765625, 2.644812744140625, 5.4503955078125, 2.645255126953125, 2.645544921875, 2.64656884765625, 2.648037353515625, 2.645263427734375, 2.64638671875, 2.6460732421875, 2.6460478515625, 2.645971923828125, 2.645147705078125, 2.646581298828125, 2.6468291015625, 2.64502880859375, 2.645927978515625, 2.64683935546875, 2.64574365234375, 2.645297119140625, 2.64578662109375, 2.645689453125, 2.64529296875, 2.645930908203125, 2.64618798828125, 2.64557373046875, 2.645583984375, 2.64784375, 2.6465341796875, 2.645923828125, 2.648280029296875, 2.647232421875, 2.64726220703125, 2.646688720703125, 2.646456298828125, 2.647128173828125, 2.64785009765625, 2.6451201171875, 2.645867431640625, 2.645412841796875, 2.647185302734375, 2.644790283203125, 2.645506103515625, 2.64781005859375, 2.6459228515625, 2.64651171875, 2.6510693359375, 2.647773193359375, 2.6475908203125, 2.648330322265625, 2.646329345703125, 2.64608349609375, 2.64578759765625, 2.64604150390625, 2.646233154296875, 2.647012451171875, 2.644991943359375, 2.6461533203125, 2.646445068359375, 2.6459423828125, 2.646255615234375, 2.647103515625, 2.64721826171875, 2.6459013671875, 2.646496337890625, 5.45085009765625, 2.646131591796875, 2.647123046875, 2.64625244140625, 2.6469755859375, 2.645900146484375, 2.64694384765625, 2.6482001953125, 2.647193603515625, 2.646340576171875, 2.646576171875, 2.647742431640625, 2.645885986328125, 2.64563916015625, 2.645662841796875, 2.646833251953125, 2.64673291015625, 2.648217529296875, 2.645130126953125, 2.648642578125, 2.64700732421875, 2.6464482421875, 2.645592041015625, 2.646584228515625, 2.64501953125, 2.64555224609375, 2.64648388671875, 2.6452490234375, 2.646088623046875, 2.64539453125, 2.646591552734375, 2.645818359375, 2.6455, 2.645107666015625, 2.647025634765625, 2.64500830078125, 2.647201904296875, 2.6471865234375, 2.647762939453125, 2.6468916015625, 2.646411376953125, 2.645792724609375, 2.647160888671875, 2.64576611328125, 2.645645263671875, 2.646044677734375, 2.647015380859375, 2.64671435546875, 2.645159912109375, 2.646703125, 2.646371337890625, 2.646978515625, 2.6460908203125, 2.648290283203125, 2.646096923828125, 2.646834228515625, 2.6468515625, 2.654116943359375, 2.646571044921875, 2.646265869140625, 2.647033935546875, 2.645507080078125, 2.64506884765625, 5.44818896484375, 2.646950927734375, 2.64728466796875, 2.64570068359375, 2.647045166015625, 2.645780517578125, 2.64650341796875, 2.647185302734375, 2.647396240234375, 2.64726416015625, 2.645887939453125, 2.646220947265625, 2.646308837890625, 2.6452724609375, 2.645916748046875, 2.6456513671875, 2.64530029296875, 2.645833740234375, 2.64532275390625, 2.646921142578125, 2.645572509765625, 2.646182861328125, 2.646026123046875, 2.646593505859375, 2.646531005859375, 2.646508544921875, 2.646131591796875, 2.647275634765625, 2.646612060546875, 2.646274169921875, 2.645583984375, 2.646960205078125, 2.647244873046875, 2.6451845703125, 2.6467060546875, 2.646052001953125, 2.64616650390625, 2.645667724609375, 2.652001220703125, 2.64635693359375, 2.645919677734375, 2.646162353515625, 2.64641845703125, 2.645307373046875, 2.644989013671875, 2.646035400390625, 2.645531494140625, 2.645157958984375, 2.645906494140625, 2.645951416015625, 2.646036376953125, 2.645796875, 2.64467041015625, 2.645960693359375, 2.64549267578125, 2.64652294921875, 2.64587255859375, 2.647626708984375, 2.646477783203125, 2.646447021484375, 2.646128662109375, 2.64745263671875, 2.646246337890625, 5.45296484375, 2.646928466796875, 2.647333984375, 2.646259765625, 2.64629443359375, 2.64618701171875, 2.6448466796875, 2.646443115234375, 2.646686767578125, 2.6447626953125, 2.64540576171875, 2.64494287109375, 2.645835693359375, 2.645675048828125, 2.64589404296875, 2.64648095703125, 2.64646044921875, 2.646432861328125, 2.646467529296875, 2.645669921875, 2.64584814453125, 2.6454580078125, 2.64523486328125, 2.650271728515625, 2.647341064453125, 2.64589306640625, 2.64591064453125, 2.64768505859375, 2.64700830078125, 2.64631201171875, 2.645865478515625, 2.647047119140625, 2.646507568359375, 2.64665087890625, 2.647560302734375, 2.645760986328125, 2.645769287109375, 2.645876708984375, 2.645303466796875, 2.646839111328125, 2.646023193359375, 2.645792724609375, 2.64610400390625, 2.646148193359375, 2.646067138671875, 2.64551416015625, 2.645665771484375, 2.64494482421875, 2.645905517578125, 2.645917724609375, 2.645688232421875, 2.64422705078125, 2.64477587890625, 2.64602001953125, 2.646118408203125, 2.645832763671875, 2.64635888671875, 2.646981689453125, 2.645694580078125, 2.6462197265625, 2.6456484375, 2.647057373046875, 2.646053955078125, 5.45440771484375, 2.64606103515625, 2.64624853515625, 2.645525390625, 2.6462197265625, 2.646288330078125, 2.64601806640625, 2.6458369140625, 2.6465341796875, 2.64682080078125, 2.6460498046875, 2.645517333984375, 2.646379638671875, 2.645887939453125, 2.645572509765625, 2.645886962890625, 2.6497626953125, 2.645772216796875, 2.64593603515625, 2.64641748046875, 2.64673681640625, 2.645414794921875, 2.644895751953125, 2.6456689453125, 2.646624267578125, 2.64646142578125, 2.645294921875, 2.646984619140625, 2.646635498046875, 2.64559619140625, 2.644748291015625, 2.64690576171875, 2.64658642578125, 2.645821533203125, 2.6462392578125, 2.647307373046875, 2.64610205078125, 2.647371826171875, 2.64780078125, 2.64745263671875, 2.646867919921875, 2.64559814453125, 2.646322265625, 2.6461328125, 2.6461328125, 2.6467666015625, 2.64734619140625, 2.647150634765625, 2.6465341796875, 2.646138916015625, 2.645937255859375, 2.64439697265625, 2.645222412109375, 2.646958984375, 2.646992919921875, 2.647083984375, 2.64567822265625, 2.646795166015625, 2.645927978515625, 2.646295654296875, 2.64530224609375, 2.646948974609375, 2.646352783203125, 5.45413623046875, 2.646625244140625, 2.6464501953125, 2.647415771484375, 2.64749462890625, 2.64658837890625, 2.645314453125, 2.647509033203125, 2.647132080078125, 2.64675537109375, 2.646547607421875, 2.646182861328125, 2.64610302734375, 2.6463896484375, 2.646077392578125, 2.64599853515625, 2.646762451171875, 2.64707080078125, 2.646077392578125, 2.64540576171875, 2.646277099609375, 2.64477392578125, 2.645159912109375, 2.6459638671875, 2.64671240234375, 2.646478759765625, 2.645445556640625, 2.646115234375, 2.646445068359375, 2.645536865234375, 2.64454443359375, 2.646118408203125, 2.645505126953125, 2.644704345703125, 2.6460517578125, 2.64701953125, 2.645275634765625, 2.64492431640625, 2.645380126953125, 2.645751708984375, 2.64485888671875, 2.64445654296875, 2.645286865234375, 2.646550537109375, 2.646221923828125, 2.645315673828125, 2.6454169921875, 2.64634375, 2.645382080078125, 2.6473984375, 2.646042724609375, 2.64547119140625, 2.644828125, 2.64538720703125, 2.646898681640625, 2.646388671875, 2.64596484375, 2.6477158203125, 2.64626171875, 2.6452685546875, 2.645350341796875, 2.6460498046875, 2.644905029296875, 5.4557060546875, 2.64730419921875, 2.64740966796875, 2.646350830078125, 2.645677978515625, 2.6465546875, 2.64831689453125, 2.64722119140625, 2.646279052734375, 2.645445556640625, 2.646408203125, 2.646245361328125, 2.646921142578125, 2.64698974609375, 2.646667236328125, 2.646288330078125, 2.64876953125, 2.647458740234375, 2.645729248046875, 2.645494873046875, 2.646150146484375, 2.6453955078125, 2.6452890625, 2.6467861328125, 2.647785400390625, 2.646699951171875, 2.65012744140625, 2.64613671875, 2.646697998046875, 2.6466845703125, 2.646642578125, 2.6476943359375, 2.64681787109375, 2.645788818359375, 2.646158203125, 2.647160888671875, 2.645675048828125, 2.64652490234375, 2.64439306640625, 2.64669189453125, 2.646427734375, 2.64625146484375, 2.646435791015625, 2.645409912109375, 2.646538330078125, 2.64618505859375, 2.6468310546875, 2.647814208984375, 2.645505126953125, 2.645622802734375, 2.6466826171875, 2.646696044921875, 2.6457548828125, 2.6469130859375, 2.6469990234375, 2.64591259765625, 2.64690185546875, 2.64622705078125, 2.64720703125, 2.646657958984375, 2.645370849609375, 2.645916748046875, 2.646266845703125]",tokens/s,0.3722418346281902,,,,, -4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-125m,,cuda,0,42,,,,,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,main,False,False,MB,1310.69952,921.174016,0.0,274.726912,220.646912,s,10,0.36048070526123044,0.03604807052612304,0.0010977874344766398,0.03588955307006836,0.036477619552612306,0.03779876155853271,0.03885567516326904,"[0.03911990356445313, 0.03553478240966797, 0.03469347381591797, 0.035817951202392576, 0.036184032440185546, 0.035873409271240234, 0.03543529510498047, 0.03591993713378906, 0.03590569686889648, 0.03599622344970703]",tokens/s,7101.628360787961,kWh,4.1762897090252873e-07,2.2884077887720709e-07,9.141624902616015e-07,1.5606322400413373e-06,tokens/kWh,164036083.2179266,MB,1311.0272,921.174016,0.0,274.726912,250.723328,s,10,21.866080322265628,2.186608032226563,0.01854961703991556,2.191078857421875,2.2041080810546876,2.2073608032226564,2.2099629809570316,"[2.1803974609375, 2.17111474609375, 2.154228759765625, 2.200171142578125, 2.210613525390625, 2.181986572265625, 2.161680419921875, 2.20071923828125, 2.20338525390625, 2.201783203125]",tokens/s,28.81174818325753,kWh,2.578141039891902e-05,1.4128932798644376e-05,5.28597971955373e-05,9.27701403931007e-05,tokens/kWh,679097.8188999841,,s,629,22.151173107147223,0.03521649142630718,0.004356343174673542,0.03486105728149414,0.035241778564453125,0.03557929000854492,0.07092830078125002,"[0.03474124908447265, 0.03529420852661133, 0.03632339096069336, 0.03650860977172852, 0.036083713531494144, 0.037152767181396484, 0.03645132827758789, 0.036311038970947264, 0.03472588729858399, 0.0354252815246582, 0.03531980895996094, 0.03479657745361328, 0.03465929412841797, 0.03616358566284179, 0.035585025787353515, 0.03515903854370117, 0.03486105728149414, 0.034651134490966795, 0.034900993347167966, 0.03449958419799805, 0.03328102493286133, 0.03349708938598633, 0.033560577392578124, 0.03324620819091797, 0.033328128814697267, 0.033255424499511715, 0.033754112243652344, 0.033223743438720706, 0.03327481460571289, 0.033438720703125, 0.0335175666809082, 0.03376332855224609, 0.03339468765258789, 0.03345305633544922, 0.03355033493041992, 0.034203647613525394, 0.03342233657836914, 0.03335372924804687, 0.0336732177734375, 0.03367116928100586, 0.03360870361328125, 0.03362508773803711, 0.03436441421508789, 0.03604377746582031, 0.03504127883911133, 0.03473920059204102, 0.0348037109375, 0.03472895812988281, 0.03472895812988281, 0.03479865646362305, 0.03479955291748047, 0.03479449462890625, 0.03474739074707031, 0.03478220748901367, 0.0347064323425293, 0.034907135009765625, 0.03474431991577148, 0.03473715209960938, 0.03485388946533203, 0.034683902740478514, 0.035019775390625, 0.0350750732421875, 0.0711178207397461, 0.0347955207824707, 0.034948097229003904, 0.03439616012573242, 0.03447808074951172, 0.03496755218505859, 0.034920448303222655, 0.0348590087890625, 0.034699264526367186, 0.0346163215637207, 0.03491430282592774, 0.03490304183959961, 0.03508019256591797, 0.034423809051513675, 0.03472793579101562, 0.03449446487426758, 0.03476377487182617, 0.03476582336425781, 0.03468902587890625, 0.03387494277954101, 0.0342210578918457, 0.034008094787597656, 0.03374998474121094, 0.03369062423706055, 0.033459201812744144, 0.03354937744140625, 0.03469715118408203, 0.03489484786987305, 0.03487027359008789, 0.03510067367553711, 0.03503513717651367, 0.03503923034667969, 0.03469619369506836, 0.03455590438842773, 0.03517852783203125, 0.03441353607177734, 0.03464704132080078, 0.03492659378051758, 0.034506752014160154, 0.03476070404052734, 0.03487846374511719, 0.03493478393554687, 0.03499212646484375, 0.034945056915283206, 0.035192798614501954, 0.03450265502929688, 0.03448524856567383, 0.03455487823486328, 0.034563072204589845, 0.03503104019165039, 0.03474431991577148, 0.03377872085571289, 0.03339875030517578, 0.03340800094604492, 0.03347558212280274, 0.03381350326538086, 0.033650688171386715, 0.033535999298095705, 0.033334270477294925, 0.03363123321533203, 0.03390771102905273, 0.03361996841430664, 0.03386880111694336, 0.06886093139648437, 0.03341823959350586, 0.03395993423461914, 0.03393228912353516, 0.03374387359619141, 0.03333529663085937, 0.0334919662475586, 0.03423027038574219, 0.03382476806640625, 0.03374387359619141, 0.0336824951171875, 0.033724353790283206, 0.03363532638549805, 0.033941505432128906, 0.03381760025024414, 0.033503231048583985, 0.03344998550415039, 0.033301502227783206, 0.033445888519287106, 0.033957889556884766, 0.03387801742553711, 0.03366092681884766, 0.03360153579711914, 0.03381043243408203, 0.033825790405273434, 0.03360870361328125, 0.03370703887939453, 0.03348271942138672, 0.03366502380371094, 0.03378995132446289, 0.03319193649291992, 0.03369267272949219, 0.03357900619506836, 0.033358848571777344, 0.0341104621887207, 0.03495731353759766, 0.03543961715698242, 0.03482316970825195, 0.034435073852539064, 0.034557952880859374, 0.034900993347167966, 0.034678783416748044, 0.03517542266845703, 0.03521023941040039, 0.035046398162841795, 0.03481702423095703, 0.03453235244750977, 0.03470131301879883, 0.03486412811279297, 0.03474739074707031, 0.03477196884155274, 0.03457843017578125, 0.034781185150146485, 0.03486310577392578, 0.035125247955322264, 0.03486617660522461, 0.03496243286132812, 0.03503513717651367, 0.03523891067504883, 0.03469823837280273, 0.0339128303527832, 0.03414220809936523, 0.035117088317871095, 0.0711659164428711, 0.03465318298339844, 0.03465727996826172, 0.03458355331420898, 0.034283519744873044, 0.034304000854492187, 0.03577958297729492, 0.03595161437988281, 0.03544473648071289, 0.03500543975830078, 0.03521843338012695, 0.03495935821533203, 0.0349194221496582, 0.0354252815246582, 0.03515903854370117, 0.0349306869506836, 0.03492454528808594, 0.036211711883544925, 0.03564441680908203, 0.03498086547851562, 0.034841598510742186, 0.034941951751708986, 0.03537612915039062, 0.034994174957275394, 0.0350382080078125, 0.03486105728149414, 0.03484672164916992, 0.03494297790527344, 0.03524095916748047, 0.03526041412353516, 0.03520614242553711, 0.03488153457641602, 0.03502796936035156, 0.03500543975830078, 0.03404185485839844, 0.033903617858886716, 0.03378073501586914, 0.03472281646728516, 0.03411251068115234, 0.03379097747802735, 0.03348787307739258, 0.0343633918762207, 0.03492659378051758, 0.03500339126586914, 0.03493788909912109, 0.03485590362548828, 0.034895870208740236, 0.034948097229003904, 0.03518771362304687, 0.034479103088378905, 0.03459174346923828, 0.035253280639648436, 0.034963424682617185, 0.034976768493652347, 0.03491843032836914, 0.034911201477050784, 0.035046398162841795, 0.03507199859619141, 0.03536383819580078, 0.03499929428100586, 0.03523379135131836, 0.03531161499023437, 0.03527679824829102, 0.07160934448242187, 0.0352624626159668, 0.03535769653320313, 0.03481087875366211, 0.03554921722412109, 0.034968544006347656, 0.03489491271972656, 0.034890689849853516, 0.03483238220214844, 0.034993152618408206, 0.035932159423828124, 0.03508838272094727, 0.03496755218505859, 0.03476377487182617, 0.03488358306884766, 0.03527884674072266, 0.03521228790283203, 0.035040256500244144, 0.03476684951782227, 0.03499622344970703, 0.03482316970825195, 0.03518259048461914, 0.03503411102294922, 0.036178943634033206, 0.039564289093017575, 0.035097599029541016, 0.03457843017578125, 0.03505561447143555, 0.034756607055664065, 0.0350013427734375, 0.03477811050415039, 0.035004417419433595, 0.03501772689819336, 0.0348671989440918, 0.03473100662231445, 0.0349224967956543, 0.03482931137084961, 0.03578777694702148, 0.0350115852355957, 0.03483443069458008, 0.03486412811279297, 0.03486207962036133, 0.034938880920410156, 0.03505561447143555, 0.035004417419433595, 0.034887680053710936, 0.034531326293945314, 0.03493580627441406, 0.03510067367553711, 0.03492454528808594, 0.034887680053710936, 0.03578879928588867, 0.034974720001220705, 0.03519180679321289, 0.034850879669189455, 0.03495315170288086, 0.035060768127441404, 0.034809825897216796, 0.03377766418457031, 0.03550620651245117, 0.034968544006347656, 0.034915328979492184, 0.03489177703857422, 0.07134719848632813, 0.03493273544311523, 0.03501465606689453, 0.03539148712158203, 0.03512319946289062, 0.034813953399658204, 0.03508019256591797, 0.03486515045166016, 0.035079166412353514, 0.03493580627441406, 0.034939903259277344, 0.034925567626953126, 0.03743027114868164, 0.034576385498046876, 0.03397836685180664, 0.03377356719970703, 0.03403878402709961, 0.03371724700927734, 0.03380223846435547, 0.03486310577392578, 0.03486412811279297, 0.03496243286132812, 0.03475046539306641, 0.03492659378051758, 0.03499622344970703, 0.034783233642578126, 0.033949695587158206, 0.033876991271972655, 0.03427123260498047, 0.03477811050415039, 0.03472076797485352, 0.034127872467041014, 0.03329740905761719, 0.0333834228515625, 0.03379814529418945, 0.03392409515380859, 0.03363532638549805, 0.03386982345581055, 0.03498086547851562, 0.034928638458251955, 0.03496345520019531, 0.03495423889160156, 0.03523891067504883, 0.03508838272094727, 0.034769920349121096, 0.03489279937744141, 0.03482624053955078, 0.034955265045166016, 0.03508838272094727, 0.03486412811279297, 0.03486515045166016, 0.03471769714355469, 0.034560001373291016, 0.034579456329345705, 0.03481292724609375, 0.03479244613647461, 0.034802688598632815, 0.03475254440307617, 0.03517948913574219, 0.03433884811401367, 0.033702880859375, 0.03359539031982422, 0.034111488342285154, 0.07044096374511719, 0.03437670516967774, 0.03483238220214844, 0.0348671989440918, 0.03470438385009766, 0.035230720520019534, 0.034816001892089846, 0.03467366409301758, 0.0347770881652832, 0.03501055908203125, 0.03429580688476563, 0.03443609619140625, 0.03509964752197266, 0.03524505615234375, 0.0347248649597168, 0.034753536224365236, 0.03496345520019531, 0.03496448135375976, 0.034770942687988284, 0.034799617767333986, 0.03484467315673828, 0.03481190490722656, 0.034353153228759765, 0.034062335968017575, 0.03349094390869141, 0.033584129333496096, 0.033587200164794925, 0.034141185760498044, 0.0347740478515625, 0.0345906867980957, 0.0344268798828125, 0.03489484786987305, 0.03420979309082031, 0.03364352035522461, 0.03377356719970703, 0.033686527252197264, 0.033290241241455076, 0.0337520637512207, 0.03401932907104492, 0.03371417617797851, 0.03375513458251953, 0.033691646575927735, 0.03370809555053711, 0.03350316619873047, 0.03325439834594727, 0.03360255813598633, 0.033860607147216795, 0.03376230239868164, 0.033691646575927735, 0.033783809661865234, 0.033797119140625, 0.03425177764892578, 0.03381452941894531, 0.033791999816894534, 0.033675262451171875, 0.03391385650634766, 0.034165760040283204, 0.03502899169921875, 0.03486822509765625, 0.034702335357666016, 0.035046398162841795, 0.03521535873413086, 0.034825214385986326, 0.07150080108642579, 0.03484364700317383, 0.034351104736328124, 0.03453747177124023, 0.035064830780029296, 0.03482931137084961, 0.03483238220214844, 0.03492659378051758, 0.034974720001220705, 0.03501465606689453, 0.034869247436523435, 0.034830337524414064, 0.034277374267578126, 0.034797569274902344, 0.03501363372802734, 0.03530342483520508, 0.0351016960144043, 0.03491635131835937, 0.03483238220214844, 0.03484672164916992, 0.03495116806030273, 0.03500646209716797, 0.03499827194213867, 0.03480678558349609, 0.03486207962036133, 0.03505152130126953, 0.034784255981445314, 0.034783233642578126, 0.034835456848144535, 0.03493580627441406, 0.034770942687988284, 0.035037185668945314, 0.035019775390625, 0.0342476806640625, 0.03505049514770508, 0.034315265655517575, 0.03477913665771484, 0.034791454315185544, 0.0353023681640625, 0.03562700653076172, 0.03510374450683594, 0.03493580627441406, 0.035156993865966796, 0.03526144027709961, 0.03488256072998047, 0.034933761596679686, 0.03517337417602539, 0.0348590087890625, 0.0352911376953125, 0.035888126373291016, 0.03522048187255859, 0.03506073760986328, 0.03483955383300781, 0.03486105728149414, 0.03498400115966797, 0.03498387145996094, 0.03485494232177734, 0.03438998413085938, 0.03479244613647461, 0.03507712173461914, 0.03500339126586914, 0.03471257781982422, 0.03476684951782227, 0.07122022247314454, 0.03482316970825195, 0.03523276901245117, 0.034941951751708986, 0.03477196884155274, 0.034991104125976565, 0.03495731353759766, 0.034939903259277344, 0.03488153457641602, 0.034900993347167966, 0.0347955207824707, 0.0350300178527832, 0.03487948989868164, 0.03520614242553711, 0.035297279357910154, 0.03513651275634765, 0.03494297790527344, 0.0349378547668457, 0.034948097229003904, 0.03488051223754883, 0.03481497573852539, 0.03454873657226563, 0.03495935821533203, 0.035062782287597655, 0.035043327331542966, 0.0349306869506836, 0.03497881698608398, 0.034909183502197266, 0.03493791961669922, 0.03483539199829101, 0.034869247436523435, 0.035031105041503904, 0.03493983840942383, 0.03483340835571289, 0.03487744140625, 0.03489484786987305, 0.03516108703613281, 0.03484364700317383, 0.034969600677490234, 0.035146751403808595, 0.035194881439208986, 0.03479244613647461, 0.03557068634033203, 0.03522969436645508, 0.035337215423583986, 0.03480473709106445, 0.03436236953735351, 0.033745918273925785, 0.03496755218505859, 0.03541401672363281, 0.03512934494018555, 0.035163135528564454, 0.0348487663269043, 0.03500339126586914, 0.03484467315673828, 0.03498495864868164, 0.03576115036010742, 0.03517030334472656, 0.03490816116333008, 0.034825214385986326, 0.0347586555480957, 0.03516723251342774, 0.035053569793701174, 0.07151315307617187, 0.03519071960449219, 0.03483852767944336, 0.03482419204711914, 0.03502592086791992, 0.03529216003417969, 0.035125247955322264, 0.034872318267822264, 0.034939903259277344, 0.0347658576965332, 0.03481699371337891, 0.03497267150878906, 0.03488665771484375, 0.034958335876464845, 0.035007488250732424, 0.03504537582397461, 0.03489382553100586, 0.03520716857910156, 0.034988033294677735, 0.034874366760253905, 0.03459481430053711, 0.03430297470092773, 0.0346879997253418, 0.03478732681274414, 0.03508736038208008, 0.0354856948852539, 0.034981952667236325, 0.0348732795715332, 0.03489382553100586, 0.03475558471679688, 0.03478732681274414, 0.03488358306884766, 0.03473920059204102, 0.03476172637939453, 0.034685951232910156, 0.035111934661865234, 0.03508736038208008, 0.03526553726196289, 0.03492147064208984, 0.03475763320922851, 0.034767871856689454, 0.03482316970825195, 0.03498092651367188, 0.03539142227172851, 0.03607961654663086, 0.03499008178710938, 0.03491123199462891, 0.034990142822265625, 0.03488249588012695, 0.03484467315673828, 0.03486207962036133, 0.0349378547668457, 0.034953216552734374, 0.03488051223754883, 0.03490816116333008, 0.0349378547668457, 0.03494604873657227, 0.03508838272094727, 0.03482726287841797, 0.03486003112792969, 0.03481702423095703, 0.034729984283447264, 0.0349224967956543]",tokens/s,28.395787300179112,,,,, -4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-6B,,cuda,0,42,,,,,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,main,False,False,MB,1575.784448,5448.925184,0.0,4802.47808,4489.252352,s,10,5.114569244384766,0.5114569244384766,0.0014045752365005437,0.51150537109375,0.5130841125488281,0.5135362274169921,0.5138979193115234,"[0.512983642578125, 0.5139883422851562, 0.5091934204101562, 0.5101927795410156, 0.5104193725585937, 0.5101881103515625, 0.5112746887207031, 0.5121220092773437, 0.5117360534667968, 0.5124708251953125]",tokens/s,500.5309103617276,kWh,6.020330099595917e-06,3.2988774182740595e-06,2.7782869448500503e-05,3.710207696637048e-05,tokens/kWh,6899883.266158921,MB,1575.784448,5448.925184,0.0,4802.47808,4557.794816,s,10,301.333625,30.133362499999997,0.011965382718741807,30.1356015625,30.146014453125,30.1513353515625,30.1555920703125,"[30.12017578125, 30.120552734375, 30.1369375, 30.140724609375, 30.15665625, 30.137947265625, 30.123359375, 30.14483203125, 30.118173828125, 30.134265625]",tokens/s,2.0907059409649356,kWh,0.00035569889715148347,0.00019495335105611957,0.001580753417379499,0.002131405665587102,tokens/kWh,29557.958401431977,,s,629,305.4536612854004,0.48561790347440437,0.060780789003776696,0.4781598815917969,0.47936102294921873,0.4797579284667969,0.9888164648437501,"[0.47731610107421873, 0.4777103271484375, 0.4780533752441406, 0.4779223022460938, 0.47746868896484373, 0.47739285278320315, 0.4777676696777344, 0.47874050903320314, 0.47792022705078124, 0.4776212463378906, 0.47766937255859376, 0.4774625244140625, 0.4779018249511719, 0.47790591430664064, 0.47894528198242187, 0.47799398803710935, 0.47878759765625, 0.47878964233398436, 0.47852645874023436, 0.47817422485351563, 0.4778905334472656, 0.47780557250976563, 0.47809637451171877, 0.47767245483398435, 0.477765625, 0.4778649597167969, 0.4779622497558594, 0.47779531860351565, 0.47874456787109376, 0.47817214965820315, 0.4782950439453125, 0.47745944213867186, 0.47759051513671874, 0.4785008544921875, 0.47783013916015626, 0.4780267639160156, 0.4782438354492187, 0.4785858459472656, 0.4793190307617187, 0.47905487060546875, 0.47873635864257813, 0.47829400634765623, 0.4777697143554688, 0.4778240661621094, 0.47742047119140624, 0.4780851135253906, 0.4782008361816406, 0.47801651000976564, 0.47831243896484377, 0.47876199340820313, 0.4778547058105469, 0.47859405517578124, 0.4782591857910156, 0.47776461791992186, 0.47824075317382814, 0.4776304626464844, 0.4778383483886719, 0.4776642456054688, 0.47760385131835936, 0.478160888671875, 0.47876300048828124, 0.4783943786621094, 0.9927915649414063, 0.47756494140625, 0.4774000549316406, 0.47765298461914063, 0.47747378540039065, 0.4782704772949219, 0.477907958984375, 0.4783944091796875, 0.477813720703125, 0.47758950805664063, 0.47759051513671874, 0.4776069030761719, 0.47729867553710936, 0.4773724060058594, 0.47751473999023436, 0.47745742797851565, 0.4780113525390625, 0.4786903076171875, 0.47860427856445314, 0.4783482971191406, 0.47794790649414065, 0.47776461791992186, 0.4778291320800781, 0.4779427795410156, 0.4779376525878906, 0.47769189453125, 0.47908352661132814, 0.47773797607421875, 0.47901287841796875, 0.4777134094238281, 0.4779151306152344, 0.47777587890625, 0.47748709106445314, 0.47771136474609377, 0.4781527099609375, 0.47798886108398436, 0.47750042724609376, 0.47809127807617186, 0.47905892944335937, 0.4782591857910156, 0.4781598815917969, 0.4782981262207031, 0.4776468505859375, 0.47786392211914064, 0.4776560668945313, 0.478023681640625, 0.48116427612304685, 0.4793456726074219, 0.4777492370605469, 0.47873126220703127, 0.47907122802734375, 0.4783626098632813, 0.4779949951171875, 0.47790286254882813, 0.4776908874511719, 0.478065673828125, 0.4777738342285156, 0.477770751953125, 0.47832369995117185, 0.47939175415039065, 0.47875686645507814, 0.47773284912109376, 0.478023681640625, 0.988906494140625, 0.47783013916015626, 0.4776437683105469, 0.4778649597167969, 0.4783902587890625, 0.47753521728515624, 0.4795494384765625, 0.47817214965820315, 0.4789770202636719, 0.4776386413574219, 0.47787213134765627, 0.47758950805664063, 0.47774514770507814, 0.47762841796875, 0.47790386962890624, 0.479541259765625, 0.4791797790527344, 0.47914700317382813, 0.4779294738769531, 0.4777625732421875, 0.47816497802734376, 0.4782356567382812, 0.47807180786132814, 0.4797726745605469, 0.4777687072753906, 0.4803246154785156, 0.4800747375488281, 0.4793231506347656, 0.4796539001464844, 0.48039935302734377, 0.4796252136230469, 0.4790927429199219, 0.477655029296875, 0.4791510925292969, 0.4787394409179688, 0.47877017211914064, 0.47938558959960936, 0.47959552001953126, 0.47838516235351564, 0.47773593139648435, 0.478129150390625, 0.4781404113769531, 0.4776857604980469, 0.4779018249511719, 0.47747378540039065, 0.4776847229003906, 0.4781803588867187, 0.4799620971679687, 0.4786053161621094, 0.47794790649414065, 0.4777123718261719, 0.4776683654785156, 0.4774696960449219, 0.47712460327148437, 0.4774481811523438, 0.47752908325195315, 0.47732632446289064, 0.4782438354492187, 0.47871487426757814, 0.47837286376953125, 0.4777123718261719, 0.47754238891601564, 0.4779929504394531, 0.9885849609375, 0.47834625244140627, 0.47808615112304687, 0.47836468505859375, 0.47836468505859375, 0.4776908874511719, 0.47762738037109376, 0.4779254455566406, 0.477751220703125, 0.4784117736816406, 0.47790286254882813, 0.4781741943359375, 0.4787138671875, 0.47869439697265626, 0.47841998291015625, 0.4790876159667969, 0.47936920166015623, 0.4789770202636719, 0.4778157958984375, 0.47849676513671874, 0.4780349426269531, 0.4785776672363281, 0.4781475830078125, 0.4785581970214844, 0.4792135620117188, 0.47874969482421875, 0.4787026062011719, 0.47837899780273435, 0.47872308349609377, 0.47839334106445314, 0.4786411437988281, 0.4790169677734375, 0.47768063354492185, 0.4778465270996094, 0.4781793212890625, 0.47906610107421876, 0.47883367919921876, 0.4782213134765625, 0.4785172424316406, 0.47826739501953125, 0.478087158203125, 0.4777277526855469, 0.4778680419921875, 0.47746456909179685, 0.4802908020019531, 0.47867086791992186, 0.47896060180664063, 0.4792279052734375, 0.4781783142089844, 0.47810867309570315, 0.4786063232421875, 0.4785162353515625, 0.4785848388671875, 0.4778659973144531, 0.4776212463378906, 0.4777840576171875, 0.4788787231445312, 0.47973171997070313, 0.47971328735351565, 0.4792842102050781, 0.477655029296875, 0.47770315551757814, 0.4779049377441406, 0.9883463134765625, 0.4791357421875, 0.4782458801269531, 0.4794173583984375, 0.478166015625, 0.47800833129882814, 0.4775157775878906, 0.4780421142578125, 0.47852032470703126, 0.4791224365234375, 0.47801651000976564, 0.47799398803710935, 0.4782530517578125, 0.47967129516601564, 0.47834417724609374, 0.4779847717285156, 0.4790302734375, 0.47956378173828124, 0.4773304443359375, 0.4774420471191406, 0.4774143981933594, 0.4773447570800781, 0.47814349365234377, 0.47828582763671873, 0.479710205078125, 0.4788910217285156, 0.479056884765625, 0.47832369995117185, 0.4782438354492187, 0.4783964233398437, 0.479025146484375, 0.4786606140136719, 0.4782909545898438, 0.47821823120117185, 0.4802129821777344, 0.47790286254882813, 0.47835751342773436, 0.4782561340332031, 0.478635009765625, 0.47870156860351565, 0.4777001037597656, 0.4775802917480469, 0.47765914916992186, 0.479578125, 0.4784527282714844, 0.47922894287109374, 0.47923199462890625, 0.4788213806152344, 0.47869439697265626, 0.4787804260253906, 0.47939993286132815, 0.47935186767578125, 0.4787639770507813, 0.4786268310546875, 0.4793589782714844, 0.48041677856445314, 0.48034817504882815, 0.482777099609375, 0.4785745849609375, 0.478571533203125, 0.47958123779296874, 0.47865029907226564, 0.4787384338378906, 0.990581787109375, 0.4781793212890625, 0.4788521423339844, 0.47797039794921875, 0.47851007080078123, 0.47790286254882813, 0.4778321838378906, 0.47813427734375, 0.4789043273925781, 0.47835134887695313, 0.47826226806640626, 0.47895858764648436, 0.4790947875976562, 0.4783759460449219, 0.47830938720703126, 0.4781803588867187, 0.477991943359375, 0.47756903076171875, 0.47764480590820313, 0.47726080322265624, 0.47773797607421875, 0.47950848388671874, 0.47766015625, 0.47899853515625, 0.47830322265625, 0.4781537170410156, 0.47825204467773436, 0.4785592346191406, 0.479025146484375, 0.47944705200195314, 0.47835134887695313, 0.478445556640625, 0.47876199340820313, 0.47891659545898435, 0.4791101379394531, 0.47820391845703125, 0.4782581787109375, 0.47862374877929686, 0.4814622802734375, 0.47808819580078127, 0.47835751342773436, 0.47878964233398436, 0.4785551452636719, 0.4777082824707031, 0.4787271728515625, 0.4784322509765625, 0.47835546875, 0.4775679931640625, 0.4779346008300781, 0.47834521484375, 0.478160888671875, 0.47767962646484374, 0.47781170654296873, 0.4783585205078125, 0.4791285705566406, 0.4783697814941406, 0.47800216674804685, 0.47763250732421875, 0.4782294921875, 0.47773696899414064, 0.47801651000976564, 0.4774718017578125, 0.47802670288085936, 0.9896888427734375, 0.47829608154296877, 0.47802975463867187, 0.47756497192382813, 0.47750039672851563, 0.47811380004882814, 0.4781025390625, 0.4778936767578125, 0.47798370361328124, 0.47788851928710935, 0.4786268310546875, 0.4782899169921875, 0.4779632568359375, 0.47777484130859377, 0.47753521728515624, 0.4779007873535156, 0.47758544921875, 0.47753726196289065, 0.47787213134765627, 0.47847628784179685, 0.4782909545898438, 0.4778014831542969, 0.47860223388671874, 0.47986483764648435, 0.47783935546875, 0.47765298461914063, 0.4779980773925781, 0.47794073486328126, 0.4787517395019531, 0.4781506652832031, 0.4780155029296875, 0.47900875854492186, 0.47818548583984377, 0.47805029296875, 0.4782233581542969, 0.47802471923828127, 0.4779346008300781, 0.47827557373046875, 0.478497802734375, 0.47849063110351564, 0.47899237060546873, 0.4786196594238281, 0.47815167236328127, 0.4799846801757813, 0.47795709228515626, 0.47788851928710935, 0.4778270568847656, 0.4778547058105469, 0.478060546875, 0.47782608032226564, 0.4778434143066406, 0.477601806640625, 0.4788572082519531, 0.47828274536132814, 0.47832986450195314, 0.4779417724609375, 0.477633544921875, 0.477907958984375, 0.478065673828125, 0.4780482482910156, 0.47808819580078127, 0.47858380126953126, 0.47811892700195313, 0.9911572265625, 0.4780707702636719, 0.47788134765625, 0.4775536499023437, 0.4774696960449219, 0.47756185913085936, 0.47758642578125, 0.4777615356445313, 0.47794790649414065, 0.4787118225097656, 0.47779736328125, 0.47749325561523437, 0.4775475158691406, 0.4777697143554688, 0.4781363220214844, 0.4779141235351563, 0.4792012939453125, 0.4778874816894531, 0.47856845092773437, 0.47768267822265625, 0.4781311950683594, 0.4785070190429688, 0.47809127807617186, 0.4776365966796875, 0.4780707702636719, 0.47816497802734376, 0.47828582763671873, 0.47777279663085936, 0.4780728454589844, 0.47835751342773436, 0.47851211547851563, 0.47822235107421873, 0.4775055236816406, 0.4781465454101563, 0.47790386962890624, 0.4779018249511719, 0.478359619140625, 0.4794295654296875, 0.4784394226074219, 0.4795248718261719, 0.4780544128417969, 0.477949951171875, 0.48094821166992185, 0.4794306640625, 0.4784814147949219, 0.4792197265625, 0.4782981262207031, 0.47831655883789065, 0.47831243896484377, 0.4784998474121094, 0.47935488891601563, 0.47914599609375, 0.48005630493164064, 0.48013311767578126, 0.47997030639648436, 0.48005426025390624, 0.4799140625, 0.4794142150878906, 0.4800153503417969, 0.4793067626953125, 0.47961505126953125, 0.47790789794921873, 0.4778670043945312, 0.9900185546875, 0.4775372924804687, 0.47779840087890624, 0.477633544921875, 0.4782950439453125, 0.47761920166015626, 0.47800421142578126, 0.478803955078125, 0.4791029663085937, 0.4782643127441406, 0.47842098999023436, 0.4776908874511719, 0.4780185546875, 0.47770932006835937, 0.4777851257324219, 0.477681640625, 0.477812744140625, 0.47979006958007814, 0.47791717529296873, 0.4780451965332031, 0.4788193359375, 0.477955078125, 0.47790591430664064, 0.47803903198242187, 0.47824697875976563, 0.4786861572265625, 0.47816192626953125, 0.477655029296875, 0.47835134887695313, 0.478919677734375, 0.47831039428710936, 0.4777205810546875, 0.4773253173828125, 0.47773284912109376, 0.4778260498046875, 0.4776069030761719, 0.47773492431640624, 0.4777461853027344, 0.478587890625, 0.47767041015625, 0.4789801330566406, 0.47915618896484374, 0.4780257263183594, 0.47762841796875, 0.4778168334960938, 0.4776714172363281, 0.47783526611328125, 0.47757107543945315, 0.4779704284667969, 0.4786677856445313, 0.47908148193359373, 0.4783206481933594, 0.4781240234375, 0.47756494140625, 0.4773294067382812, 0.47757208251953126, 0.477348876953125, 0.47757720947265625, 0.47773492431640624, 0.47816192626953125, 0.47770932006835937, 0.4789381103515625, 0.47819366455078127, 0.9904609375, 0.47848550415039065, 0.478013427734375, 0.4775751647949219, 0.4774912109375, 0.4782847900390625, 0.4794931335449219, 0.4796549072265625, 0.47851318359375, 0.4785254211425781, 0.47859506225585935, 0.4788695068359375, 0.47816293334960935, 0.47827969360351563, 0.4785244140625, 0.4797358093261719, 0.4782438354492187, 0.4776570739746094, 0.4789125061035156, 0.47882650756835937, 0.4777082824707031, 0.47753216552734373, 0.47809637451171877, 0.47831655883789065, 0.4776488952636719, 0.4774471740722656, 0.4796170349121094, 0.47986483764648435, 0.4793456726074219, 0.479599609375, 0.4791654357910156, 0.4776488952636719, 0.4775475158691406, 0.47792538452148436, 0.4778240051269531, 0.4774819946289062, 0.4782612915039062, 0.4777738037109375, 0.4777543640136719, 0.47963134765625, 0.47828070068359374, 0.4778486022949219, 0.4777625427246094, 0.477601806640625, 0.4777185363769531, 0.47762432861328125, 0.47770932006835937, 0.4777062377929687, 0.47948800659179686, 0.4784066467285156, 0.47847015380859376, 0.4775106506347656, 0.4777420654296875, 0.47842098999023436, 0.47805233764648436, 0.47766015625, 0.47758746337890623, 0.47821926879882815, 0.4781087036132812, 0.4776806640625, 0.47913876342773437, 0.47851828002929686, 0.4783575744628906]",tokens/s,2.059232151132392,,,,, -4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-2.7B,,cuda,0,42,,,,,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,main,False,False,MB,2236.862464,2932.342784,0.0,2285.89568,2082.706944,s,10,2.5091812438964842,0.25091812438964844,0.0016359490817563227,0.25024176025390626,0.25317278594970705,0.2535381980895996,0.25383052780151366,"[0.25309158325195313, 0.2539036102294922, 0.24987110900878906, 0.24888946533203124, 0.2497696075439453, 0.24914796447753906, 0.2504080047607422, 0.25179104614257813, 0.25223333740234377, 0.2500755157470703]",tokens/s,1020.253122897013,kWh,2.948670335578402e-06,1.6157431971317612e-06,1.3066561943753649e-05,1.763097547646381e-05,tokens/kWh,14519899.953450851,MB,2238.38208,2959.60576,0.0,2313.158656,2180.685312,s,10,143.8805810546875,14.38805810546875,0.011353783216855234,14.382828125,14.40145888671875,14.403668896484376,14.405436904296876,"[14.4009677734375, 14.40587890625, 14.39833203125, 14.3805361328125, 14.374423828125, 14.3830166015625, 14.400140625, 14.373748046875, 14.3808974609375, 14.3826396484375]",tokens/s,4.378631191102458,kWh,0.000169606630340582,9.295822156634463e-05,0.0007504743544440634,0.0010130392063509902,tokens/kWh,62189.10344736671,,s,629,145.90565167236326,0.2319644700673502,0.029875164751919897,0.22824755859375,0.22927770080566406,0.22946876831054688,0.47820606079101563,"[0.22992076110839843, 0.22815129089355468, 0.22794752502441407, 0.22817485046386718, 0.2277969970703125, 0.22800997924804686, 0.22908108520507814, 0.22838169860839844, 0.2285506591796875, 0.2286049346923828, 0.22830694580078126, 0.22801516723632811, 0.22815020751953125, 0.22767308044433593, 0.22776422119140624, 0.2276741180419922, 0.22829362487792967, 0.22767205810546876, 0.22779493713378907, 0.2277058563232422, 0.22765670776367186, 0.22813594055175782, 0.22820147705078125, 0.2283335723876953, 0.2275809326171875, 0.22812364196777343, 0.2283008575439453, 0.22861715698242188, 0.2281564178466797, 0.22812979125976562, 0.2289090576171875, 0.2288476104736328, 0.22834585571289062, 0.22857522583007814, 0.2290155487060547, 0.22790553283691406, 0.22831922912597657, 0.22821784973144532, 0.22858444213867188, 0.2286612548828125, 0.22839808654785157, 0.22902169799804686, 0.22916812133789063, 0.22946917724609375, 0.22901350402832032, 0.22913536071777343, 0.22909747314453124, 0.229138427734375, 0.22936679077148436, 0.22892851257324218, 0.22936268615722658, 0.22932582092285156, 0.22920909118652344, 0.22904730224609374, 0.2291425323486328, 0.22898892211914063, 0.22877183532714843, 0.2286202850341797, 0.229064697265625, 0.22876876831054688, 0.2292131805419922, 0.22902476501464844, 0.48329931640625, 0.2282854461669922, 0.228822021484375, 0.2279741516113281, 0.22834687805175782, 0.22893157958984375, 0.22888038635253907, 0.22889677429199218, 0.22806629943847656, 0.22842678833007812, 0.22791778564453125, 0.22817791748046876, 0.22819737243652344, 0.228274169921875, 0.22817791748046876, 0.2278656005859375, 0.22779391479492186, 0.22778982543945311, 0.22769561767578125, 0.22855783081054687, 0.2280273895263672, 0.22766490173339843, 0.2277734375, 0.2279331817626953, 0.22785125732421874, 0.2278594512939453, 0.227852294921875, 0.2281492462158203, 0.2276505584716797, 0.22739762878417968, 0.22808883666992188, 0.22794650268554686, 0.22817485046386718, 0.228094970703125, 0.22861415100097657, 0.22928793334960937, 0.22927769470214843, 0.22896333312988282, 0.22915072631835937, 0.22908006286621094, 0.22935040283203126, 0.22938829040527345, 0.22915583801269532, 0.22902578735351561, 0.22928282165527344, 0.22933914184570312, 0.23262924194335938, 0.22968319702148438, 0.22875648498535156, 0.22789529418945312, 0.2278666229248047, 0.2295767059326172, 0.22941900634765625, 0.22929306030273439, 0.22928282165527344, 0.22947532653808594, 0.2292490234375, 0.22923365783691407, 0.22946815490722655, 0.22925106811523438, 0.22942207336425782, 0.22925315856933592, 0.2293155517578125, 0.4786954345703125, 0.22789324951171874, 0.22778778076171874, 0.22853427124023437, 0.2276505584716797, 0.22818611145019532, 0.22808677673339844, 0.22816160583496095, 0.22942201232910156, 0.22774887084960938, 0.22915277099609374, 0.2292162628173828, 0.22927772521972656, 0.22797821044921876, 0.2286940155029297, 0.22868377685546876, 0.22756658935546875, 0.22776832580566406, 0.22760652160644532, 0.22787379455566406, 0.2284400634765625, 0.22826905822753907, 0.22876876831054688, 0.22959616088867188, 0.22788198852539063, 0.22754815673828124, 0.2290493469238281, 0.22905958557128905, 0.22827314758300782, 0.22785842895507813, 0.22869094848632812, 0.2282854461669922, 0.22885580444335937, 0.22893157958984375, 0.2289971160888672, 0.22875340270996095, 0.22852301025390626, 0.22836224365234375, 0.22938829040527345, 0.2279147491455078, 0.22865306091308593, 0.22844825744628905, 0.22864588928222657, 0.22837759399414062, 0.2290636749267578, 0.2289213409423828, 0.22904115295410157, 0.23109939575195312, 0.22880665588378907, 0.22881996154785156, 0.2284707794189453, 0.229232666015625, 0.22840829467773438, 0.2283520050048828, 0.23005081176757813, 0.22963821411132812, 0.22803654479980467, 0.22828851318359375, 0.22792909240722656, 0.22795878601074218, 0.22777548217773438, 0.22785331726074218, 0.22797004699707032, 0.47824075317382814, 0.22801408386230468, 0.22845132446289063, 0.22803660583496094, 0.22865408325195313, 0.22791679382324218, 0.22993101501464844, 0.22821376037597657, 0.22777548217773438, 0.2276874237060547, 0.22872781372070314, 0.22810418701171875, 0.22863462829589845, 0.2276290588378906, 0.22807244873046875, 0.2276546630859375, 0.22767514038085937, 0.22787481689453126, 0.22811648559570313, 0.22791270446777342, 0.22781336975097657, 0.22764236450195313, 0.2276822967529297, 0.22771916198730469, 0.22807347106933593, 0.22767718505859375, 0.22773248291015624, 0.2274396209716797, 0.2276177978515625, 0.2276126708984375, 0.22771405029296876, 0.2276884460449219, 0.22757273864746094, 0.22881689453125, 0.22877593994140624, 0.22816461181640624, 0.22797311401367187, 0.22859878540039064, 0.22825677490234375, 0.22843699645996093, 0.22875750732421876, 0.22935040283203126, 0.22820352172851563, 0.22832333374023436, 0.23120999145507812, 0.22912101745605468, 0.22907904052734376, 0.22970060729980468, 0.22873907470703125, 0.22823014831542968, 0.22959414672851564, 0.22853219604492186, 0.22784101867675782, 0.22805708312988282, 0.22899507141113282, 0.2288885803222656, 0.227852294921875, 0.2276669464111328, 0.2279219207763672, 0.22771916198730469, 0.22784921264648436, 0.2281492462158203, 0.22833561706542968, 0.47811685180664065, 0.22757478332519532, 0.2276433868408203, 0.227662841796875, 0.22771510314941407, 0.22770684814453124, 0.22767001342773438, 0.2279884796142578, 0.22770176696777344, 0.2278707275390625, 0.22784512329101564, 0.22757171630859374, 0.22767922973632812, 0.22853836059570312, 0.22817181396484376, 0.22842568969726562, 0.22801408386230468, 0.2303057861328125, 0.22843084716796874, 0.22846669006347656, 0.22829670715332032, 0.227631103515625, 0.22766796875, 0.22799462890625, 0.22806732177734376, 0.22791270446777342, 0.22781747436523436, 0.22777548217773438, 0.22781234741210937, 0.22849331665039063, 0.22844009399414061, 0.22815536499023437, 0.22805509948730468, 0.22854751586914063, 0.22936679077148436, 0.2276741180419922, 0.22772940063476563, 0.22759014892578125, 0.22804173278808593, 0.22774169921875, 0.22859373474121095, 0.22844819641113281, 0.22795266723632812, 0.22765052795410157, 0.22817893981933593, 0.22788096618652343, 0.2275768280029297, 0.227915771484375, 0.22789631652832032, 0.22774783325195314, 0.22783692932128907, 0.22787890625, 0.22841958618164063, 0.22780006408691406, 0.22799154663085938, 0.2284862060546875, 0.22839187622070312, 0.23194111633300782, 0.22911180114746094, 0.22951731872558595, 0.22850355529785157, 0.22849740600585938, 0.22812570190429687, 0.47752294921875, 0.2276556854248047, 0.22764544677734375, 0.22774169921875, 0.22828953552246095, 0.22794444274902342, 0.22778880310058594, 0.2285332489013672, 0.22908927917480468, 0.2287073211669922, 0.2291087341308594, 0.22870118713378906, 0.22815948486328125, 0.22893466186523437, 0.2281328582763672, 0.22798233032226561, 0.22857215881347656, 0.22788812255859375, 0.2276177978515625, 0.22820658874511718, 0.22790963745117188, 0.2278912353515625, 0.22779286193847656, 0.22789631652832032, 0.2286622772216797, 0.2280079345703125, 0.22793728637695312, 0.22807142639160155, 0.22804888916015625, 0.2280335388183594, 0.22812159729003906, 0.2278830108642578, 0.2276884460449219, 0.22792807006835938, 0.22778675842285157, 0.22810009765625, 0.22793215942382813, 0.22936473083496095, 0.23167283630371094, 0.22877081298828125, 0.22863258361816408, 0.22877593994140624, 0.22783282470703126, 0.2277181396484375, 0.22791270446777342, 0.22783183288574219, 0.22780720520019532, 0.22802841186523437, 0.2279536590576172, 0.2279619140625, 0.22795872497558595, 0.22792909240722656, 0.22810009765625, 0.22938214111328126, 0.22828031921386718, 0.228242431640625, 0.2287636413574219, 0.22942311096191406, 0.22908114624023437, 0.22857618713378905, 0.22855372619628905, 0.22922035217285155, 0.2285117492675781, 0.4789893188476563, 0.22813900756835936, 0.22828134155273438, 0.22873292541503906, 0.2284707794189453, 0.22896640014648437, 0.22858956909179687, 0.2285096893310547, 0.22846669006347656, 0.22824858093261718, 0.2286878662109375, 0.2290083770751953, 0.22912716674804687, 0.22828851318359375, 0.22853836059570312, 0.228384765625, 0.22901145935058595, 0.2284523468017578, 0.22929306030273439, 0.22884352111816406, 0.2288916473388672, 0.22799874877929688, 0.22763005065917968, 0.2294476776123047, 0.22789324951171874, 0.22849740600585938, 0.2280437774658203, 0.22870835876464843, 0.228890625, 0.228600830078125, 0.22893466186523437, 0.22901248168945312, 0.22887936401367187, 0.22905445861816406, 0.2291025848388672, 0.22861824035644532, 0.2289040069580078, 0.22949267578125, 0.22833255004882813, 0.22794650268554686, 0.2277928924560547, 0.2285117492675781, 0.2278656005859375, 0.22826905822753907, 0.22805914306640626, 0.22932992553710937, 0.2291025848388672, 0.2288046112060547, 0.22843597412109376, 0.22823526000976563, 0.2282977294921875, 0.22911488342285155, 0.22893772888183594, 0.2283100128173828, 0.22832127380371095, 0.22796287536621093, 0.22796389770507813, 0.22786151123046874, 0.22802024841308594, 0.22792703247070312, 0.2290247344970703, 0.2290872344970703, 0.2293729248046875, 0.48112127685546874, 0.22774476623535156, 0.22767205810546876, 0.22760140991210936, 0.22770790100097657, 0.228706298828125, 0.22839602661132813, 0.22828953552246095, 0.22751437377929687, 0.22827008056640624, 0.22855577087402343, 0.22760858154296876, 0.22853733825683595, 0.2290882568359375, 0.22853427124023437, 0.22787174987792969, 0.22959922790527343, 0.22843597412109376, 0.22793830871582033, 0.22853427124023437, 0.22780621337890625, 0.22782566833496093, 0.2280755157470703, 0.22808781433105468, 0.22815948486328125, 0.22787686157226564, 0.22864691162109374, 0.2278481903076172, 0.22766387939453125, 0.22787174987792969, 0.22767514038085937, 0.22795161437988282, 0.22794650268554686, 0.22880563354492187, 0.22775808715820312, 0.22767514038085937, 0.22767718505859375, 0.2277734375, 0.22772122192382813, 0.22761984252929687, 0.2289653778076172, 0.22932582092285156, 0.2285506591796875, 0.22796083068847656, 0.2276986846923828, 0.22787583923339844, 0.2282608642578125, 0.2280263671875, 0.22794137573242187, 0.2278461456298828, 0.22779391479492186, 0.22773248291015624, 0.2276259765625, 0.22793215942382813, 0.22887014770507813, 0.2283756103515625, 0.22759724426269531, 0.22863462829589845, 0.22802841186523437, 0.228279296875, 0.22952755737304686, 0.22824755859375, 0.2283223114013672, 0.4810905456542969, 0.22825372314453124, 0.22767202758789062, 0.22804582214355468, 0.22862745666503906, 0.227810302734375, 0.22774989318847655, 0.22795980834960938, 0.22772940063476563, 0.22866943359375, 0.22906982421875, 0.2284390411376953, 0.22901657104492187, 0.22875852966308594, 0.22855270385742188, 0.22873805236816405, 0.22831513977050782, 0.22964530944824219, 0.2282782745361328, 0.228890625, 0.22893772888183594, 0.22941389465332032, 0.22863871765136717, 0.22856192016601562, 0.2286868438720703, 0.22785536193847655, 0.22773554992675782, 0.22780723571777345, 0.22774887084960938, 0.22783795166015625, 0.2286028747558594, 0.2292316131591797, 0.22855474853515625, 0.2284277801513672, 0.22815129089355468, 0.22762086486816407, 0.22760345458984374, 0.2294599609375, 0.228251708984375, 0.22796998596191406, 0.22916300964355468, 0.22799667358398437, 0.22786355590820312, 0.2276986846923828, 0.22783282470703126, 0.22780621337890625, 0.22774681091308593, 0.22811955261230468, 0.22787890625, 0.22774374389648439, 0.2275000305175781, 0.22767001342773438, 0.2276884460449219, 0.22773554992675782, 0.2278778839111328, 0.22794650268554686, 0.22774476623535156, 0.22773452758789062, 0.22781234741210937, 0.22782975769042968, 0.22782054138183594, 0.22786151123046874, 0.22886604309082031, 0.4816783447265625, 0.22901145935058595, 0.22866943359375, 0.2290145263671875, 0.22938316345214843, 0.22859571838378906, 0.22785023498535156, 0.22841139221191406, 0.22864895629882812, 0.2287615966796875, 0.22899302673339844, 0.2279720916748047, 0.2287073211669922, 0.22850355529785157, 0.22853631591796875, 0.22866021728515626, 0.22911077880859376, 0.22943026733398436, 0.22803558349609376, 0.22808781433105468, 0.2292725830078125, 0.22914457702636717, 0.22839295959472655, 0.22797314453125, 0.22791267395019532, 0.22795161437988282, 0.22988902282714843, 0.2280447998046875, 0.22778163146972658, 0.2276444091796875, 0.2277928924560547, 0.2285015106201172, 0.22870835876464843, 0.22830181884765624, 0.22769766235351563, 0.22775193786621092, 0.22775091552734375, 0.22768333435058594, 0.22769664001464843, 0.22772633361816405, 0.22788607788085938, 0.22782054138183594, 0.22759837341308595, 0.22872572326660157, 0.2291568603515625, 0.22867762756347657, 0.228094970703125, 0.2280990753173828, 0.22783795166015625, 0.2277724151611328, 0.22763827514648438, 0.2278164520263672, 0.22774681091308593, 0.2276433868408203, 0.2277181396484375, 0.22805708312988282, 0.2277232666015625, 0.228068359375, 0.2278154296875, 0.2276864013671875, 0.22811546325683593, 0.22927462768554688, 0.2289459228515625]",tokens/s,4.311005041891343,,,,, -4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciCoder-1b,Deci/DeciCoder-1b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,TencentARC/Mistral_Pro_8B_v0.1,TencentARC/Mistral_Pro_8B_v0.1,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,2452.258816,7298.613248,0.0,6652.166144,6323.221504,s,10,7.6798339843750005,0.7679833984375,0.0025974055850476797,0.7679208068847656,0.7711121887207032,0.7721169158935547,0.772920697631836,"[0.7685753173828125, 0.7731216430664063, 0.7647974853515624, 0.7675167846679688, 0.76627783203125, 0.7641251220703125, 0.7666953735351563, 0.7683248291015625, 0.770888916015625, 0.7695106811523438]",tokens/s,333.34053902837564,kWh,9.039419593792114e-06,4.95321018510627e-06,4.32610068310048e-05,5.725363660990319e-05,tokens/kWh,4471331.694513176,MB,2452.258816,7298.613248,0.0,6652.166144,6382.564864,s,10,455.74574609375,45.57457460937499,0.015546779743152406,45.57488671875,45.595179296874996,45.5973923828125,45.5991628515625,"[45.5867109375, 45.5946875, 45.59960546875, 45.58337109375, 45.55990625, 45.5529375, 45.558640625, 45.56011328125, 45.5776015625, 45.572171875]",tokens/s,1.3823497101175461,kWh,0.000537889561981909,0.0002948106017116671,0.0025152069288305857,0.0033479070925241613,tokens/kWh,18817.726495660016,,s,629,461.93301361083974,0.7343927084433065,0.09125959398126374,0.7233239135742188,0.7243169799804687,0.7247333496093751,1.4901356689453125,"[0.7238604736328125, 0.7241830444335937, 0.7225497436523437, 0.7234365234375, 0.7223142700195313, 0.7223203735351562, 0.7245137939453125, 0.7236771850585938, 0.7226992797851562, 0.723535888671875, 0.7234600830078125, 0.7246929931640625, 0.7229706420898437, 0.7237723999023438, 0.7233392944335938, 0.7235133666992187, 0.7225231323242187, 0.7234631958007812, 0.723367919921875, 0.7226572875976562, 0.7243929443359375, 0.7232921752929687, 0.7236546630859375, 0.7240488891601562, 0.7235286865234375, 0.7245383911132812, 0.7233167114257812, 0.7230576782226562, 0.7237959594726563, 0.7243140869140625, 0.722629638671875, 0.7226122436523438, 0.7224381713867187, 0.7233873901367187, 0.7249028930664062, 0.7240294189453125, 0.723904541015625, 0.7241912231445312, 0.7243069458007813, 0.7245864868164062, 0.7234006958007813, 0.723114990234375, 0.7231324462890625, 0.7226593017578125, 0.7241062622070312, 0.7237662963867187, 0.7239597778320312, 0.724094970703125, 0.7240611572265625, 0.7237017822265625, 0.7238123779296874, 0.72437451171875, 0.7245363159179687, 0.7248076782226562, 0.7224637451171875, 0.7232000122070312, 0.7239700317382812, 0.7240325317382813, 0.7229603881835938, 0.7228866577148437, 0.7229910888671875, 0.7245496215820313, 1.497416748046875, 0.7238062133789063, 0.7237662963867187, 0.7238911743164063, 0.7240550537109375, 0.7237857055664062, 0.7238358764648437, 0.7228784790039062, 0.72378369140625, 0.7233925170898438, 0.7229235229492188, 0.722534423828125, 0.7236341552734376, 0.723399658203125, 0.7241830444335937, 0.7228897094726563, 0.7232071533203125, 0.7245209350585937, 0.7240325317382813, 0.7224688720703125, 0.7228262329101562, 0.7225323486328125, 0.7236700439453125, 0.7239026489257813, 0.72288037109375, 0.7225128784179687, 0.7245783081054687, 0.7235983276367187, 0.7235532836914063, 0.725017578125, 0.724094970703125, 0.7250708618164062, 0.7244400634765625, 0.7243571166992188, 0.7240745239257812, 0.7232041015625, 0.7236218872070312, 0.7237877807617188, 0.7239588012695313, 0.7241809692382812, 0.7240929565429688, 0.7230279541015625, 0.7229634399414062, 0.7229419555664063, 0.72318359375, 0.7240601806640625, 0.7232184448242187, 0.7226746826171875, 0.7235880737304687, 0.724453369140625, 0.7243407592773438, 0.7238184814453125, 0.7240242919921875, 0.7247308959960937, 0.7247349853515626, 0.7235543212890625, 0.7233402709960938, 0.7242977294921875, 0.7243202514648438, 0.723862548828125, 0.724084716796875, 0.723768310546875, 0.7227473754882813, 1.490913330078125, 0.7229655151367187, 0.7226480712890625, 0.7240775756835938, 0.7235686645507813, 0.7231815795898437, 0.72302490234375, 0.7227381591796875, 0.7236761474609374, 0.724653076171875, 0.724200439453125, 0.7239618530273437, 0.7246776123046875, 0.7250585327148438, 0.7243140869140625, 0.7238225708007813, 0.7248977661132813, 0.725359619140625, 0.7238379516601563, 0.7237171020507812, 0.724158447265625, 0.7239188232421875, 0.7248445434570312, 0.7251189575195313, 0.7247544555664063, 0.7250022583007812, 0.7248936767578125, 0.7236198120117188, 0.7234979858398437, 0.723462158203125, 0.7234559936523437, 0.724263916015625, 0.7227955322265625, 0.7225702514648438, 0.7241932983398438, 0.7234805908203125, 0.7234949340820312, 0.72275146484375, 0.7225385131835937, 0.7228804931640626, 0.7242403564453125, 0.7236218872070312, 0.7236986694335937, 0.723472412109375, 0.724126708984375, 0.7239505615234375, 0.7238901977539063, 0.7237283935546875, 0.7242025146484375, 0.724116455078125, 0.7241031494140625, 0.72374169921875, 0.7228671875, 0.7233228759765625, 0.7237857055664062, 0.7240171508789063, 0.72399462890625, 0.7229276123046875, 0.7241410522460937, 0.7236751098632812, 0.7226521606445313, 0.7233648681640625, 0.7235952758789063, 1.49184814453125, 0.7231047973632813, 0.7224780883789063, 0.7224258422851563, 0.72266650390625, 0.7234283447265625, 0.72353076171875, 0.7227269287109375, 0.7243991088867188, 0.7234119873046875, 0.72285595703125, 0.7229951782226562, 0.7234692993164062, 0.7242373046875, 0.7239915771484375, 0.7242465209960938, 0.7243253784179687, 0.7225200805664063, 0.7226972045898438, 0.7227289428710938, 0.7236956176757813, 0.7238881225585938, 0.7240274047851563, 0.7241492309570312, 0.7236741333007812, 0.7231661987304687, 0.723114990234375, 0.7243571166992188, 0.7238656005859375, 0.723240966796875, 0.7226122436523438, 0.7238615112304687, 0.7242168579101562, 0.7235440673828125, 0.7236239624023437, 0.723926025390625, 0.7243919067382812, 0.7229173583984375, 0.7226542358398438, 0.7227238159179687, 0.7229859619140625, 0.7234990234375, 0.7239567260742188, 0.7231416015625, 0.7239454956054687, 0.7230084838867188, 0.7233607788085937, 0.7238359375, 0.7240099487304688, 0.7241431274414063, 0.7246141357421875, 0.7241359252929688, 0.7245772705078125, 0.7245137939453125, 0.725411865234375, 0.723578857421875, 0.7241543579101563, 0.7230054321289062, 0.7233526000976562, 0.7226992797851562, 0.7232420043945312, 0.7230075073242187, 0.7230187377929688, 1.490323486328125, 0.7229685668945313, 0.723535888671875, 0.72338330078125, 0.723409912109375, 0.7233341674804687, 0.7232112426757813, 0.72279345703125, 0.7230853271484375, 0.7230105590820313, 0.7226060791015625, 0.723346435546875, 0.7229296875, 0.72382568359375, 0.7241666259765625, 0.72303515625, 0.7230812377929687, 0.7226583251953125, 0.7230535888671875, 0.7229890747070312, 0.7228118896484375, 0.7222753295898438, 0.7228641357421876, 0.722740234375, 0.72254052734375, 0.7228671875, 0.7227391967773438, 0.7233505249023438, 0.722619384765625, 0.72300439453125, 0.7234150390625, 0.7230556030273437, 0.7235297241210937, 0.722819091796875, 0.723610595703125, 0.7228836059570313, 0.7235348510742188, 0.7226911010742187, 0.7227914428710938, 0.7233935546875, 0.7231590576171875, 0.7238717651367188, 0.72331982421875, 0.7227811889648438, 0.7233136596679688, 0.7229542236328125, 0.7231754150390625, 0.7233607788085937, 0.7233925170898438, 0.7229081420898438, 0.7231047973632813, 0.7224483642578124, 0.723252197265625, 0.722640869140625, 0.7238778686523437, 0.7232604370117187, 0.7231641845703125, 0.7244615478515625, 0.7236874389648438, 0.72321533203125, 0.7241400146484375, 0.7235686645507813, 0.7226695556640625, 1.48693603515625, 0.7224832153320313, 0.7234774780273437, 0.7231426391601562, 0.724173828125, 0.7240755004882813, 0.7229542236328125, 0.72241357421875, 0.7228876953125, 0.7233106079101562, 0.7236618041992188, 0.7231641845703125, 0.7225897216796875, 0.7224903564453125, 0.7225599975585938, 0.7225702514648438, 0.7227361450195312, 0.7226112060546875, 0.7231498413085937, 0.723304443359375, 0.722703369140625, 0.7227955322265625, 0.7233474731445313, 0.7234703369140625, 0.7267368774414062, 0.7227781982421875, 0.7225128173828125, 0.7224033203125, 0.7225589599609376, 0.7226634521484375, 0.7223306274414063, 0.7225364990234375, 0.7230811767578125, 0.72255078125, 0.7227003173828125, 0.7224832153320313, 0.7223961791992187, 0.7230310668945312, 0.7257907104492187, 0.7230904541015625, 0.7228538818359375, 0.7229788208007812, 0.7225180053710938, 0.7223818359375, 0.7229481201171875, 0.7229378662109375, 0.72367822265625, 0.7230084838867188, 0.7230146484375, 0.7226326904296875, 0.7231416015625, 0.7237335205078125, 0.7241860961914063, 0.7245127563476562, 0.722572265625, 0.7229020385742188, 0.72300341796875, 0.72266650390625, 0.7228159790039063, 0.7224207153320312, 0.7239270629882812, 0.72285693359375, 0.722966552734375, 1.4896527099609376, 0.7232348022460937, 0.7239567260742188, 0.7227739868164063, 0.723078125, 0.7226429443359375, 0.7226798095703125, 0.7228703002929687, 0.7228917846679688, 0.7229112548828125, 0.722935791015625, 0.7229030151367187, 0.7226050415039063, 0.7226798095703125, 0.7230105590820313, 0.7234058227539063, 0.7233239135742188, 0.7225077514648437, 0.7225938110351563, 0.7227412719726563, 0.7225303344726562, 0.72310888671875, 0.7248046264648438, 0.7248180541992187, 0.723907470703125, 0.72346826171875, 0.7232081909179687, 0.7231743774414062, 0.7237877807617188, 0.7234160766601563, 0.7235860595703125, 0.7227647705078125, 0.7225620727539063, 0.7228334350585938, 0.7226992797851562, 0.7227996215820313, 0.7227381591796875, 0.7233106079101562, 0.723267578125, 0.7230556030273437, 0.7226695556640625, 0.723061767578125, 0.7235317993164062, 0.7233689575195312, 0.7238010864257812, 0.7234078979492188, 0.7232849731445312, 0.7231876831054688, 0.7232327880859375, 0.7230422973632813, 0.7230341186523438, 0.7235563354492187, 0.7233157348632813, 0.7235266723632813, 0.723040283203125, 0.7234692993164062, 0.7238154296875, 0.7239547119140625, 0.7235758056640625, 0.7227924194335937, 0.7224013061523438, 0.7225753784179687, 0.7224832153320313, 1.4909122314453125, 0.7237437744140625, 0.7234396362304687, 0.7226849365234375, 0.7229102172851563, 0.7227125854492188, 0.7236638793945313, 0.7237918701171875, 0.722882568359375, 0.7225845947265624, 0.7231959228515625, 0.7231826171875, 0.722893798828125, 0.7226316528320312, 0.7233382568359376, 0.7234171142578125, 0.7229818725585937, 0.722428955078125, 0.7228836059570313, 0.722682861328125, 0.7232973022460938, 0.7227125854492188, 0.7229685668945313, 0.723040283203125, 0.7242997436523437, 0.7234037475585937, 0.7232604370117187, 0.7234345092773438, 0.72426904296875, 0.7237816162109375, 0.7232747802734375, 0.7224575805664063, 0.7224043579101562, 0.7228845825195312, 0.7226439819335938, 0.7229317016601563, 0.722651123046875, 0.7223142700195313, 0.7236024169921875, 0.7237283935546875, 0.7242454833984375, 0.72339453125, 0.7232337646484375, 0.7232767944335937, 0.7239905395507813, 0.7230320434570312, 0.7228037109375, 0.7227996215820313, 0.722904052734375, 0.7226705932617188, 0.7241994018554687, 0.7242670288085937, 0.7234703369140625, 0.723314697265625, 0.7229102172851563, 0.722967529296875, 0.7240089721679688, 0.723979248046875, 0.7236167602539062, 0.7229450073242187, 0.7226654663085937, 0.7226122436523438, 0.722820068359375, 1.495736328125, 0.724738037109375, 0.7240570678710937, 0.72363623046875, 0.72394140625, 0.7237017822265625, 0.7253534545898438, 0.7251988525390625, 0.7241973876953125, 0.7238829956054688, 0.7238615112304687, 0.7239464721679687, 0.7240714111328125, 0.7241809692382812, 0.7237929077148437, 0.72289892578125, 0.7232706298828125, 0.722914306640625, 0.7235338134765625, 0.7231948852539063, 0.7239393310546876, 0.7230525512695313, 0.7231815795898437, 0.7229603881835938, 0.7230853271484375, 0.72344677734375, 0.7233054809570313, 0.72283544921875, 0.7226583251953125, 0.7230003051757813, 0.72287744140625, 0.722608154296875, 0.7231324462890625, 0.7233526000976562, 0.7234529418945312, 0.72300439453125, 0.7229020385742188, 0.7233484497070313, 0.723610595703125, 0.7233925170898438, 0.72308837890625, 0.7229317016601563, 0.7229389038085937, 0.7231682739257812, 0.7229910888671875, 0.7232286987304688, 0.72271875, 0.723019775390625, 0.723493896484375, 0.72239208984375, 0.722845703125, 0.7230084838867188, 0.7235686645507813, 0.7236188354492188, 0.7237345581054687, 0.72363623046875, 0.7234805908203125, 0.7234385986328125, 0.7238164672851563, 0.724041748046875, 0.723356689453125, 0.723030029296875, 0.7237140502929688, 1.4925987548828126, 0.7230996704101562, 0.72384716796875, 0.7239147338867188, 0.7256944580078125, 0.7247103881835938, 0.7244021606445312, 0.7243776245117187, 0.724316162109375, 0.7242546997070313, 0.7244994506835938, 0.724832275390625, 0.7239239501953125, 0.723041259765625, 0.7240745239257812, 0.7239096069335937, 0.723515380859375, 0.7226603393554687, 0.7232348022460937, 0.7224627075195312, 0.7226132202148438, 0.7228047485351563, 0.72265625, 0.7226470947265625, 0.7226756591796875, 0.7231713256835938, 0.72317236328125, 0.722787353515625, 0.7223971557617187, 0.7228057861328125, 0.7233065185546875, 0.7235245971679688, 0.7238748168945313, 0.72374169921875, 0.7234816284179687, 0.7235297241210937, 0.7231426391601562, 0.7237867431640626, 0.7235635375976562, 0.7232808837890625, 0.72345703125, 0.7230924682617188, 0.72296142578125, 0.7232747802734375, 0.7239485473632813, 0.724284423828125, 0.722783203125, 0.7224873046875, 0.7223971557617187, 0.7222200927734375, 0.7224954833984375, 0.7242076416015625, 0.722787353515625, 0.7238604736328125, 0.7227545776367188, 0.7224268798828125, 0.7232020263671874, 0.7240253295898438, 0.722998291015625, 0.7229900512695312, 0.72289892578125, 0.7227340698242187, 0.7230996704101562]",tokens/s,1.36166929287697,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-564M,facebook/xglm-564M,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 64, in run - report = scenario.run(backend) - File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run - _ = backend.generate(self.inputs, self.config.generate_kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate - return self.pretrained_model.generate(**inputs, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1576, in generate - result = self._greedy_search( - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2494, in _greedy_search - outputs = self( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1211, in forward - outputs = self.model( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1018, in forward - layer_outputs = decoder_layer( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 741, in forward - hidden_states, self_attn_weights, present_key_value = self.self_attn( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) -TypeError: DeciCoderAttention.forward() got an unexpected keyword argument 'cache_position' + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: XGLMForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciLM-7B,Deci/DeciLM-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-350m,facebook/opt-350m,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 64, in run - report = scenario.run(backend) - File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run - _ = backend.generate(self.inputs, self.config.generate_kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate - return self.pretrained_model.generate(**inputs, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1576, in generate - result = self._greedy_search( - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2494, in _greedy_search - outputs = self( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/root/.cache/huggingface/modules/transformers_modules/Deci/DeciLM-7B/c3c9f4226801dc0433f32aebffe0aac68ee2f051/transformers_v4_35_2__modeling_llama.py"", line 1034, in forward - outputs = self.model( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/root/.cache/huggingface/modules/transformers_modules/Deci/DeciLM-7B/c3c9f4226801dc0433f32aebffe0aac68ee2f051/modeling_decilm.py"", line 274, in forward - layer_outputs = decoder_layer( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/root/.cache/huggingface/modules/transformers_modules/Deci/DeciLM-7B/c3c9f4226801dc0433f32aebffe0aac68ee2f051/transformers_v4_35_2__modeling_llama.py"", line 672, in forward - hidden_states, self_attn_weights, present_key_value = self.self_attn( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/root/.cache/huggingface/modules/transformers_modules/Deci/DeciLM-7B/c3c9f4226801dc0433f32aebffe0aac68ee2f051/modeling_decilm.py"", line 84, in forward - query_states = self.q_proj(hidden_states) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward - assert AWQ_INSTALLED, ( -AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: OPTForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-j-6b,EleutherAI/gpt-j-6b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-6.7b,facebook/opt-6.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 64, in run - report = scenario.run(backend) - File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run - _ = backend.generate(self.inputs, self.config.generate_kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate - return self.pretrained_model.generate(**inputs, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1576, in generate - result = self._greedy_search( - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2494, in _greedy_search - outputs = self( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gptj/modeling_gptj.py"", line 1127, in forward - transformer_outputs = self.transformer( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gptj/modeling_gptj.py"", line 953, in forward - outputs = block( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gptj/modeling_gptj.py"", line 581, in forward - attn_outputs = self.attn( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gptj/modeling_gptj.py"", line 320, in forward - query = self.q_proj(hidden_states) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward - assert AWQ_INSTALLED, ( -AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: OPTForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-1.3B,EleutherAI/gpt-neo-1.3B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-72B,Qwen/Qwen-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 64, in run - report = scenario.run(backend) - File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run - _ = backend.generate(self.inputs, self.config.generate_kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate - return self.pretrained_model.generate(**inputs, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1576, in generate - result = self._greedy_search( - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2494, in _greedy_search - outputs = self( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 974, in forward - transformer_outputs = self.transformer( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 842, in forward - outputs = block( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 569, in forward - attn_outputs = self.attn( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 521, in forward - return self.attention( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 322, in forward - query = self.q_proj(hidden_states) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward - assert AWQ_INSTALLED, ( -AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 550, in from_pretrained + model_class = get_class_from_dynamic_module( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 498, in get_class_from_dynamic_module + final_module = get_cached_module_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 323, in get_cached_module_file + modules_needed = check_imports(resolved_module_file) + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 181, in check_imports + raise ImportError( +ImportError: This modeling file requires the following packages that were not found in your environment: transformers_stream_generator. Run `pip install transformers_stream_generator` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-34B,01-ai/Yi-34B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-6B,,cuda,0,42,,,,,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,,MB,1571.749888,5448.925184,0.0,4802.47808,4489.12128,s,10,5.0903373718261715,0.5090337371826171,0.0024963690031113454,0.5090882415771485,0.5114007537841796,0.5133009628295898,0.514821130065918,"[0.515201171875, 0.5058653564453125, 0.5075380249023438, 0.5109784851074218, 0.5074734191894531, 0.5066469421386719, 0.5090344543457032, 0.5092752685546875, 0.5091822204589844, 0.5091420288085937]",tokens/s,502.91362104386286,kWh,5.98198832737075e-06,3.277856253707796e-06,2.697877158300277e-05,3.623861616408132e-05,tokens/kWh,7064287.412104325,MB,1571.749888,5448.925184,0.0,4802.47808,4557.793792,s,10,299.8803828125,29.98803828125,0.03808075694133287,29.9707392578125,30.046216015625,30.058939062500002,30.0691175,"[29.972552734375, 30.043388671875, 30.071662109375, 30.010330078125, 29.96, 29.95865625, 29.96925, 29.972228515625, 29.963451171875, 29.95886328125]",tokens/s,2.1008376543053404,kWh,0.00035374844759702685,0.00019388459414373753,0.001556149911585597,0.0021037829533263617,tokens/kWh,29946.05498651303,,s,629,303.97862612915014,0.48327285553124066,0.06045359810307915,0.47556915283203127,0.47775867309570313,0.47801220703124997,0.9833881201171876,"[0.4750715026855469, 0.4762265625, 0.4773417053222656, 0.47625729370117187, 0.47571148681640624, 0.47557223510742186, 0.47629312133789065, 0.47471615600585937, 0.475789306640625, 0.47461376953125, 0.4746967163085937, 0.474913818359375, 0.47592657470703126, 0.47551071166992187, 0.47595623779296875, 0.47574118041992186, 0.4750817260742187, 0.47586611938476564, 0.4771154479980469, 0.47594082641601565, 0.4757462768554688, 0.47569818115234375, 0.4764293212890625, 0.475978759765625, 0.47632794189453126, 0.47545343017578123, 0.47498138427734377, 0.47522305297851564, 0.47550360107421874, 0.47533575439453124, 0.4750652770996094, 0.4751790161132812, 0.47525888061523436, 0.4754544677734375, 0.4752803955078125, 0.4756971435546875, 0.4763504638671875, 0.4758845520019531, 0.47809332275390626, 0.475462646484375, 0.47741644287109375, 0.47680307006835937, 0.47630642700195314, 0.47653070068359377, 0.47590911865234375, 0.4751380615234375, 0.47556710815429687, 0.475146240234375, 0.47564901733398435, 0.4757647399902344, 0.47530291748046877, 0.4750807189941406, 0.47518719482421873, 0.47629925537109374, 0.4754646911621094, 0.4754155578613281, 0.4753121337890625, 0.4750469055175781, 0.47505612182617185, 0.4759122009277344, 0.47506842041015623, 0.4758763427734375, 0.9829324951171875, 0.4772618103027344, 0.47679385375976563, 0.4752302551269531, 0.47515850830078127, 0.4750141296386719, 0.4753879089355469, 0.47504486083984376, 0.47488204956054686, 0.47477862548828126, 0.4748114013671875, 0.47600946044921877, 0.4750940246582031, 0.4750837707519531, 0.4750745544433594, 0.4750592041015625, 0.47484109497070315, 0.4761292724609375, 0.4751933288574219, 0.4750182495117187, 0.475357177734375, 0.47510528564453125, 0.4762142639160156, 0.47762432861328125, 0.4771512451171875, 0.478097412109375, 0.47743179321289064, 0.4781097106933594, 0.4774246520996094, 0.47769497680664064, 0.4778270568847656, 0.4778741760253906, 0.47795404052734375, 0.47758233642578124, 0.47754238891601564, 0.47758438110351564, 0.4775854187011719, 0.4776581115722656, 0.4781363220214844, 0.47784140014648435, 0.4785827941894531, 0.4778670043945312, 0.478013427734375, 0.4813271179199219, 0.4787435607910156, 0.4778823547363281, 0.4775577697753906, 0.47766937255859376, 0.4776509399414062, 0.47802163696289063, 0.4775925903320313, 0.4771829833984375, 0.47707955932617185, 0.4779130859375, 0.47752191162109375, 0.4774225769042969, 0.47764581298828124, 0.47809228515625, 0.47725054931640626, 0.47956683349609375, 0.47532647705078124, 0.47634738159179685, 0.4757237854003906, 0.9846497192382813, 0.4779151306152344, 0.47775845336914063, 0.4778516540527344, 0.4773468017578125, 0.47776461791992186, 0.47770932006835937, 0.4769669189453125, 0.4777676696777344, 0.4777635803222656, 0.47618765258789064, 0.4749752197265625, 0.4755538024902344, 0.4747980651855469, 0.4747796630859375, 0.4748114013671875, 0.47680514526367185, 0.4753622741699219, 0.47532440185546876, 0.4750438537597656, 0.4750325622558594, 0.47742669677734373, 0.4790394897460937, 0.47991500854492186, 0.4776028137207031, 0.4770672607421875, 0.47781991577148436, 0.4773304443359375, 0.4778946533203125, 0.4778496398925781, 0.4774788818359375, 0.4779632568359375, 0.47771136474609377, 0.47775640869140623, 0.47768267822265625, 0.47796121215820314, 0.4775454711914062, 0.4780103759765625, 0.4779438171386719, 0.4773253173828125, 0.4780001220703125, 0.47749530029296877, 0.478482421875, 0.4767160339355469, 0.47870156860351565, 0.47761203002929686, 0.47860427856445314, 0.47783221435546874, 0.4779366149902344, 0.4776212463378906, 0.4775301208496094, 0.477328369140625, 0.478376953125, 0.4772362365722656, 0.47751168823242185, 0.47775955200195314, 0.47782699584960936, 0.47732632446289064, 0.47806362915039063, 0.4772454528808594, 0.4775782775878906, 0.47722698974609373, 0.477765625, 0.9892290649414063, 0.47758950805664063, 0.47766937255859376, 0.47711026000976564, 0.4785479736328125, 0.4779674072265625, 0.47848751831054687, 0.4773918762207031, 0.47714407348632815, 0.475968505859375, 0.47733248901367187, 0.4777830505371094, 0.4774912109375, 0.4774410095214844, 0.4776990661621094, 0.47711026000976564, 0.47647540283203127, 0.4762900390625, 0.47571044921875, 0.47531622314453126, 0.47745126342773436, 0.4778526611328125, 0.4754595947265625, 0.4753295593261719, 0.47519024658203124, 0.47728536987304687, 0.4776847229003906, 0.47796734619140624, 0.47679693603515627, 0.4769525756835937, 0.47623678588867185, 0.47598489379882813, 0.47588760375976563, 0.47573504638671876, 0.4752127990722656, 0.47496600341796874, 0.4763709411621094, 0.47549542236328124, 0.4751493225097656, 0.4747591552734375, 0.47531109619140627, 0.47625933837890627, 0.47604122924804687, 0.4752691345214844, 0.478866455078125, 0.4751441650390625, 0.4752322692871094, 0.47499981689453125, 0.47643954467773436, 0.4755630187988281, 0.4757166137695312, 0.47565005493164064, 0.476015625, 0.4754155578613281, 0.4754124755859375, 0.47525070190429686, 0.47504281616210936, 0.4759347229003906, 0.47554763793945315, 0.47518923950195313, 0.47525274658203126, 0.475404296875, 0.47679489135742187, 0.9835653076171875, 0.4750100708007812, 0.47497518920898435, 0.47521792602539065, 0.47529168701171876, 0.47549435424804687, 0.47515341186523435, 0.47609548950195313, 0.47539712524414063, 0.47541351318359376, 0.47511859130859374, 0.47484622192382814, 0.47537664794921874, 0.475104248046875, 0.47634738159179685, 0.4752896118164063, 0.47548724365234374, 0.4761077880859375, 0.4751452026367188, 0.4771768188476562, 0.4755333251953125, 0.4755230712890625, 0.4750796813964844, 0.4751923217773438, 0.4752496643066406, 0.4748308410644531, 0.47501516723632814, 0.47498751831054686, 0.4759531555175781, 0.4754883117675781, 0.47538067626953123, 0.4753909912109375, 0.47539712524414063, 0.4753735656738281, 0.475494384765625, 0.47589376831054686, 0.4761006164550781, 0.47518206787109374, 0.4750602111816406, 0.47489434814453124, 0.47634228515625, 0.47604940795898437, 0.47567974853515627, 0.47519744873046876, 0.4751749267578125, 0.47551077270507813, 0.4754002685546875, 0.47505914306640623, 0.47493426513671877, 0.475788330078125, 0.4752465515136719, 0.47501516723632814, 0.47556915283203127, 0.4761507873535156, 0.47565823364257814, 0.47601458740234376, 0.47671194458007815, 0.478308349609375, 0.4755906677246094, 0.47542578125, 0.4754708557128906, 0.4771061706542969, 0.4754810791015625, 0.9827368774414063, 0.47548416137695315, 0.4752916564941406, 0.4750274658203125, 0.4750274658203125, 0.47477658081054686, 0.4755199890136719, 0.47510833740234376, 0.4752332763671875, 0.47511962890625, 0.4750837707519531, 0.47506329345703124, 0.4751718444824219, 0.477048828125, 0.4756756591796875, 0.475315185546875, 0.47508685302734377, 0.474925048828125, 0.476084228515625, 0.4754022521972656, 0.47531417846679686, 0.47490765380859373, 0.47505612182617185, 0.4752547912597656, 0.47493939208984376, 0.4751216735839844, 0.4752363586425781, 0.47553741455078125, 0.47550872802734373, 0.4752025451660156, 0.47503768920898437, 0.47521588134765624, 0.47512063598632814, 0.4751523742675781, 0.4767621154785156, 0.4764487609863281, 0.47532235717773436, 0.4752414855957031, 0.47862374877929686, 0.47638223266601565, 0.47562542724609375, 0.47544525146484373, 0.47627877807617186, 0.47506329345703124, 0.47538995361328124, 0.4754227294921875, 0.47551693725585936, 0.4754288635253906, 0.47596337890625, 0.4758896789550781, 0.4756899719238281, 0.4755988464355469, 0.4755916748046875, 0.4752066650390625, 0.4755640258789062, 0.47575653076171875, 0.47548416137695315, 0.4752998962402344, 0.4755332336425781, 0.4752363586425781, 0.4765552673339844, 0.4758056945800781, 0.47643853759765625, 0.9858508911132813, 0.4760975341796875, 0.4763627624511719, 0.47618048095703125, 0.47583026123046873, 0.4758814697265625, 0.47627365112304687, 0.47550360107421874, 0.4748226623535156, 0.4754380798339844, 0.4754176025390625, 0.47634841918945314, 0.47666278076171875, 0.4759449462890625, 0.47538177490234373, 0.475430908203125, 0.4754606018066406, 0.4765736999511719, 0.47562957763671876, 0.47619277954101563, 0.4762623901367187, 0.47586407470703124, 0.4763156433105469, 0.47800421142578126, 0.47563058471679687, 0.47566949462890623, 0.4764661865234375, 0.47568487548828126, 0.4763525085449219, 0.47546881103515626, 0.4754503784179688, 0.47535821533203126, 0.4751155090332031, 0.476685302734375, 0.4755333251953125, 0.4750540771484375, 0.4750325622558594, 0.4753387451171875, 0.47632281494140627, 0.47543603515625, 0.47514727783203126, 0.47536947631835935, 0.4750110778808594, 0.4750960693359375, 0.47505816650390625, 0.47526605224609375, 0.47514215087890627, 0.47597158813476564, 0.47550054931640623, 0.47509503173828127, 0.47519845581054687, 0.4751452026367188, 0.4753704833984375, 0.474967041015625, 0.47536639404296877, 0.47516876220703125, 0.47490457153320315, 0.4751523742675781, 0.47530905151367187, 0.4764610595703125, 0.47634228515625, 0.47607601928710935, 0.4758845520019531, 0.9857208251953125, 0.4755548095703125, 0.47583026123046873, 0.4755302734375, 0.4761640625, 0.4756316223144531, 0.475904052734375, 0.4751758728027344, 0.47539813232421874, 0.475536376953125, 0.4755210266113281, 0.475514892578125, 0.4755937805175781, 0.47548818969726564, 0.47541656494140627, 0.47544216918945315, 0.477454345703125, 0.4756387939453125, 0.4758138732910156, 0.47603302001953124, 0.47580465698242186, 0.47644158935546876, 0.4751994934082031, 0.47532339477539065, 0.4750745544433594, 0.47546981811523437, 0.4754565124511719, 0.47503768920898437, 0.47527935791015624, 0.4750858154296875, 0.4753018798828125, 0.4747683715820312, 0.4749916076660156, 0.47602789306640625, 0.4755906677246094, 0.4751790161132812, 0.47570123291015626, 0.477876220703125, 0.477338623046875, 0.4760637512207031, 0.4767999877929687, 0.47610983276367186, 0.47703347778320315, 0.47675802612304685, 0.47620712280273436, 0.4752158508300781, 0.47609036254882814, 0.4771328125, 0.4762552185058594, 0.4760504455566406, 0.47554458618164064, 0.475283447265625, 0.47543603515625, 0.4751523742675781, 0.475335693359375, 0.4752138366699219, 0.47528140258789064, 0.475109375, 0.47681332397460935, 0.47562240600585937, 0.47532339477539065, 0.47533465576171874, 0.47510833740234376, 0.984848388671875, 0.4750960693359375, 0.47506024169921873, 0.4759930419921875, 0.47555789184570313, 0.4750254211425781, 0.47533978271484373, 0.4758394775390625, 0.475610107421875, 0.47590911865234375, 0.4757596130371094, 0.4766791687011719, 0.4766371765136719, 0.4757176208496094, 0.4764549255371094, 0.4773990478515625, 0.4758917236328125, 0.476790771484375, 0.47595416259765627, 0.4752404479980469, 0.47515545654296876, 0.475030517578125, 0.4752916564941406, 0.4751278076171875, 0.4755138549804688, 0.47549234008789065, 0.47554150390625, 0.4758425598144531, 0.47556607055664063, 0.4752015380859375, 0.47505612182617185, 0.4750540771484375, 0.47509811401367186, 0.4751697998046875, 0.4750817260742187, 0.4752066650390625, 0.4769587097167969, 0.4757596130371094, 0.47552410888671875, 0.4753387451171875, 0.4751247253417969, 0.47600946044921877, 0.47491787719726564, 0.47531417846679686, 0.47513394165039063, 0.4757739562988281, 0.4750469055175781, 0.475030517578125, 0.4758507385253906, 0.4755599365234375, 0.47600741577148437, 0.4752414855957031, 0.4751247253417969, 0.47610470581054687, 0.4753530883789063, 0.475404296875, 0.4752005004882813, 0.476626953125, 0.4756654052734375, 0.47670578002929687, 0.4761466979980469, 0.4752547912597656, 0.4752209777832031, 0.9850972290039063, 0.47500698852539064, 0.47586099243164065, 0.475536376953125, 0.4751933288574219, 0.4753070068359375, 0.47523434448242186, 0.475157470703125, 0.4753039245605469, 0.4752762756347656, 0.4760606689453125, 0.47565216064453125, 0.4750796203613281, 0.47518106079101563, 0.4763607177734375, 0.47558758544921875, 0.47543499755859375, 0.47558349609375, 0.4752384033203125, 0.47536639404296877, 0.47527423095703125, 0.47547494506835936, 0.4750335998535156, 0.4753049621582031, 0.47516571044921874, 0.4754155578613281, 0.4772812805175781, 0.4751769714355469, 0.4758814697265625, 0.4761343994140625, 0.4757074279785156, 0.4766033630371094, 0.47504281616210936, 0.47508480834960937, 0.4752629699707031, 0.47697714233398436, 0.4764979248046875, 0.4756357421875, 0.4753837890625, 0.4751473083496094, 0.475273193359375, 0.475177978515625, 0.47527835083007813, 0.4753541259765625, 0.4753623046875, 0.4753930358886719, 0.4750796813964844, 0.47559576416015625, 0.47536334228515625, 0.4764241943359375, 0.47548724365234374, 0.47530087280273436, 0.47603302001953124, 0.4752906188964844, 0.47511962890625, 0.47501516723632814, 0.4768563232421875, 0.4758026123046875, 0.47588864135742187, 0.4753950805664062, 0.4752138366699219, 0.47535000610351563, 0.4750120849609375]",tokens/s,2.069224432025556,,,main,False,False,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen2-beta-72B,Qwen/Qwen2-beta-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 64, in run - report = scenario.run(backend) - File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run - _ = backend.generate(self.inputs, self.config.generate_kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate - return self.pretrained_model.generate(**inputs, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1576, in generate - result = self._greedy_search( - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2494, in _greedy_search - outputs = self( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1211, in forward - outputs = self.model( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1018, in forward - layer_outputs = decoder_layer( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 741, in forward - hidden_states, self_attn_weights, present_key_value = self.self_attn( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 435, in forward - query_states = self.q_proj(hidden_states) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward - assert AWQ_INSTALLED, ( -AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 96.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-2.7B,EleutherAI/gpt-neo-2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-125m,EleutherAI/gpt-neo-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-66b,facebook/opt-66b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 64, in run - report = scenario.run(backend) - File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run - _ = backend.generate(self.inputs, self.config.generate_kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate - return self.pretrained_model.generate(**inputs, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1576, in generate - result = self._greedy_search( - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2494, in _greedy_search - outputs = self( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 974, in forward - transformer_outputs = self.transformer( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 842, in forward - outputs = block( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 569, in forward - attn_outputs = self.attn( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 521, in forward - return self.attention( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 322, in forward - query = self.q_proj(hidden_states) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward - assert AWQ_INSTALLED, ( -AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: OPTForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-6B,01-ai/Yi-6B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,internlm/internlm2-20b,internlm/internlm2-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 64, in run - report = scenario.run(backend) - File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run - _ = backend.generate(self.inputs, self.config.generate_kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate - return self.pretrained_model.generate(**inputs, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1576, in generate - result = self._greedy_search( - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2494, in _greedy_search - outputs = self( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1211, in forward - outputs = self.model( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1018, in forward - layer_outputs = decoder_layer( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 741, in forward - hidden_states, self_attn_weights, present_key_value = self.self_attn( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 435, in forward - query_states = self.q_proj(hidden_states) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward - assert AWQ_INSTALLED, ( -AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 558, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3626, in from_pretrained + model = cls(config, *model_args, **model_kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm2-20b/f363ea8a116b3ea829c7a068ca24bc9d3e668083/modeling_internlm2.py"", line 974, in __init__ + self.model = InternLM2Model(config) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm2-20b/f363ea8a116b3ea829c7a068ca24bc9d3e668083/modeling_internlm2.py"", line 796, in __init__ + self.layers = nn.ModuleList([InternLM2DecoderLayer(config) for _ in range(config.num_hidden_layers)]) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm2-20b/f363ea8a116b3ea829c7a068ca24bc9d3e668083/modeling_internlm2.py"", line 796, in + self.layers = nn.ModuleList([InternLM2DecoderLayer(config) for _ in range(config.num_hidden_layers)]) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm2-20b/f363ea8a116b3ea829c7a068ca24bc9d3e668083/modeling_internlm2.py"", line 598, in __init__ + self.attention = INTERNLM2_ATTENTION_CLASSES[config.attn_implementation](config=config) +KeyError: 'sdpa' -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-2.7B,EleutherAI/gpt-neo-2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-6.7b,EleutherAI/pythia-6.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 64, in run - report = scenario.run(backend) - File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run - _ = backend.generate(self.inputs, self.config.generate_kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate - return self.pretrained_model.generate(**inputs, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1576, in generate - result = self._greedy_search( - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2494, in _greedy_search - outputs = self( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 974, in forward - transformer_outputs = self.transformer( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 842, in forward - outputs = block( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 569, in forward - attn_outputs = self.attn( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 521, in forward - return self.attention( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 322, in forward - query = self.q_proj(hidden_states) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward - assert AWQ_INSTALLED, ( -AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciCoder-1b,Deci/DeciCoder-1b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-1.4b,EleutherAI/pythia-1.4b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): - File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 600, in resolve_trust_remote_code - answer = input( -EOFError: EOF when reading a line - -During handling of the above exception, another exception occurred: - -Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 228, in load_model_with_no_weights - self.create_no_weights_model() - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 210, in create_no_weights_model - meta_model = self.automodel_class.from_config(self.pretrained_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 418, in from_config - trust_remote_code = resolve_trust_remote_code( - File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 613, in resolve_trust_remote_code + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa raise ValueError( -ValueError: The repository for Deci/DeciCoder-1b contains custom code which must be executed to correctly load the model. You can inspect the repository content at https://hf.co/Deci/DeciCoder-1b. -Please pass the argument `trust_remote_code=True` to allow custom code to be run. +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 -4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciLM-7B,Deci/DeciLM-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Salesforce/codegen-6B-nl,Salesforce/codegen-6B-nl,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): - File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 600, in resolve_trust_remote_code - answer = input( -EOFError: EOF when reading a line - -During handling of the above exception, another exception occurred: - -Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 228, in load_model_with_no_weights - self.create_no_weights_model() - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 210, in create_no_weights_model - meta_model = self.automodel_class.from_config(self.pretrained_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 418, in from_config - trust_remote_code = resolve_trust_remote_code( - File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 613, in resolve_trust_remote_code + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa raise ValueError( -ValueError: The repository for Deci/DeciLM-7B contains custom code which must be executed to correctly load the model. You can inspect the repository content at https://hf.co/Deci/DeciLM-7B. -Please pass the argument `trust_remote_code=True` to allow custom code to be run. +ValueError: CodeGenForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 -4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-j-6b,EleutherAI/gpt-j-6b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-70m,EleutherAI/pythia-70m,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -1406,32 +1610,26 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained - hf_quantizer.postprocess_model(model) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model - return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading - model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 436, in post_init_awq_exllama_modules - model = exllamav2_post_init( - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init - submodule.post_init(scratch_space=model.scratch_spaces[device]) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init - self.q_handle = exlv2_ext.make_q_matrix( -NameError: name 'exlv2_ext' is not defined + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 -4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-1.3B,EleutherAI/gpt-neo-1.3B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-110B,Qwen/Qwen1.5-110B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -1441,32 +1639,39 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained - hf_quantizer.postprocess_model(model) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model - return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading - model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 436, in post_init_awq_exllama_modules - model = exllamav2_post_init( - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init - submodule.post_init(scratch_space=model.scratch_spaces[device]) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init - self.q_handle = exlv2_ext.make_q_matrix( -NameError: name 'exlv2_ext' is not defined + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 32.00 MiB. GPU -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 -4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-34B,01-ai/Yi-34B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciCoder-1b,Deci/DeciCoder-1b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) - File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): - File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -1474,69 +1679,109 @@ ChildProcessError: Traceback (most recent call last): self.load_model_from_pretrained() File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained self.pretrained_model = self.automodel_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 558, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained hf_quantizer.postprocess_model(model) File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 436, in post_init_awq_exllama_modules - model = exllamav2_post_init( - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init - submodule.post_init(scratch_space=model.scratch_spaces[device]) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init - self.q_handle = exlv2_ext.make_q_matrix( -NameError: name 'exlv2_ext' is not defined + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 -4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-125m,EleutherAI/gpt-neo-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/recurrentgemma-2b,google/recurrentgemma-2b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/recurrentgemma-2b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aaa67-3a456d5f69f2dc1d10604993;b290a931-c314-4c73-9a5d-14be5f736551) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/recurrentgemma-2b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ - self.load_model_with_no_weights() - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights - self.load_model_from_pretrained() - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained - self.pretrained_model = self.automodel_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained - return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained - hf_quantizer.postprocess_model(model) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model - return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading - model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 436, in post_init_awq_exllama_modules - model = exllamav2_post_init( - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init - submodule.post_init(scratch_space=model.scratch_spaces[device]) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init - self.q_handle = exlv2_ext.make_q_matrix( -NameError: name 'exlv2_ext' is not defined + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/recurrentgemma-2b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 -4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-6B,01-ai/Yi-6B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,internlm/internlm-20b,internlm/internlm-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -1544,34 +1789,32 @@ ChildProcessError: Traceback (most recent call last): self.load_model_from_pretrained() File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained self.pretrained_model = self.automodel_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 558, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained - hf_quantizer.postprocess_model(model) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model - return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading - model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 436, in post_init_awq_exllama_modules - model = exllamav2_post_init( - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init - submodule.post_init(scratch_space=model.scratch_spaces[device]) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init - self.q_handle = exlv2_ext.make_q_matrix( -NameError: name 'exlv2_ext' is not defined + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3626, in from_pretrained + model = cls(config, *model_args, **model_kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm-20b/80729bcf52fbc4553d965926b27304ac5e156d98/modeling_internlm.py"", line 906, in __init__ + self.model = InternLMModel(config) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm-20b/80729bcf52fbc4553d965926b27304ac5e156d98/modeling_internlm.py"", line 729, in __init__ + self.layers = nn.ModuleList([InternLMDecoderLayer(config) for _ in range(config.num_hidden_layers)]) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm-20b/80729bcf52fbc4553d965926b27304ac5e156d98/modeling_internlm.py"", line 729, in + self.layers = nn.ModuleList([InternLMDecoderLayer(config) for _ in range(config.num_hidden_layers)]) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm-20b/80729bcf52fbc4553d965926b27304ac5e156d98/modeling_internlm.py"", line 545, in __init__ + self.self_attn = INTERNLM_ATTENTION_CLASSES[config.attn_implementation](config=config) +KeyError: 'sdpa' -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 -4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-2.7B,EleutherAI/gpt-neo-2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-2.7b,facebook/opt-2.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -1581,32 +1824,26 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained - hf_quantizer.postprocess_model(model) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model - return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading - model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 436, in post_init_awq_exllama_modules - model = exllamav2_post_init( - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init - submodule.post_init(scratch_space=model.scratch_spaces[device]) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init - self.q_handle = exlv2_ext.make_q_matrix( -NameError: name 'exlv2_ext' is not defined + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: OPTForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 -4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciCoder-1b,Deci/DeciCoder-1b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-125m,facebook/opt-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -1614,69 +1851,103 @@ ChildProcessError: Traceback (most recent call last): self.load_model_from_pretrained() File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained self.pretrained_model = self.automodel_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 558, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained - hf_quantizer.postprocess_model(model) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model - return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading - model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 432, in post_init_awq_exllama_modules - model = exllama_post_init(model) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init - submodule.post_init() - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init - self.q4 = exl_ext.make_q4( -NameError: name 'exl_ext' is not defined + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: OPTForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 -4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciLM-7B,Deci/DeciLM-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/gemma-7b,google/gemma-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/gemma-7b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aaa15-551341a36ec2255638579554;94b4bc4d-df28-42dc-a162-9a127ff0a232) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/gemma-7b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ - self.load_model_with_no_weights() - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights - self.load_model_from_pretrained() - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained - self.pretrained_model = self.automodel_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 558, in from_pretrained - return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained - hf_quantizer.postprocess_model(model) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model - return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading - model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 432, in post_init_awq_exllama_modules - model = exllama_post_init(model) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init - submodule.post_init() - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init - self.q4 = exl_ext.make_q4( -NameError: name 'exl_ext' is not defined + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/gemma-7b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 -4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-j-6b,EleutherAI/gpt-j-6b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,huggyllama/llama-13b,huggyllama/llama-13b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -1686,13 +1957,13 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained hf_quantizer.postprocess_model(model) File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 432, in post_init_awq_exllama_modules + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules model = exllama_post_init(model) File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init submodule.post_init() @@ -1700,18 +1971,18 @@ ChildProcessError: Traceback (most recent call last): self.q4 = exl_ext.make_q4( NameError: name 'exl_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 -4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-1.3B,EleutherAI/gpt-neo-1.3B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-7B,Qwen/Qwen-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -1719,15 +1990,144 @@ ChildProcessError: Traceback (most recent call last): self.load_model_from_pretrained() File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained self.pretrained_model = self.automodel_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 550, in from_pretrained + model_class = get_class_from_dynamic_module( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 498, in get_class_from_dynamic_module + final_module = get_cached_module_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 323, in get_cached_module_file + modules_needed = check_imports(resolved_module_file) + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 181, in check_imports + raise ImportError( +ImportError: This modeling file requires the following packages that were not found in your environment: transformers_stream_generator. Run `pip install transformers_stream_generator` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-72B,Qwen/Qwen1.5-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 96.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciLM-7B,Deci/DeciLM-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 558, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: DeciLMForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-j-6b,EleutherAI/gpt-j-6b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTJForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-0.5B,Qwen/Qwen1.5-0.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained hf_quantizer.postprocess_model(model) File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 432, in post_init_awq_exllama_modules + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules model = exllama_post_init(model) File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init submodule.post_init() @@ -1735,18 +2135,76 @@ ChildProcessError: Traceback (most recent call last): self.q4 = exl_ext.make_q4( NameError: name 'exl_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 -4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neox-20b,EleutherAI/gpt-neox-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-160m,EleutherAI/pythia-160m,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Salesforce/codegen-16B-nl,Salesforce/codegen-16B-nl,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: CodeGenForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-MoE-A2.7B,Qwen/Qwen1.5-MoE-A2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -1756,13 +2214,13 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained hf_quantizer.postprocess_model(model) File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 432, in post_init_awq_exllama_modules + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules model = exllama_post_init(model) File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init submodule.post_init() @@ -1770,18 +2228,76 @@ ChildProcessError: Traceback (most recent call last): self.q4 = exl_ext.make_q4( NameError: name 'exl_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 -4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-34B,01-ai/Yi-34B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-1.3b,EleutherAI/pythia-1.3b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-2.7b,EleutherAI/pythia-2.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-14B,Qwen/Qwen1.5-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -1791,13 +2307,13 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained hf_quantizer.postprocess_model(model) File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 432, in post_init_awq_exllama_modules + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules model = exllama_post_init(model) File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init submodule.post_init() @@ -1805,18 +2321,18 @@ ChildProcessError: Traceback (most recent call last): self.q4 = exl_ext.make_q4( NameError: name 'exl_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 -4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-125m,EleutherAI/gpt-neo-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen2-beta-14B,Qwen/Qwen2-beta-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -1826,13 +2342,13 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained hf_quantizer.postprocess_model(model) File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 432, in post_init_awq_exllama_modules + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules model = exllama_post_init(model) File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init submodule.post_init() @@ -1840,18 +2356,151 @@ ChildProcessError: Traceback (most recent call last): self.q4 = exl_ext.make_q4( NameError: name 'exl_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 -4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-6B,01-ai/Yi-6B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-30b,facebook/opt-30b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: OPTForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-1.3B,EleutherAI/gpt-neo-1.3B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,databricks/dbrx-base,databricks/dbrx-base,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/databricks/dbrx-base/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aa530-0be33bce01a4e1e87987d70d;13bbd174-f909-4c06-84f7-58966286c385) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/databricks/dbrx-base/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like databricks/dbrx-base is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,huggyllama/llama-7b,huggyllama/llama-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -1861,13 +2510,13 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained hf_quantizer.postprocess_model(model) File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 432, in post_init_awq_exllama_modules + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules model = exllama_post_init(model) File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init submodule.post_init() @@ -1875,18 +2524,47 @@ ChildProcessError: Traceback (most recent call last): self.q4 = exl_ext.make_q4( NameError: name 'exl_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 -4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-2.7B,EleutherAI/gpt-neo-2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neox-20b,EleutherAI/gpt-neox-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-4B,Qwen/Qwen1.5-4B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -1896,13 +2574,13 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained hf_quantizer.postprocess_model(model) File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 432, in post_init_awq_exllama_modules + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules model = exllama_post_init(model) File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init submodule.post_init() @@ -1910,311 +2588,18586 @@ ChildProcessError: Traceback (most recent call last): self.q4 = exl_ext.make_q4( NameError: name 'exl_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 -4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciCoder-1b,Deci/DeciCoder-1b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-1.8B,Qwen/Qwen1.5-1.8B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 64, in run - report = scenario.run(backend) - File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run - _ = backend.generate(self.inputs, self.config.generate_kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate - return self.pretrained_model.generate(**inputs, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1576, in generate - result = self._greedy_search( - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2494, in _greedy_search - outputs = self( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1211, in forward - outputs = self.model( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1018, in forward - layer_outputs = decoder_layer( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 741, in forward - hidden_states, self_attn_weights, present_key_value = self.self_attn( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) -TypeError: DeciCoderAttention.forward() got an unexpected keyword argument 'cache_position' + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciLM-7B,Deci/DeciLM-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-14B,Qwen/Qwen-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 64, in run - report = scenario.run(backend) - File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run - _ = backend.generate(self.inputs, self.config.generate_kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate - return self.pretrained_model.generate(**inputs, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1576, in generate - result = self._greedy_search( - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2494, in _greedy_search - outputs = self( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/root/.cache/huggingface/modules/transformers_modules/Deci/DeciLM-7B/c3c9f4226801dc0433f32aebffe0aac68ee2f051/transformers_v4_35_2__modeling_llama.py"", line 1034, in forward - outputs = self.model( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/root/.cache/huggingface/modules/transformers_modules/Deci/DeciLM-7B/c3c9f4226801dc0433f32aebffe0aac68ee2f051/modeling_decilm.py"", line 274, in forward - layer_outputs = decoder_layer( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/root/.cache/huggingface/modules/transformers_modules/Deci/DeciLM-7B/c3c9f4226801dc0433f32aebffe0aac68ee2f051/transformers_v4_35_2__modeling_llama.py"", line 672, in forward - hidden_states, self_attn_weights, present_key_value = self.self_attn( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/root/.cache/huggingface/modules/transformers_modules/Deci/DeciLM-7B/c3c9f4226801dc0433f32aebffe0aac68ee2f051/modeling_decilm.py"", line 84, in forward - query_states = self.q_proj(hidden_states) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward - assert AWQ_INSTALLED, ( -AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 550, in from_pretrained + model_class = get_class_from_dynamic_module( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 498, in get_class_from_dynamic_module + final_module = get_cached_module_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 323, in get_cached_module_file + modules_needed = check_imports(resolved_module_file) + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 181, in check_imports + raise ImportError( +ImportError: This modeling file requires the following packages that were not found in your environment: transformers_stream_generator. Run `pip install transformers_stream_generator` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-j-6b,EleutherAI/gpt-j-6b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-7.5B,facebook/xglm-7.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 64, in run - report = scenario.run(backend) - File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run - _ = backend.generate(self.inputs, self.config.generate_kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate - return self.pretrained_model.generate(**inputs, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1576, in generate - result = self._greedy_search( - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2494, in _greedy_search - outputs = self( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gptj/modeling_gptj.py"", line 1127, in forward - transformer_outputs = self.transformer( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gptj/modeling_gptj.py"", line 953, in forward - outputs = block( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gptj/modeling_gptj.py"", line 581, in forward - attn_outputs = self.attn( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gptj/modeling_gptj.py"", line 227, in forward - query = self.q_proj(hidden_states) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward - assert AWQ_INSTALLED, ( -AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: XGLMForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-1.3B,EleutherAI/gpt-neo-1.3B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-12b,EleutherAI/pythia-12b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 64, in run - report = scenario.run(backend) - File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run - _ = backend.generate(self.inputs, self.config.generate_kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate - return self.pretrained_model.generate(**inputs, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1576, in generate - result = self._greedy_search( - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2494, in _greedy_search - outputs = self( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 974, in forward - transformer_outputs = self.transformer( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 842, in forward - outputs = block( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 569, in forward - attn_outputs = self.attn( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 521, in forward - return self.attention( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 263, in forward - query = self.q_proj(hidden_states) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward - assert AWQ_INSTALLED, ( -AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neox-20b,EleutherAI/gpt-neox-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-7B,Qwen/Qwen1.5-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 64, in run - report = scenario.run(backend) - File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run - _ = backend.generate(self.inputs, self.config.generate_kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate - return self.pretrained_model.generate(**inputs, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1576, in generate - result = self._greedy_search( - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2494, in _greedy_search - outputs = self( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 1034, in forward - outputs = self.gpt_neox( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 925, in forward - outputs = layer( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 690, in forward - attention_layer_outputs = self.attention( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 175, in forward - qkv = self.query_key_value(hidden_states) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward - assert AWQ_INSTALLED, ( -AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-34B,01-ai/Yi-34B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-34B,01-ai/Yi-34B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) - File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): - File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 64, in run - report = scenario.run(backend) - File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/gemma-2b,google/gemma-2b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/gemma-2b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aa9b9-3140fca435e1f16f153b386d;3c6cdadf-785a-4bae-a17b-e298342c1b63) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/gemma-2b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/gemma-2b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/polyglot-ko-12.8b,EleutherAI/polyglot-ko-12.8b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-125m,EleutherAI/gpt-neo-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-13b,facebook/opt-13b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: OPTForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/recurrentgemma-7b,google/recurrentgemma-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 404 Client Error: Not Found for url: https://huggingface.co/google/recurrentgemma-7b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1823, in _raise_on_head_call_error + raise head_call_error + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 352, in hf_raise_for_status + raise RepositoryNotFoundError(message, response) from e +huggingface_hub.utils._errors.RepositoryNotFoundError: 404 Client Error. (Request ID: Root=1-664aaab9-28822f101473f11448384b6f;72ccc4be-99b5-4346-adf1-79176bebb380) + +Repository Not Found for url: https://huggingface.co/google/recurrentgemma-7b/resolve/main/config.json. +Please make sure you specified the correct `repo_id` and `repo_type`. +If you are trying to access a private or gated repo, make sure you are authenticated. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 422, in cached_file + raise EnvironmentError( +OSError: google/recurrentgemma-7b is not a local folder and is not a valid model identifier listed on 'https://huggingface.co/models' +If this is a private repository, make sure to pass a token having permission to this repo either by logging in with `huggingface-cli login` or by passing `token=` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-410m,EleutherAI/pythia-410m,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-32B,Qwen/Qwen1.5-32B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-4.5B,facebook/xglm-4.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: XGLMForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,TencentARC/Mistral_Pro_8B_v0.1,TencentARC/Mistral_Pro_8B_v0.1,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-564M,facebook/xglm-564M,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: XGLMForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-350m,facebook/opt-350m,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: OPTForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-6.7b,facebook/opt-6.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: OPTForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-72B,Qwen/Qwen-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 550, in from_pretrained + model_class = get_class_from_dynamic_module( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 498, in get_class_from_dynamic_module + final_module = get_cached_module_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 323, in get_cached_module_file + modules_needed = check_imports(resolved_module_file) + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 181, in check_imports + raise ImportError( +ImportError: This modeling file requires the following packages that were not found in your environment: transformers_stream_generator. Run `pip install transformers_stream_generator` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-6B,01-ai/Yi-6B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen2-beta-72B,Qwen/Qwen2-beta-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 96.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-2.7B,EleutherAI/gpt-neo-2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-66b,facebook/opt-66b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: OPTForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,internlm/internlm2-20b,internlm/internlm2-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 558, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3626, in from_pretrained + model = cls(config, *model_args, **model_kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm2-20b/f363ea8a116b3ea829c7a068ca24bc9d3e668083/modeling_internlm2.py"", line 974, in __init__ + self.model = InternLM2Model(config) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm2-20b/f363ea8a116b3ea829c7a068ca24bc9d3e668083/modeling_internlm2.py"", line 796, in __init__ + self.layers = nn.ModuleList([InternLM2DecoderLayer(config) for _ in range(config.num_hidden_layers)]) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm2-20b/f363ea8a116b3ea829c7a068ca24bc9d3e668083/modeling_internlm2.py"", line 796, in + self.layers = nn.ModuleList([InternLM2DecoderLayer(config) for _ in range(config.num_hidden_layers)]) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm2-20b/f363ea8a116b3ea829c7a068ca24bc9d3e668083/modeling_internlm2.py"", line 598, in __init__ + self.attention = INTERNLM2_ATTENTION_CLASSES[config.attn_implementation](config=config) +KeyError: 'sdpa' + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-6.7b,EleutherAI/pythia-6.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-1.4b,EleutherAI/pythia-1.4b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Salesforce/codegen-6B-nl,Salesforce/codegen-6B-nl,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1469, in _autoset_attn_implementation + cls._check_and_enable_flash_attn_2( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1560, in _check_and_enable_flash_attn_2 + raise ValueError( +ValueError: CodeGenForCausalLM does not support Flash Attention 2.0 yet. Please request to add support where the model is hosted, on its model hub page: https://huggingface.co//tmp/tmpficmgqtz/no_weights_model/discussions/new or in the Transformers GitHub repo: https://github.com/huggingface/transformers/issues/new + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-70m,EleutherAI/pythia-70m,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-110B,Qwen/Qwen1.5-110B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 32.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciCoder-1b,Deci/DeciCoder-1b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 558, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/recurrentgemma-2b,google/recurrentgemma-2b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/recurrentgemma-2b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aaa82-3e32a0ae2496f29275d4bfa2;f3444d97-fdfa-4a98-9cee-0ff83cccadf3) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/recurrentgemma-2b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/recurrentgemma-2b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,internlm/internlm-20b,internlm/internlm-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 558, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-2.7b,facebook/opt-2.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-125m,facebook/opt-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/gemma-7b,google/gemma-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/gemma-7b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aaa30-5b4e2f1260ca513e5d6f53de;fc8a85a9-2778-4c51-ae44-93359d4081dd) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/gemma-7b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/gemma-7b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-7B,Qwen/Qwen-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 550, in from_pretrained + model_class = get_class_from_dynamic_module( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 498, in get_class_from_dynamic_module + final_module = get_cached_module_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 323, in get_cached_module_file + modules_needed = check_imports(resolved_module_file) + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 181, in check_imports + raise ImportError( +ImportError: This modeling file requires the following packages that were not found in your environment: transformers_stream_generator. Run `pip install transformers_stream_generator` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-72B,Qwen/Qwen1.5-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 96.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciLM-7B,Deci/DeciLM-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 558, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-j-6b,EleutherAI/gpt-j-6b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-0.5B,Qwen/Qwen1.5-0.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-160m,EleutherAI/pythia-160m,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Salesforce/codegen-16B-nl,Salesforce/codegen-16B-nl,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1469, in _autoset_attn_implementation + cls._check_and_enable_flash_attn_2( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1560, in _check_and_enable_flash_attn_2 + raise ValueError( +ValueError: CodeGenForCausalLM does not support Flash Attention 2.0 yet. Please request to add support where the model is hosted, on its model hub page: https://huggingface.co//tmp/tmp9a964gjq/no_weights_model/discussions/new or in the Transformers GitHub repo: https://github.com/huggingface/transformers/issues/new + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-MoE-A2.7B,Qwen/Qwen1.5-MoE-A2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-1.3b,EleutherAI/pythia-1.3b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-2.7b,EleutherAI/pythia-2.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-14B,Qwen/Qwen1.5-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen2-beta-14B,Qwen/Qwen2-beta-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-30b,facebook/opt-30b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-1.3B,EleutherAI/gpt-neo-1.3B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,databricks/dbrx-base,databricks/dbrx-base,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/databricks/dbrx-base/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aa54c-4acded0c06b30b6e5a37b45a;53d275f7-fbe2-40c5-b28d-3a53f3890719) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/databricks/dbrx-base/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like databricks/dbrx-base is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,huggyllama/llama-7b,huggyllama/llama-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neox-20b,EleutherAI/gpt-neox-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-4B,Qwen/Qwen1.5-4B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-1.8B,Qwen/Qwen1.5-1.8B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-14B,Qwen/Qwen-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 550, in from_pretrained + model_class = get_class_from_dynamic_module( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 498, in get_class_from_dynamic_module + final_module = get_cached_module_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 323, in get_cached_module_file + modules_needed = check_imports(resolved_module_file) + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 181, in check_imports + raise ImportError( +ImportError: This modeling file requires the following packages that were not found in your environment: transformers_stream_generator. Run `pip install transformers_stream_generator` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-7.5B,facebook/xglm-7.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1469, in _autoset_attn_implementation + cls._check_and_enable_flash_attn_2( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1560, in _check_and_enable_flash_attn_2 + raise ValueError( +ValueError: XGLMForCausalLM does not support Flash Attention 2.0 yet. Please request to add support where the model is hosted, on its model hub page: https://huggingface.co//tmp/tmpk_zyud20/no_weights_model/discussions/new or in the Transformers GitHub repo: https://github.com/huggingface/transformers/issues/new + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-12b,EleutherAI/pythia-12b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-7B,Qwen/Qwen1.5-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-34B,01-ai/Yi-34B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/gemma-2b,google/gemma-2b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/gemma-2b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aa9d9-47f011087cce4ea9787e6bd6;8e8654e4-7767-4efd-81bb-673d1c1a23fb) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/gemma-2b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/gemma-2b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/polyglot-ko-12.8b,EleutherAI/polyglot-ko-12.8b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-125m,EleutherAI/gpt-neo-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-13b,facebook/opt-13b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/recurrentgemma-7b,google/recurrentgemma-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 404 Client Error: Not Found for url: https://huggingface.co/google/recurrentgemma-7b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1823, in _raise_on_head_call_error + raise head_call_error + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 352, in hf_raise_for_status + raise RepositoryNotFoundError(message, response) from e +huggingface_hub.utils._errors.RepositoryNotFoundError: 404 Client Error. (Request ID: Root=1-664aaad5-780a7a8c56e27c0379263083;2cd16c37-da9d-46f5-86b4-4c585d348ee7) + +Repository Not Found for url: https://huggingface.co/google/recurrentgemma-7b/resolve/main/config.json. +Please make sure you specified the correct `repo_id` and `repo_type`. +If you are trying to access a private or gated repo, make sure you are authenticated. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 422, in cached_file + raise EnvironmentError( +OSError: google/recurrentgemma-7b is not a local folder and is not a valid model identifier listed on 'https://huggingface.co/models' +If this is a private repository, make sure to pass a token having permission to this repo either by logging in with `huggingface-cli login` or by passing `token=` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-410m,EleutherAI/pythia-410m,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-32B,Qwen/Qwen1.5-32B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-4.5B,facebook/xglm-4.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1469, in _autoset_attn_implementation + cls._check_and_enable_flash_attn_2( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1560, in _check_and_enable_flash_attn_2 + raise ValueError( +ValueError: XGLMForCausalLM does not support Flash Attention 2.0 yet. Please request to add support where the model is hosted, on its model hub page: https://huggingface.co//tmp/tmpbiaq465j/no_weights_model/discussions/new or in the Transformers GitHub repo: https://github.com/huggingface/transformers/issues/new + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,TencentARC/Mistral_Pro_8B_v0.1,TencentARC/Mistral_Pro_8B_v0.1,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-564M,facebook/xglm-564M,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1469, in _autoset_attn_implementation + cls._check_and_enable_flash_attn_2( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1560, in _check_and_enable_flash_attn_2 + raise ValueError( +ValueError: XGLMForCausalLM does not support Flash Attention 2.0 yet. Please request to add support where the model is hosted, on its model hub page: https://huggingface.co//tmp/tmpvqw4a06w/no_weights_model/discussions/new or in the Transformers GitHub repo: https://github.com/huggingface/transformers/issues/new + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-350m,facebook/opt-350m,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-6.7b,facebook/opt-6.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-72B,Qwen/Qwen-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 550, in from_pretrained + model_class = get_class_from_dynamic_module( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 498, in get_class_from_dynamic_module + final_module = get_cached_module_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 323, in get_cached_module_file + modules_needed = check_imports(resolved_module_file) + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 181, in check_imports + raise ImportError( +ImportError: This modeling file requires the following packages that were not found in your environment: transformers_stream_generator. Run `pip install transformers_stream_generator` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-6B,01-ai/Yi-6B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen2-beta-72B,Qwen/Qwen2-beta-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 96.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-2.7B,EleutherAI/gpt-neo-2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-66b,facebook/opt-66b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 162.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,internlm/internlm2-20b,internlm/internlm2-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 558, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-6.7b,EleutherAI/pythia-6.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-1.4b,EleutherAI/pythia-1.4b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,,,MB,1591.713792,2250.768384,0.0,1604.32128,1463.693312,s,10,1.2648313903808592,0.12648313903808595,0.0013989604614154947,0.1262478713989258,0.1271006202697754,0.1287280216217041,0.13002994270324708,"[0.1303554229736328, 0.12560326385498047, 0.12551904296875, 0.12673897552490235, 0.12573398590087892, 0.12508892822265624, 0.12599820709228515, 0.1264975357055664, 0.1265692138671875, 0.12672681427001953]",tokens/s,2023.9851884361806,kWh,1.4855278241965507e-06,8.139949754399822e-07,6.45409544104969e-06,8.753618240686224e-06,tokens/kWh,29245049.64245863,MB,1591.713792,2250.768384,0.0,1604.32128,1560.975872,s,10,72.9124365234375,7.291243652343749,0.003929707108860259,7.2908349609375005,7.297245703125,7.2972966796875,7.2973374609375,"[7.297234375, 7.2856533203125, 7.29734765625, 7.29276513671875, 7.29289794921875, 7.29063818359375, 7.29089404296875, 7.29077587890625, 7.28937060546875, 7.284859375]",tokens/s,8.640501264794358,kWh,8.609067341933647e-05,4.718392125277432e-05,0.00037183251274354733,0.0005051071074156582,tokens/kWh,124726.02162011672,,s,629,73.93082885742194,0.11753708880353239,0.015022143769719637,0.11565261077880859,0.11614679260253907,0.11651727142333984,0.24159293273925783,"[0.11554611206054688, 0.11542835235595703, 0.11560959625244141, 0.11652607727050782, 0.11558604431152343, 0.1156341781616211, 0.11573760223388672, 0.1157570571899414, 0.11565670776367187, 0.11562393951416015, 0.11558502197265624, 0.11574476623535156, 0.1157580795288086, 0.11542835235595703, 0.11543961334228516, 0.11608268737792969, 0.1160478744506836, 0.11690393829345704, 0.11582669067382813, 0.11567513275146485, 0.1154303970336914, 0.11542835235595703, 0.1157232666015625, 0.1157694091796875, 0.11561158752441406, 0.11565261077880859, 0.11558297729492187, 0.11557170867919922, 0.11555225372314454, 0.11560857391357422, 0.11558604431152343, 0.11568946838378906, 0.11568537902832031, 0.11553177642822265, 0.1154703369140625, 0.11592601776123047, 0.11566284942626953, 0.11548159790039063, 0.11579289245605469, 0.11569561767578125, 0.11579801940917969, 0.11614924621582032, 0.11590553283691406, 0.1172838363647461, 0.11599565124511718, 0.11589427185058594, 0.11579392242431641, 0.11582054138183594, 0.11560550689697266, 0.11576937866210937, 0.11679228973388672, 0.11623014068603515, 0.115884033203125, 0.11555532836914062, 0.11562290954589843, 0.11581132507324218, 0.11705241394042969, 0.11600486755371094, 0.11588607788085938, 0.115736572265625, 0.11612569427490234, 0.11604176330566406, 0.24190666198730468, 0.11594547271728516, 0.11553177642822265, 0.1155819549560547, 0.1155051498413086, 0.11557990264892579, 0.11673600006103516, 0.11557376098632813, 0.11559117126464843, 0.1155072021484375, 0.11576525115966797, 0.11566079711914062, 0.11590656280517578, 0.115525634765625, 0.11559219360351562, 0.1160847396850586, 0.11556658935546875, 0.1156147232055664, 0.11551641845703126, 0.11533209228515626, 0.11555328369140624, 0.11560147094726562, 0.11554502105712891, 0.11557376098632813, 0.1158809585571289, 0.11545088195800782, 0.11558604431152343, 0.11544985961914063, 0.11559321594238281, 0.11556761932373047, 0.11569868469238281, 0.11571097564697265, 0.11561676788330078, 0.11552051544189453, 0.11553997039794922, 0.11532393646240234, 0.11541606140136719, 0.11545494079589844, 0.11565055847167968, 0.11606527709960937, 0.11634278106689454, 0.1155440673828125, 0.11550105285644531, 0.11538432312011719, 0.11557785797119141, 0.11540582275390625, 0.11706163024902344, 0.11553382110595703, 0.11559321594238281, 0.11543244934082031, 0.11545804595947265, 0.11543142700195312, 0.11540480041503906, 0.11530547332763671, 0.11539667510986328, 0.11571295928955078, 0.11585740661621094, 0.11560447692871094, 0.11563622283935547, 0.11560550689697266, 0.11555430603027343, 0.11554303741455078, 0.1156147232055664, 0.24147865295410156, 0.11553382110595703, 0.11642678070068359, 0.11596393585205078, 0.11609900665283203, 0.11560550689697266, 0.11545906829833984, 0.11533824157714843, 0.11538944244384766, 0.11619737243652344, 0.11612876892089843, 0.11601612854003907, 0.1155758056640625, 0.11589433288574219, 0.1156719970703125, 0.11552870178222656, 0.11574681854248046, 0.11581337738037109, 0.11625574493408203, 0.11572531127929687, 0.11590656280517578, 0.11556454467773437, 0.11557075500488281, 0.1164062042236328, 0.11572736358642578, 0.11559935760498047, 0.1159710693359375, 0.11614105224609375, 0.11683942413330078, 0.11663980865478515, 0.11592697906494141, 0.11562598419189453, 0.11584307098388671, 0.11552973175048828, 0.11597516632080078, 0.11550310516357422, 0.11585126495361328, 0.11633869171142579, 0.11576121520996094, 0.11590239715576171, 0.11548057556152344, 0.11562700653076172, 0.11580518341064452, 0.11599974060058593, 0.11608370971679688, 0.11642265319824219, 0.11600179290771484, 0.11641446685791015, 0.11563212585449219, 0.11547647857666016, 0.11545088195800782, 0.11568032073974609, 0.11588703918457031, 0.11573350524902344, 0.1157201919555664, 0.11566182708740234, 0.11576627349853516, 0.11582566070556641, 0.11572736358642578, 0.11544780731201172, 0.11562393951416015, 0.11546828460693359, 0.11547853088378907, 0.2430136260986328, 0.1159004135131836, 0.11564236450195313, 0.11563827514648438, 0.11565567779541015, 0.11559526062011719, 0.11573350524902344, 0.11556658935546875, 0.1157396469116211, 0.1160284194946289, 0.11578982543945313, 0.11594035339355468, 0.11562290954589843, 0.11544985961914063, 0.11558502197265624, 0.11559117126464843, 0.11605811309814452, 0.11588301086425781, 0.1157949447631836, 0.11583795166015624, 0.11589734649658204, 0.11614617919921875, 0.11590348815917968, 0.1155440673828125, 0.1166714859008789, 0.11601510620117188, 0.11548365020751954, 0.11576012420654297, 0.11538944244384766, 0.11538944244384766, 0.11550617980957031, 0.1161164779663086, 0.11625676727294922, 0.11571814727783203, 0.11565875244140625, 0.1156485137939453, 0.11559117126464843, 0.11533106994628907, 0.11534438323974609, 0.115957763671875, 0.11574066925048829, 0.11546316528320312, 0.11652095794677735, 0.11555532836914062, 0.11549081420898437, 0.11547443389892578, 0.11542630767822265, 0.11534438323974609, 0.11527986907958984, 0.11548365020751954, 0.11580825805664062, 0.1169459228515625, 0.11576729583740235, 0.11597516632080078, 0.11551849365234375, 0.11617174530029296, 0.1161553955078125, 0.11608678436279297, 0.11566079711914062, 0.11565261077880859, 0.11548365020751954, 0.1155758056640625, 0.11557376098632813, 0.2416373748779297, 0.11573760223388672, 0.11561881256103515, 0.11579596710205078, 0.11593830108642578, 0.115810302734375, 0.11552870178222656, 0.11581542205810547, 0.11552665710449218, 0.11582771301269532, 0.1157396469116211, 0.11553791809082031, 0.1157027816772461, 0.11587789154052734, 0.11567513275146485, 0.11574169921875, 0.11559833526611328, 0.11615334320068359, 0.11550822448730469, 0.11546316528320312, 0.11542733001708984, 0.11593523406982421, 0.11576217651367188, 0.11546521759033203, 0.11530751800537109, 0.11534848022460938, 0.11562496185302734, 0.11573248291015625, 0.11568946838378906, 0.11568025970458984, 0.1157570571899414, 0.11558809661865234, 0.11546009826660156, 0.11551436614990235, 0.11556864166259766, 0.11568851470947265, 0.11561977386474609, 0.11564441680908204, 0.11560243225097656, 0.11574374389648437, 0.11693260955810547, 0.11600179290771484, 0.11581951904296875, 0.11551337432861328, 0.11621782684326172, 0.11561574554443359, 0.11616153717041015, 0.11619328308105469, 0.11597926330566406, 0.11552153778076171, 0.11582157135009766, 0.11549798583984375, 0.11565773010253906, 0.11573868560791016, 0.11633657836914063, 0.11591986846923828, 0.11580723571777343, 0.11681075286865235, 0.11580210876464844, 0.11566284942626953, 0.11573554992675782, 0.11571405029296875, 0.11579392242431641, 0.24123802185058593, 0.11577037048339844, 0.11569152069091797, 0.11550924682617188, 0.11548262023925782, 0.11541299438476563, 0.11537100982666015, 0.11548979187011718, 0.11538540649414063, 0.11682399749755859, 0.11572531127929687, 0.11558911895751953, 0.11550003051757812, 0.11544576263427735, 0.11544678497314453, 0.11560755157470703, 0.11546828460693359, 0.11556249237060547, 0.1156485137939453, 0.11555023956298828, 0.11552047729492188, 0.11564236450195313, 0.11559014129638671, 0.11573452758789063, 0.11543859100341797, 0.11551641845703126, 0.11565465545654297, 0.11574988555908203, 0.11569971466064453, 0.11559935760498047, 0.11563724517822266, 0.11556044769287109, 0.11579084777832031, 0.11558399963378906, 0.11560345458984375, 0.11583385467529297, 0.11566182708740234, 0.11573350524902344, 0.11707392120361328, 0.11604377746582031, 0.11578166198730469, 0.11570275115966797, 0.11584614562988281, 0.11643289947509766, 0.11766886138916016, 0.11568742370605468, 0.11568131256103516, 0.1155184326171875, 0.11561779022216796, 0.11559219360351562, 0.1157734375, 0.1157580795288086, 0.11578675079345703, 0.11578163146972656, 0.11558399963378906, 0.1155758056640625, 0.11558399963378906, 0.11569766235351563, 0.11568230438232421, 0.11554412841796875, 0.11554502105712891, 0.11572531127929687, 0.11579392242431641, 0.24222003173828124, 0.11544473266601563, 0.11548876953125, 0.11571814727783203, 0.11562905883789062, 0.1165486068725586, 0.11599871826171874, 0.11563724517822266, 0.11574374389648437, 0.11579801940917969, 0.11553382110595703, 0.11554611206054688, 0.11684249877929688, 0.11557273864746094, 0.11547853088378907, 0.1154877471923828, 0.11573350524902344, 0.11547853088378907, 0.11548876953125, 0.11556556701660156, 0.1155973129272461, 0.11554815673828125, 0.1156280288696289, 0.1164400634765625, 0.115525634765625, 0.11554815673828125, 0.11559321594238281, 0.11538022613525391, 0.11541510772705078, 0.11557574462890625, 0.11551538848876954, 0.11658854675292969, 0.11569561767578125, 0.11581132507324218, 0.11544371032714844, 0.1155594253540039, 0.11566284942626953, 0.11575603485107422, 0.1155072021484375, 0.11597618865966797, 0.11624857330322266, 0.11573554992675782, 0.11539968109130859, 0.11556352233886719, 0.11553997039794922, 0.1154119644165039, 0.11562086486816406, 0.11566899108886719, 0.11578470611572265, 0.11621580505371094, 0.11563520050048828, 0.11619840240478516, 0.11569459533691406, 0.11568230438232421, 0.11591270446777344, 0.11579801940917969, 0.11572531127929687, 0.11583692932128906, 0.11580006408691407, 0.11551436614990235, 0.11591372680664062, 0.11565055847167968, 0.11566796875, 0.2425927734375, 0.11592601776123047, 0.11652812957763672, 0.1160263671875, 0.1156280288696289, 0.1154703369140625, 0.11540684509277344, 0.11567411041259766, 0.11576422119140625, 0.11549388885498046, 0.11578163146972656, 0.11591372680664062, 0.11623423767089844, 0.11559321594238281, 0.11569459533691406, 0.11551744079589844, 0.11644620513916015, 0.1155645751953125, 0.11573654174804687, 0.11594445037841797, 0.11607039642333984, 0.1157232666015625, 0.11550310516357422, 0.11549900817871094, 0.11566694641113281, 0.11567616271972657, 0.11566902160644531, 0.1161082534790039, 0.11594137573242187, 0.11565158081054687, 0.11552153778076171, 0.11549702453613281, 0.11560646057128907, 0.11540275573730469, 0.11600281524658203, 0.11596390533447265, 0.11555123138427735, 0.11542118072509766, 0.11550617980957031, 0.11543654632568359, 0.11562700653076172, 0.11552051544189453, 0.11569356536865234, 0.11555020904541016, 0.11564236450195313, 0.11563622283935547, 0.11566694641113281, 0.11554611206054688, 0.11565567779541015, 0.11542425537109376, 0.11545600128173829, 0.11553587341308594, 0.11576729583740235, 0.11622911834716797, 0.11651174163818359, 0.11568844604492187, 0.11576831817626954, 0.11554713439941407, 0.11568025970458984, 0.11566694641113281, 0.11580620574951171, 0.11560345458984375, 0.11570893096923827, 0.24270541381835936, 0.11573554992675782, 0.1154734115600586, 0.115525634765625, 0.11553279876708984, 0.11555430603027343, 0.11576525115966797, 0.11567922973632813, 0.11552665710449218, 0.11537715148925781, 0.1155041275024414, 0.11540991973876953, 0.11538329315185547, 0.11551641845703126, 0.11542937469482421, 0.11544576263427735, 0.11548365020751954, 0.11621273803710938, 0.11560447692871094, 0.11579801940917969, 0.11566796875, 0.11572531127929687, 0.11551750183105469, 0.1155849609375, 0.11561062622070313, 0.11598028564453125, 0.11561676788330078, 0.11556147003173828, 0.11561676788330078, 0.11551436614990235, 0.11571302032470702, 0.11565875244140625, 0.11572223663330078, 0.11552973175048828, 0.11562290954589843, 0.11551747131347656, 0.11566793823242187, 0.11573760223388672, 0.11573760223388672, 0.11582463836669922, 0.1156147232055664, 0.11548880004882812, 0.1162034912109375, 0.11573350524902344, 0.11611344146728515, 0.11572425842285156, 0.11567616271972657, 0.11576217651367188, 0.11560352325439453, 0.11609388732910156, 0.11659980773925781, 0.11586048126220704, 0.11570381164550782, 0.11573248291015625, 0.11557170867919922, 0.11563827514648438, 0.11605709075927734, 0.1166714859008789, 0.1158123550415039, 0.11569971466064453, 0.11564543914794922, 0.11563629150390625, 0.11554707336425782, 0.24293376159667968, 0.1157570571899414, 0.11543551635742187, 0.11561164855957032, 0.11565161895751953, 0.11541910552978515, 0.11535257720947266, 0.11540790557861329, 0.1153371810913086, 0.11557478332519532, 0.11545394897460938, 0.11548365020751954, 0.11608576202392579, 0.11585740661621094, 0.1155758056640625, 0.11546419525146484, 0.11544882965087891, 0.11567922973632813, 0.11579596710205078, 0.11548467254638672, 0.11567820739746094, 0.11553075408935547, 0.11560447692871094, 0.11557683563232422, 0.11538841247558594, 0.11541709136962891, 0.11647283172607421, 0.11568946838378906, 0.11550924682617188, 0.11575603485107422, 0.11594035339355468, 0.11566387176513672, 0.11554713439941407, 0.11556761932373047, 0.11553997039794922, 0.11574578857421874, 0.1154959716796875, 0.11552867126464844, 0.11542527770996094, 0.11553689575195313, 0.11548262023925782, 0.11609190368652343, 0.11563314819335938, 0.11550105285644531, 0.11568844604492187, 0.1156341781616211, 0.11574681854248046, 0.11630079650878906, 0.11589119720458985, 0.1157027816772461, 0.11557170867919922, 0.11571609497070312, 0.11571916961669922, 0.11559321594238281, 0.1154549789428711, 0.1155389404296875, 0.11564543914794922, 0.11553587341308594, 0.11558604431152343, 0.11541913604736329, 0.11550822448730469, 0.11574272155761718, 0.11551026916503906]",tokens/s,8.507952767756032,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Salesforce/codegen-6B-nl,Salesforce/codegen-6B-nl,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,1400.639488,6094.848,0.0,5448.400896,5215.942144,s,10,5.915999877929687,0.5915999877929687,0.0028828720813889874,0.5905971069335938,0.5921576904296875,0.5961826965332031,0.5994027014160156,"[0.6002077026367187, 0.5907606201171876, 0.5909608764648437, 0.5905933227539063, 0.5902119140625, 0.5906008911132813, 0.5904852294921875, 0.5912632446289062, 0.5904329223632813, 0.590483154296875]",tokens/s,432.7248229923689,kWh,6.983380516370138e-06,3.826209227811299e-06,3.1648620090131334e-05,4.2458209834312766e-05,tokens/kWh,6029458.166017933,MB,1400.967168,6094.848,0.0,5448.400896,5405.644288,s,10,341.33091015625,34.133091015625,0.0089451948085196,34.130179687500004,34.14430546875,34.149687890625,34.153993828125,"[34.13034765625, 34.1351484375, 34.13001171875, 34.128140625, 34.12757421875, 34.12398828125, 34.13193359375, 34.1255859375, 34.143109375, 34.1550703125]",tokens/s,1.8457162280193342,kWh,0.00040300285582741096,0.00022088092240917615,0.0018324814267682385,0.002456365205004826,tokens/kWh,25647.65201511484,,s,629,346.1052627563478,0.5502468406301234,0.07016038395822202,0.5417257080078125,0.5423982788085937,0.54271630859375,1.1323166894531251,"[0.5417594604492187, 0.5414471435546875, 0.5411215209960938, 0.5414297485351562, 0.5414563598632812, 0.541576171875, 0.5412239379882813, 0.5414502563476562, 0.5416427612304687, 0.54096484375, 0.54148095703125, 0.5411840209960938, 0.5414339599609375, 0.5409760131835938, 0.5409024047851563, 0.5412208862304687, 0.5413017578125, 0.5412515869140625, 0.5410549926757813, 0.5412177734375, 0.5411133422851563, 0.54131201171875, 0.5422049560546875, 0.5416530151367187, 0.54127001953125, 0.5410795288085938, 0.5415239868164062, 0.5422069702148438, 0.5421731567382813, 0.5426155395507812, 0.542755859375, 0.541576171875, 0.5415362548828125, 0.5411993408203125, 0.5416642456054688, 0.541591552734375, 0.541427734375, 0.5414830322265625, 0.5413529663085938, 0.5414666137695312, 0.5418700561523437, 0.5421414184570312, 0.5417953491210937, 0.5417728271484376, 0.5417267456054687, 0.5425919799804687, 0.5420779418945313, 0.5423953857421875, 0.542466064453125, 0.5427732543945313, 0.5427783813476562, 0.5423319091796875, 0.5419212646484375, 0.5422335815429687, 0.5419622192382813, 0.542166015625, 0.542202880859375, 0.5419612426757813, 0.541928466796875, 0.5428971557617187, 0.5427886352539063, 0.542540771484375, 1.132921875, 0.5414379272460937, 0.541675537109375, 0.5412976684570312, 0.5413980102539062, 0.541432861328125, 0.5411819458007813, 0.5411235961914063, 0.5415782470703125, 0.5412136840820313, 0.5412894897460937, 0.5412925415039063, 0.5418301391601562, 0.541549560546875, 0.5421035766601563, 0.5417698974609375, 0.5415985717773437, 0.5415782470703125, 0.5415618286132813, 0.5418792724609375, 0.5416151123046875, 0.5421527099609375, 0.5418721313476562, 0.5414307861328125, 0.5414574584960937, 0.5414255981445313, 0.5418076171875, 0.5416806640625, 0.5416058959960938, 0.5413232421875, 0.5415167846679687, 0.5413140869140625, 0.5417481689453125, 0.5427220458984375, 0.5421270751953124, 0.541823974609375, 0.54205029296875, 0.5420144653320312, 0.5415424194335937, 0.541971435546875, 0.5416202392578126, 0.541264892578125, 0.5415219116210938, 0.541216796875, 0.5416058959960938, 0.541259765625, 0.5414993896484375, 0.5412925415039063, 0.5416744995117188, 0.5412996826171875, 0.5416222534179688, 0.5422120971679687, 0.5429841918945313, 0.542635009765625, 0.5428388061523437, 0.542835693359375, 0.5432268676757812, 0.5427363891601562, 0.5429586181640625, 0.5425602416992188, 0.5428838500976563, 0.5426534423828125, 0.5426380615234375, 1.1323843994140625, 0.5415637817382812, 0.5415362548828125, 0.541212646484375, 0.5413027954101562, 0.54141748046875, 0.54186083984375, 0.542003173828125, 0.5418147583007813, 0.5417257080078125, 0.5426544799804688, 0.5414871215820313, 0.5419520263671875, 0.5414297485351562, 0.5414932250976563, 0.5413765258789063, 0.5414635620117187, 0.5415690307617187, 0.5419008178710938, 0.5413099365234375, 0.5413130493164062, 0.5413457641601562, 0.5415126953125, 0.541191162109375, 0.541591552734375, 0.5423012084960938, 0.5414676513671876, 0.5414563598632812, 0.5417738037109375, 0.541896728515625, 0.5416048583984375, 0.5420687255859375, 0.5419673461914063, 0.5420953369140625, 0.541780029296875, 0.541761474609375, 0.5418792724609375, 0.5419478759765625, 0.54186083984375, 0.5423062744140625, 0.54225, 0.5419407348632812, 0.542171142578125, 0.5424015502929688, 0.5422223510742188, 0.5414348754882813, 0.5414635620117187, 0.54171337890625, 0.5415720825195313, 0.5418065795898438, 0.5415321655273437, 0.54148095703125, 0.5417645874023438, 0.54140625, 0.5415218505859375, 0.5416734619140625, 0.5417164916992188, 0.5417820434570313, 0.5417778930664062, 0.54166015625, 0.5416734619140625, 0.5428797607421875, 0.542571533203125, 1.1325224609375, 0.5419622192382813, 0.541770751953125, 0.5419089965820313, 0.5416908569335938, 0.5413734130859374, 0.5414747924804687, 0.5413980102539062, 0.5412987060546876, 0.5413621826171875, 0.5414256591796875, 0.5414717407226562, 0.5418414306640625, 0.5414390258789062, 0.5416283569335938, 0.5422162475585938, 0.5423297729492188, 0.5422459106445312, 0.5416038208007813, 0.5418384399414062, 0.5420675659179688, 0.5416325073242187, 0.5418936157226563, 0.54164892578125, 0.5416837158203125, 0.5414542846679687, 0.5416089477539062, 0.541623291015625, 0.541696044921875, 0.5414962768554688, 0.5420676879882812, 0.5414850463867188, 0.5416734619140625, 0.5414481811523437, 0.5424260864257813, 0.5415341796875, 0.5418352661132813, 0.5416427612304687, 0.5419017944335938, 0.5418506469726563, 0.5419898681640625, 0.541454345703125, 0.5417994384765625, 0.5414666137695312, 0.541697021484375, 0.5416581420898438, 0.5419612426757813, 0.5414061889648437, 0.5416857299804687, 0.5413396606445312, 0.5420892333984375, 0.5414430541992188, 0.5420545654296876, 0.5417859497070312, 0.5416704711914062, 0.5413395385742188, 0.5417564086914063, 0.5420462036132813, 0.5418311767578124, 0.5419110107421875, 0.5419161376953125, 0.5414850463867188, 0.541528076171875, 1.131968505859375, 0.5414400024414062, 0.5414932250976563, 0.5411287231445312, 0.541296630859375, 0.5412792358398437, 0.5417031860351562, 0.5415956420898438, 0.5414348754882813, 0.5413161010742188, 0.5413847045898438, 0.541760498046875, 0.5419417724609376, 0.5416417236328125, 0.5418035278320312, 0.54144921875, 0.54145947265625, 0.5413785400390625, 0.541380615234375, 0.5414000854492188, 0.5414727783203125, 0.5413887939453125, 0.5414727783203125, 0.5413324584960938, 0.5414451293945313, 0.54200830078125, 0.541663330078125, 0.5414224853515625, 0.5416151123046875, 0.5414993896484375, 0.54185986328125, 0.5417636108398437, 0.5419540405273438, 0.5416038208007813, 0.5418005981445313, 0.541447021484375, 0.541686767578125, 0.5417062377929688, 0.5416693725585937, 0.5417902221679688, 0.5416990966796875, 0.5416304931640625, 0.5417277221679687, 0.5417277221679687, 0.5418322143554688, 0.5419100341796875, 0.5417410888671875, 0.5418536987304687, 0.5424578247070313, 0.5422274780273437, 0.5421045532226563, 0.5419489135742187, 0.5420206298828125, 0.5418895263671875, 0.541897705078125, 0.5418690795898438, 0.5419857788085938, 0.5421117553710938, 0.54192333984375, 0.5418588256835938, 0.5421434936523437, 0.5418905639648437, 0.5424384155273437, 1.1324405517578124, 0.5413406982421874, 0.5414522705078125, 0.5422376708984376, 0.5428193359375, 0.541686767578125, 0.5414686889648438, 0.5414215698242187, 0.541285400390625, 0.5416571044921875, 0.5415997314453125, 0.5419386596679687, 0.5416673583984375, 0.541638671875, 0.5414850463867188, 0.5416683349609375, 0.5423175659179688, 0.5418322143554688, 0.5414451904296875, 0.5412515258789062, 0.5415239868164062, 0.542023681640625, 0.541918212890625, 0.5417359619140625, 0.541549560546875, 0.5413662719726563, 0.5415792846679688, 0.5414010620117188, 0.5414441528320313, 0.5412525634765625, 0.54166015625, 0.5412556762695313, 0.541591552734375, 0.5411932373046875, 0.541432861328125, 0.5416161499023437, 0.5414912109375, 0.5413294067382812, 0.541365234375, 0.5413304443359375, 0.5423472900390625, 0.541759521484375, 0.5416038208007813, 0.5411983642578125, 0.5415557250976563, 0.5414194946289063, 0.5415966796875, 0.54135498046875, 0.5417625732421875, 0.5416345825195312, 0.5418465576171875, 0.541486083984375, 0.5419673461914063, 0.541765625, 0.542076904296875, 0.5416171264648437, 0.5418035278320312, 0.5415731201171875, 0.541892578125, 0.54164892578125, 0.5422459106445312, 0.5417850952148437, 0.5420390625, 1.132142578125, 0.5416161499023437, 0.541454345703125, 0.5413867797851563, 0.5415997314453125, 0.541613037109375, 0.5415372924804688, 0.5415844116210937, 0.5418772583007813, 0.5416714477539063, 0.5416253662109375, 0.5415557250976563, 0.5419458618164062, 0.5417195434570312, 0.5419008178710938, 0.541591552734375, 0.5416642456054688, 0.5415946044921875, 0.5415925903320312, 0.5416468505859375, 0.541644775390625, 0.5415782470703125, 0.5415966796875, 0.541538330078125, 0.541675537109375, 0.5414788818359375, 0.5420431518554687, 0.5420267333984375, 0.5419776000976563, 0.5416171264648437, 0.541822998046875, 0.54211376953125, 0.5417615356445312, 0.5421096801757812, 0.5417891845703126, 0.5419161376953125, 0.541697021484375, 0.5416591186523437, 0.5419724731445312, 0.5416560668945313, 0.54164892578125, 0.5418291015625, 0.5418055419921874, 0.541601806640625, 0.5418721313476562, 0.54198681640625, 0.5417297973632812, 0.5416796264648438, 0.5417349243164062, 0.5415823364257812, 0.5425029296875, 0.542382080078125, 0.5421414184570312, 0.5416663208007813, 0.5417011108398437, 0.54160302734375, 0.5418053588867188, 0.5416949462890625, 0.5418045654296875, 0.5421281127929688, 0.5423175659179688, 0.5419632568359375, 0.542244873046875, 1.1331278076171876, 0.5414716186523437, 0.54160791015625, 0.54152294921875, 0.5418578491210938, 0.5414696655273438, 0.5420236206054687, 0.5413458862304688, 0.5417337646484375, 0.5414185180664063, 0.5415403442382812, 0.5418291015625, 0.5415823974609375, 0.5414583740234375, 0.541802490234375, 0.541475830078125, 0.5415629272460938, 0.5419693603515625, 0.5415751953125, 0.5413673095703125, 0.5415239868164062, 0.5413990478515625, 0.5414297485351562, 0.5419171752929688, 0.542044189453125, 0.5415536499023438, 0.5416273803710937, 0.5416406860351562, 0.5414400024414062, 0.541470703125, 0.5419960327148438, 0.5414901733398437, 0.5415792846679688, 0.5415465087890625, 0.5416714477539063, 0.5413161010742188, 0.5425889282226563, 0.5413949584960938, 0.5415833740234375, 0.541106201171875, 0.5417615356445312, 0.5417666625976563, 0.541760498046875, 0.5414307861328125, 0.5415894775390625, 0.5413898315429687, 0.5415833740234375, 0.5412874145507812, 0.5417349243164062, 0.5415741577148437, 0.5416611938476562, 0.5413416748046875, 0.5422161865234375, 0.5416673583984375, 0.5423308715820313, 0.5420431518554687, 0.5420534057617188, 0.5417984008789063, 0.5419888916015625, 0.5413898315429687, 0.5425121459960938, 0.541970458984375, 0.5419970092773437, 1.13246826171875, 0.5414359130859375, 0.5415659790039062, 0.54150244140625, 0.5414912719726562, 0.5414020385742188, 0.541581298828125, 0.5415567626953125, 0.5416243286132812, 0.5418076171875, 0.5416089477539062, 0.5414696655273438, 0.5415587768554687, 0.541931640625, 0.542152587890625, 0.541929443359375, 0.541939697265625, 0.54158642578125, 0.5415321655273437, 0.5415823364257812, 0.5417267456054687, 0.5416222534179688, 0.54167041015625, 0.54148095703125, 0.5416376342773438, 0.5415116577148438, 0.5416591186523437, 0.5424230346679687, 0.5425469360351562, 0.5417778930664062, 0.5417523193359375, 0.5418987426757812, 0.5419951171875, 0.5423696899414062, 0.5424568481445312, 0.5426206665039063, 0.542086181640625, 0.541823974609375, 0.54186083984375, 0.5420349731445312, 0.5424475708007812, 0.5419776000976563, 0.5422203369140625, 0.5420625, 0.542244873046875, 0.5426073608398437, 0.5429688110351563, 0.5421936645507812, 0.542075927734375, 0.5420676879882812, 0.5423175659179688, 0.5427466430664063, 0.5428531494140625, 0.542118896484375, 0.5421475830078125, 0.542002197265625, 0.5421066284179688, 0.5419857788085938, 0.5419765625, 0.5419990844726562, 0.5420123901367188, 0.5418936157226563, 0.5418700561523437, 1.1330999755859374, 0.5421752319335937, 0.5420472412109375, 0.5415352172851563, 0.541517822265625, 0.5416509399414062, 0.54187109375, 0.5422120971679687, 0.5417625732421875, 0.5419806518554687, 0.5415413818359375, 0.54137548828125, 0.5424609375, 0.5421240234375, 0.54167041015625, 0.5416581420898438, 0.5418803100585937, 0.5415997314453125, 0.5424701538085938, 0.5428551635742187, 0.5424916381835938, 0.542624755859375, 0.542044189453125, 0.54175537109375, 0.541865966796875, 0.54215576171875, 0.5419192504882813, 0.5422418212890625, 0.54240869140625, 0.5427630004882813, 0.543088623046875, 0.5421854858398437, 0.5420318603515625, 0.5419612426757813, 0.5423749389648438, 0.541897705078125, 0.5418147583007813, 0.5428131713867187, 0.542508056640625, 0.5420390625, 0.5422202758789062, 0.542045166015625, 0.542202880859375, 0.5423974609375, 0.5422069702148438, 0.541749267578125, 0.5420185546875, 0.5419468994140625, 0.5423267822265625, 0.5417062377929688, 0.5420277709960938, 0.5420349731445312, 0.5422673950195313, 0.5421270751953124, 0.5421915893554687, 0.5427077026367187, 0.5426472778320313, 0.542382080078125, 0.54236572265625, 0.542023681640625, 0.5423565063476562, 0.542160888671875, 0.5429483642578125]",tokens/s,1.8173661821571478,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-70m,EleutherAI/pythia-70m,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,1298.415616,872.93952,0.0,226.492416,184.397824,s,11,0.17165715312957763,0.015605195739052512,0.0007607833862353127,0.01534540843963623,0.01575004768371582,0.016863200187683107,0.017753722190856934,"[0.01797635269165039, 0.015421407699584961, 0.015347392082214355, 0.015315903663635253, 0.015256128311157226, 0.01575004768371582, 0.01534540843963623, 0.015287008285522461, 0.015340383529663087, 0.01527171230316162, 0.01534540843963623]",tokens/s,16404.792626814135,kWh,1.8500959462776245e-07,1.0137606156642736e-07,3.4862617005003736e-07,6.350118262442272e-07,tokens/kWh,403142098.1780924,MB,1298.743296,872.93952,0.0,226.492416,197.933568,s,11,10.171151123046876,0.9246501020951705,0.010660233315434522,0.920117919921875,0.9293056640625,0.9430220336914062,0.9539951293945312,"[0.9567384033203125, 0.9293056640625, 0.9245167846679687, 0.9185679931640625, 0.920117919921875, 0.9228417358398437, 0.919580810546875, 0.9193172607421874, 0.9187877807617187, 0.923688232421875, 0.9176885375976562]",tokens/s,68.13388097535262,kWh,1.1024352834572037e-05,6.040874968975529e-06,1.993414258140665e-05,3.6999370384954216e-05,tokens/kWh,1702731.6774454878,,s,692,10.307773435592651,0.014895626352012501,0.0018980452913926306,0.014577664375305176,0.01503354845046997,0.015412480163574216,0.030279741020202637,"[0.016365568161010743, 0.016268287658691406, 0.01589350414276123, 0.01510707187652588, 0.015166463851928711, 0.015711232185363768, 0.015400959968566894, 0.015467519760131837, 0.015654911994934084, 0.015433728218078613, 0.015542304039001465, 0.015136799812316894, 0.015172543525695801, 0.015097855567932129, 0.015466496467590332, 0.016034816741943358, 0.015383551597595215, 0.015619071960449218, 0.015451135635375977, 0.015178751945495606, 0.015625215530395507, 0.015426560401916504, 0.015222847938537598, 0.015156160354614258, 0.015565823554992676, 0.016327680587768553, 0.015833087921142578, 0.01530675220489502, 0.015086591720581055, 0.01507737636566162, 0.01501088047027588, 0.01481721591949463, 0.015222784042358398, 0.015755264282226563, 0.016021503448486327, 0.015339520454406739, 0.015148032188415527, 0.01509171199798584, 0.015104000091552734, 0.015050751686096191, 0.015105024337768554, 0.014737407684326171, 0.014621696472167968, 0.014722047805786133, 0.014584896087646484, 0.014627776145935058, 0.014577664375305176, 0.014598143577575684, 0.014589952468872071, 0.014524415969848633, 0.014585856437683106, 0.01456332778930664, 0.014573568344116212, 0.014537728309631348, 0.014552063941955566, 0.014576640129089356, 0.014576640129089356, 0.014548992156982422, 0.01457049560546875, 0.014616576194763184, 0.014621696472167968, 0.014593024253845215, 0.030418943405151368, 0.014628864288330079, 0.014632960319519044, 0.0145797119140625, 0.01460531234741211, 0.014616576194763184, 0.01458073616027832, 0.014573568344116212, 0.014584832191467285, 0.014578687667846679, 0.014615551948547363, 0.014621696472167968, 0.01458790397644043, 0.014673919677734374, 0.014622719764709472, 0.01457151985168457, 0.014611455917358398, 0.0145797119140625, 0.014619647979736328, 0.014697471618652343, 0.014582783699035644, 0.014603263854980468, 0.01460223960876465, 0.01459404754638672, 0.014568448066711426, 0.014610431671142577, 0.014583807945251465, 0.014577664375305176, 0.014595071792602539, 0.014558208465576173, 0.01457254409790039, 0.014645248413085938, 0.014598143577575684, 0.015670271873474122, 0.015376383781433106, 0.015114239692687988, 0.015088640213012695, 0.015019007682800293, 0.015113216400146484, 0.015021056175231933, 0.014995455741882324, 0.014962688446044922, 0.014998527526855468, 0.014962688446044922, 0.014985216140747071, 0.014946304321289062, 0.01500876808166504, 0.014965760231018066, 0.015024191856384277, 0.014966719627380371, 0.014658559799194336, 0.014590975761413574, 0.014565376281738282, 0.01461350440979004, 0.014582783699035644, 0.015023103713989258, 0.01509887981414795, 0.014726143836975097, 0.014541824340820313, 0.0145797119140625, 0.014606335639953612, 0.014665727615356445, 0.01459609603881836, 0.030462976455688476, 0.014574591636657714, 0.014746623992919922, 0.015026176452636719, 0.015056960105895996, 0.015804351806640624, 0.015242239952087403, 0.015034367561340332, 0.015006719589233398, 0.014931967735290527, 0.01477734375, 0.0145797119140625, 0.014584832191467285, 0.014592000007629394, 0.014552063941955566, 0.014543871879577636, 0.014574591636657714, 0.014581760406494141, 0.014575615882873535, 0.014657535552978516, 0.014608384132385254, 0.014558208465576173, 0.014607359886169433, 0.014606335639953612, 0.01459609603881836, 0.014619647979736328, 0.014576640129089356, 0.014553088188171387, 0.014573568344116212, 0.014575615882873535, 0.014550016403198243, 0.01458892822265625, 0.014589952468872071, 0.014590975761413574, 0.014647295951843262, 0.01455513572692871, 0.014576640129089356, 0.014583807945251465, 0.014652416229248047, 0.01460223960876465, 0.014617600440979005, 0.014598143577575684, 0.0145797119140625, 0.01457254409790039, 0.014576640129089356, 0.014589952468872071, 0.01458790397644043, 0.014652416229248047, 0.014550016403198243, 0.01457049560546875, 0.01458892822265625, 0.014541824340820313, 0.0145797119140625, 0.0145633602142334, 0.014960607528686524, 0.014947327613830566, 0.014850111961364746, 0.014465984344482421, 0.014386176109313965, 0.014425087928771972, 0.014609408378601075, 0.014697471618652343, 0.014664704322814942, 0.030324735641479493, 0.014525440216064453, 0.014568448066711426, 0.014598143577575684, 0.01457049560546875, 0.014592000007629394, 0.014541824340820313, 0.014660639762878418, 0.014574560165405273, 0.014515199661254884, 0.014586879730224609, 0.014577664375305176, 0.01461350440979004, 0.014574591636657714, 0.01455513572692871, 0.014520319938659668, 0.014518272399902344, 0.01457151985168457, 0.014554112434387208, 0.01456332778930664, 0.014540800094604492, 0.01457049560546875, 0.01456332778930664, 0.014504960060119629, 0.014553088188171387, 0.014619647979736328, 0.014583807945251465, 0.014557184219360352, 0.014566399574279786, 0.014529536247253418, 0.014600192070007324, 0.014612480163574219, 0.01456332778930664, 0.014575615882873535, 0.014627840042114258, 0.01467084789276123, 0.014575615882873535, 0.014560256004333496, 0.014663680076599121, 0.014564352035522461, 0.014524415969848633, 0.01454694366455078, 0.01458073616027832, 0.014548992156982422, 0.01452444839477539, 0.014602208137512207, 0.01455513572692871, 0.01456332778930664, 0.014532608032226562, 0.01461350440979004, 0.014622719764709472, 0.014517248153686524, 0.01459404754638672, 0.014558208465576173, 0.014556159973144531, 0.014548992156982422, 0.01457049560546875, 0.014508031845092773, 0.014590975761413574, 0.014540800094604492, 0.014600192070007324, 0.01469644832611084, 0.014692352294921876, 0.030276607513427735, 0.01461350440979004, 0.014553088188171387, 0.014508031845092773, 0.014564352035522461, 0.014553088188171387, 0.014565376281738282, 0.014566399574279786, 0.014537728309631348, 0.014530559539794922, 0.01459609603881836, 0.014569472312927247, 0.014567423820495605, 0.014508031845092773, 0.01455513572692871, 0.014628864288330079, 0.014535679817199706, 0.01458790397644043, 0.014557184219360352, 0.014535712242126466, 0.014557151794433594, 0.014535679817199706, 0.014534655570983887, 0.014531583786010742, 0.014545920372009278, 0.014566399574279786, 0.014538751602172852, 0.014553088188171387, 0.01456332778930664, 0.014638079643249511, 0.014538751602172852, 0.014615551948547363, 0.014623744010925293, 0.014533663749694824, 0.014525407791137696, 0.014531583786010742, 0.0145797119140625, 0.01457049560546875, 0.014547967910766601, 0.014557184219360352, 0.014577664375305176, 0.014550016403198243, 0.014560256004333496, 0.014547967910766601, 0.01447935962677002, 0.014482432365417481, 0.014364671707153321, 0.014386176109313965, 0.014722047805786133, 0.015221759796142579, 0.01508351993560791, 0.015017984390258789, 0.014990336418151855, 0.014946304321289062, 0.014996479988098145, 0.014521344184875488, 0.014540800094604492, 0.014592000007629394, 0.014558208465576173, 0.014535679817199706, 0.01457049560546875, 0.014556159973144531, 0.014526464462280274, 0.031153152465820313, 0.014898176193237305, 0.01458790397644043, 0.014568448066711426, 0.014693375587463378, 0.014625791549682618, 0.015431679725646973, 0.014708736419677734, 0.014601216316223145, 0.01455513572692871, 0.014576640129089356, 0.014567423820495605, 0.014589952468872071, 0.014566399574279786, 0.014551039695739745, 0.01480191993713379, 0.014784511566162109, 0.015605759620666505, 0.015138815879821778, 0.014937088012695313, 0.014920703887939453, 0.014586879730224609, 0.014550047874450684, 0.014574624061584473, 0.01459712028503418, 0.014562239646911621, 0.014526464462280274, 0.014525440216064453, 0.014536704063415527, 0.014520319938659668, 0.014515199661254884, 0.014625791549682618, 0.014533632278442383, 0.014556159973144531, 0.014573568344116212, 0.014589952468872071, 0.01458073616027832, 0.014565376281738282, 0.014610431671142577, 0.01459609603881836, 0.01466982364654541, 0.014590975761413574, 0.014558208465576173, 0.014566399574279786, 0.014538751602172852, 0.014564352035522461, 0.014567423820495605, 0.014530559539794922, 0.014538751602172852, 0.014535679817199706, 0.014551039695739745, 0.01457151985168457, 0.014675968170166016, 0.014567423820495605, 0.014585856437683106, 0.014559231758117675, 0.014527487754821777, 0.014577664375305176, 0.014561280250549317, 0.014562303543090821, 0.014578687667846679, 0.014515199661254884, 0.01454086399078369, 0.03033900833129883, 0.014519295692443847, 0.014643199920654297, 0.014578687667846679, 0.014610431671142577, 0.014585856437683106, 0.014574591636657714, 0.014586879730224609, 0.01457254409790039, 0.014628864288330079, 0.01458790397644043, 0.01529856014251709, 0.014918656349182128, 0.014584832191467285, 0.014532608032226562, 0.014585920333862305, 0.014574527740478516, 0.014573568344116212, 0.014543871879577636, 0.014592063903808595, 0.01458886432647705, 0.014574591636657714, 0.01458892822265625, 0.014633983612060546, 0.014540800094604492, 0.014567423820495605, 0.014564352035522461, 0.014526464462280274, 0.014562303543090821, 0.01458073616027832, 0.014523391723632812, 0.014614527702331542, 0.014553088188171387, 0.014541824340820313, 0.014567423820495605, 0.014635007858276367, 0.014566399574279786, 0.01459404754638672, 0.014535679817199706, 0.014558208465576173, 0.014565376281738282, 0.01458073616027832, 0.014585856437683106, 0.014607359886169433, 0.014540800094604492, 0.014544896125793457, 0.014617600440979005, 0.014566399574279786, 0.01457254409790039, 0.014559231758117675, 0.014561280250549317, 0.01458073616027832, 0.014527551651000977, 0.014563263893127442, 0.014543871879577636, 0.014548992156982422, 0.014582783699035644, 0.014598143577575684, 0.014575615882873535, 0.014552063941955566, 0.014589952468872071, 0.01457254409790039, 0.014573568344116212, 0.030249984741210937, 0.014523391723632812, 0.014586943626403809, 0.014542783737182617, 0.014595104217529296, 0.014566368103027343, 0.014517248153686524, 0.014551039695739745, 0.014527487754821777, 0.014636032104492188, 0.014616576194763184, 0.014551072120666505, 0.014565343856811524, 0.014598143577575684, 0.0146626558303833, 0.014781439781188965, 0.015035391807556153, 0.014568448066711426, 0.01457254409790039, 0.014618623733520507, 0.014560256004333496, 0.014574591636657714, 0.014547967910766601, 0.014540800094604492, 0.014562303543090821, 0.014603263854980468, 0.01456332778930664, 0.014577664375305176, 0.014547967910766601, 0.01459404754638672, 0.01458790397644043, 0.014542847633361817, 0.014595071792602539, 0.01458790397644043, 0.014665792465209961, 0.014555071830749512, 0.014537728309631348, 0.014530559539794922, 0.014581760406494141, 0.014584832191467285, 0.014564352035522461, 0.014573568344116212, 0.014548992156982422, 0.014700544357299805, 0.014585856437683106, 0.014574591636657714, 0.014560256004333496, 0.014617600440979005, 0.014556159973144531, 0.014578687667846679, 0.01455513572692871, 0.01466982364654541, 0.014557184219360352, 0.014561280250549317, 0.014525440216064453, 0.014565376281738282, 0.014586879730224609, 0.0145664644241333, 0.014582719802856445, 0.014539775848388671, 0.014560256004333496, 0.01460428810119629, 0.014592000007629394, 0.030311424255371092, 0.014531583786010742, 0.014569472312927247, 0.014636032104492188, 0.014564352035522461, 0.014538751602172852, 0.01458892822265625, 0.014531583786010742, 0.014553088188171387, 0.014525440216064453, 0.014530624389648437, 0.014523327827453613, 0.014575615882873535, 0.014558208465576173, 0.014516223907470703, 0.014637056350708008, 0.014518272399902344, 0.014577664375305176, 0.01495961570739746, 0.014617600440979005, 0.014619711875915528, 0.014569408416748047, 0.014639103889465332, 0.014576640129089356, 0.014560256004333496, 0.014568448066711426, 0.0145797119140625, 0.014550016403198243, 0.014577664375305176, 0.014582783699035644, 0.014611488342285156, 0.014558176040649413, 0.014578687667846679, 0.014569472312927247, 0.014663680076599121, 0.014584832191467285, 0.014565376281738282, 0.01456332778930664, 0.014550016403198243, 0.014538751602172852, 0.014530559539794922, 0.014534655570983887, 0.014559231758117675, 0.01455513572692871, 0.014628864288330079, 0.014585856437683106, 0.014578687667846679, 0.014625791549682618, 0.014540800094604492, 0.014582783699035644, 0.014582783699035644, 0.014595071792602539, 0.014568448066711426, 0.014524415969848633, 0.014545920372009278, 0.014603263854980468, 0.01457049560546875, 0.014561280250549317, 0.014623744010925293, 0.014557184219360352, 0.014583807945251465, 0.014567487716674805, 0.014582719802856445, 0.03019980812072754, 0.014551039695739745, 0.01457151985168457, 0.014541824340820313, 0.014585856437683106, 0.01539891242980957, 0.015851519584655763, 0.014816255569458007, 0.014586879730224609, 0.014638079643249511, 0.014639103889465332, 0.014531583786010742, 0.014663680076599121, 0.014520319938659668, 0.014543871879577636, 0.014617600440979005, 0.014495743751525878, 0.01458073616027832, 0.014540800094604492, 0.014554112434387208, 0.014567423820495605, 0.01455513572692871, 0.014504960060119629, 0.01479475212097168, 0.015308799743652344, 0.014611455917358398, 0.014626815795898437, 0.014567423820495605, 0.014535679817199706, 0.01460223960876465, 0.01458892822265625, 0.014566399574279786, 0.014566399574279786, 0.014557184219360352, 0.014724096298217774, 0.014551039695739745, 0.014535712242126466, 0.014597087860107423, 0.014573599815368652, 0.014573535919189452, 0.014564352035522461, 0.01457049560546875, 0.014740480422973632, 0.014905344009399414, 0.014684160232543946, 0.014547967910766601, 0.014768128395080566, 0.015746047973632812, 0.014673919677734374, 0.014562303543090821, 0.014558208465576173, 0.014558208465576173, 0.01455513572692871, 0.014536704063415527, 0.01458790397644043, 0.014529536247253418, 0.014589952468872071, 0.014554112434387208, 0.014553088188171387, 0.014586879730224609, 0.014547967910766601, 0.014573599815368652, 0.014557151794433594, 0.03037593650817871, 0.014526464462280274, 0.014798848152160645, 0.014523391723632812, 0.01459404754638672, 0.014567423820495605, 0.014531583786010742, 0.014575615882873535, 0.014536704063415527, 0.014520319938659668, 0.014532608032226562, 0.014419967651367188, 0.014387200355529785, 0.01439027214050293, 0.014443519592285157, 0.014552063941955566, 0.014498815536499024, 0.014529536247253418, 0.014651391983032227, 0.014534655570983887, 0.014548992156982422, 0.014628864288330079, 0.014568448066711426, 0.01458790397644043, 0.01457151985168457, 0.014544896125793457, 0.014608384132385254, 0.014728192329406739, 0.014638079643249511, 0.014592000007629394, 0.014573568344116212, 0.014636032104492188, 0.014647295951843262, 0.014457856178283691, 0.014402560234069824, 0.014632960319519044, 0.014524415969848633, 0.01457049560546875, 0.01459712028503418, 0.014574591636657714, 0.014561280250549317, 0.01458790397644043, 0.014598143577575684, 0.014618623733520507, 0.014533632278442383, 0.014523391723632812, 0.014573568344116212, 0.01456332778930664, 0.01457151985168457, 0.01458073616027832, 0.014542847633361817, 0.014547967910766601, 0.014568448066711426, 0.014557184219360352, 0.014553088188171387, 0.014541824340820313, 0.014518272399902344, 0.014573568344116212, 0.014523455619812012, 0.014740415573120117, 0.014707776069641113, 0.014453696250915528, 0.014403583526611329]",tokens/s,67.13379997376835,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-110B,Qwen/Qwen1.5-110B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 32.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciCoder-1b,Deci/DeciCoder-1b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 68, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1164, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 968, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 713, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) +TypeError: DeciCoderAttention.forward() got an unexpected keyword argument 'cache_position' + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/recurrentgemma-2b,google/recurrentgemma-2b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/recurrentgemma-2b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aaa3e-125f61b1450fdd331a65c1d8;e0614d6c-abd0-498c-b212-f8d285a40174) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/recurrentgemma-2b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/recurrentgemma-2b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,internlm/internlm-20b,internlm/internlm-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,4294.987776,14621.868032,0.0,13975.420928,13366.068224,s,10,16.495098876953126,1.6495098876953125,0.0005626635207808348,1.6493680419921875,1.6498059448242186,1.6504339782714843,1.650936405029297,"[1.65106201171875, 1.6493150634765625, 1.6490108642578125, 1.64951513671875, 1.6496663818359374, 1.6489195556640626, 1.649384521484375, 1.6492862548828124, 1.6495875244140625, 1.6493515625]",tokens/s,155.19761470340868,kWh,1.9482200029823513e-05,1.067632958434842e-05,9.151657321320173e-05,0.00012167510282737365,tokens/kWh,2103963.7037595077,MB,4294.987776,14621.868032,0.0,13975.420928,13814.414848,s,10,977.7317578124998,97.77317578125,0.006501026483324052,97.774078125,97.7795859375,97.78208203125,97.78407890625,"[97.7740859375, 97.76746875, 97.7754296875, 97.784578125, 97.7782421875, 97.7740703125, 97.77903125, 97.76209375, 97.764875, 97.7718828125]",tokens/s,0.6443485086436308,kWh,0.0011542207027475041,0.0006326151578875942,0.005356944729996799,0.007143780590631898,tokens/kWh,8818.859874086276,,s,629,991.0235808105483,1.5755541825286914,0.19600214423351972,1.55186181640625,1.5529105224609374,1.5533291259765625,3.2017358984375,"[1.5526031494140624, 1.5511746826171875, 1.5512442626953125, 1.5510794677734374, 1.55169482421875, 1.5518966064453126, 1.5526973876953125, 1.5532349853515626, 1.55080810546875, 1.55123095703125, 1.5509954833984374, 1.5513375244140626, 1.551034423828125, 1.552995361328125, 1.5513057861328126, 1.55249462890625, 1.5510445556640624, 1.551331298828125, 1.5506575927734374, 1.55245263671875, 1.5513251953125, 1.5512381591796875, 1.551388671875, 1.5524095458984375, 1.5524761962890625, 1.5533260498046875, 1.553095703125, 1.5521700439453125, 1.551240234375, 1.551688720703125, 1.5509698486328125, 1.55129443359375, 1.5527669677734375, 1.5515802001953125, 1.551595458984375, 1.5514736328125, 1.552204833984375, 1.5519498291015625, 1.55194677734375, 1.55318994140625, 1.55307421875, 1.5523338623046874, 1.5521546630859375, 1.55177978515625, 1.5524197998046876, 1.5533404541015625, 1.5529237060546874, 1.5524351806640626, 1.55150439453125, 1.55194775390625, 1.552548828125, 1.5523931884765625, 1.55225390625, 1.5514603271484375, 1.5512484130859374, 1.5517244873046876, 1.5518822021484375, 1.552468017578125, 1.55346630859375, 1.5518555908203124, 1.5518289794921876, 1.5518668212890625, 3.202376708984375, 1.55222216796875, 1.5525550537109376, 1.551751220703125, 1.5510302734375, 1.55141943359375, 1.5515596923828125, 1.5518863525390625, 1.5530260009765624, 1.5516702880859374, 1.5516968994140625, 1.55188330078125, 1.55222119140625, 1.5524822998046874, 1.5535472412109375, 1.552794677734375, 1.550878662109375, 1.551520751953125, 1.55161083984375, 1.5515064697265626, 1.5530260009765624, 1.5537049560546874, 1.5521280517578124, 1.5518392333984374, 1.5516507568359375, 1.5517174072265625, 1.5512647705078124, 1.5524556884765626, 1.5519549560546875, 1.552384033203125, 1.5521689453125, 1.5524669189453124, 1.5522928466796875, 1.5529041748046875, 1.552691162109375, 1.55226318359375, 1.552175048828125, 1.5517655029296875, 1.5517972412109375, 1.55106201171875, 1.5525765380859375, 1.551730712890625, 1.5510968017578124, 1.550551025390625, 1.5507015380859375, 1.5512001953125, 1.5515802001953125, 1.5522396240234375, 1.5507763671875, 1.5507752685546874, 1.551515625, 1.551489990234375, 1.5510538330078125, 1.552689208984375, 1.5505909423828126, 1.55152587890625, 1.551647705078125, 1.5511357421875, 1.5508541259765625, 1.5514869384765626, 1.5528365478515624, 1.5521771240234374, 1.5512657470703124, 3.201815673828125, 1.5507783203125, 1.5535196533203126, 1.55132421875, 1.551177734375, 1.55173681640625, 1.5521044921875, 1.5513375244140626, 1.5526204833984374, 1.5516661376953125, 1.5514337158203124, 1.551310791015625, 1.5512391357421875, 1.5521761474609375, 1.5526307373046875, 1.553349609375, 1.551321044921875, 1.5509493408203125, 1.5514971923828125, 1.5512432861328125, 1.551467529296875, 1.552627685546875, 1.552321533203125, 1.5515975341796875, 1.5512330322265624, 1.55135595703125, 1.5512811279296874, 1.55175732421875, 1.5532247314453125, 1.5510487060546876, 1.55163232421875, 1.551236083984375, 1.55205224609375, 1.551705078125, 1.5527781982421875, 1.55230517578125, 1.5521064453125, 1.5526748046875, 1.551935546875, 1.5518392333984374, 1.553038330078125, 1.552990234375, 1.5518115234375, 1.551340576171875, 1.5512269287109375, 1.5518863525390625, 1.551740966796875, 1.553649658203125, 1.551879150390625, 1.5528212890625, 1.552669677734375, 1.5528980712890625, 1.5524495849609374, 1.5538206787109374, 1.55340185546875, 1.5514490966796874, 1.55186181640625, 1.5515064697265626, 1.5515709228515624, 1.5511982421875, 1.5528519287109375, 1.5516334228515625, 1.5513446044921875, 3.2030166015625, 1.55180029296875, 1.552606201171875, 1.5531837158203126, 1.551551513671875, 1.55184130859375, 1.5515115966796875, 1.551283203125, 1.551794189453125, 1.5518525390625, 1.5510743408203125, 1.551730712890625, 1.551551513671875, 1.5510927734375, 1.550951416015625, 1.5540325927734375, 1.55168359375, 1.5516273193359376, 1.5512708740234376, 1.5510076904296874, 1.55110302734375, 1.552822265625, 1.5531990966796876, 1.551572021484375, 1.552153564453125, 1.5514736328125, 1.5518802490234376, 1.551657958984375, 1.5528406982421874, 1.5522344970703126, 1.5521812744140624, 1.5515064697265626, 1.55281103515625, 1.55241064453125, 1.552889892578125, 1.5524730224609375, 1.5517911376953124, 1.5517276611328126, 1.5515074462890626, 1.550802978515625, 1.552126953125, 1.55321240234375, 1.5522139892578124, 1.5518084716796876, 1.5521444091796874, 1.5523614501953125, 1.55158837890625, 1.5536046142578126, 1.5520296630859376, 1.5530731201171875, 1.5527608642578126, 1.552880615234375, 1.5533311767578124, 1.5541422119140624, 1.55335888671875, 1.551869873046875, 1.55161083984375, 1.5517900390625, 1.5514500732421874, 1.551742919921875, 1.553448974609375, 1.551916015625, 1.552142333984375, 3.20272802734375, 1.5516005859375, 1.55409814453125, 1.552426025390625, 1.5515802001953125, 1.5510958251953124, 1.5515125732421875, 1.55104052734375, 1.5517767333984376, 1.5526820068359375, 1.5530455322265626, 1.5519539794921875, 1.5517962646484376, 1.552195556640625, 1.5515853271484374, 1.5515821533203125, 1.5523768310546875, 1.5513426513671875, 1.551593505859375, 1.5516016845703124, 1.5524136962890625, 1.551958984375, 1.5531632080078126, 1.5515289306640625, 1.55102099609375, 1.5516641845703125, 1.5517286376953126, 1.5518515625, 1.5519652099609376, 1.5515596923828125, 1.55163134765625, 1.5519027099609375, 1.5522764892578125, 1.5519969482421876, 1.5524884033203126, 1.5527935791015626, 1.5509124755859376, 1.55156884765625, 1.551499267578125, 1.5518935546875, 1.551709228515625, 1.552568359375, 1.552501708984375, 1.5518760986328124, 1.5515699462890624, 1.551488037109375, 1.5521546630859375, 1.5521934814453124, 1.5524454345703125, 1.553112060546875, 1.5525970458984375, 1.551457275390625, 1.551424560546875, 1.5510076904296874, 1.5529554443359375, 1.5528099365234376, 1.552541748046875, 1.5528201904296874, 1.5534459228515625, 1.5517440185546876, 1.5520296630859376, 1.5527730712890624, 1.55129345703125, 3.201416259765625, 1.5519610595703126, 1.551810546875, 1.5517440185546876, 1.551605712890625, 1.5520235595703125, 1.5515657958984375, 1.552027587890625, 1.551488037109375, 1.5524935302734375, 1.5531314697265626, 1.5514276123046875, 1.5520655517578126, 1.5519825439453125, 1.551853515625, 1.5516712646484374, 1.5525252685546875, 1.551730712890625, 1.5519703369140625, 1.55103125, 1.55213720703125, 1.552764892578125, 1.553755126953125, 1.5528775634765626, 1.5516845703125, 1.5517542724609374, 1.5515330810546875, 1.5523287353515625, 1.552110595703125, 1.5518494873046875, 1.5517244873046876, 1.55184228515625, 1.551310791015625, 1.5519027099609375, 1.5507896728515624, 1.552606201171875, 1.5512637939453124, 1.5521474609375, 1.5516856689453125, 1.551458251953125, 1.5516610107421875, 1.552607177734375, 1.5517041015625, 1.5525919189453126, 1.55161083984375, 1.552396240234375, 1.550496826171875, 1.551563720703125, 1.5521248779296875, 1.5515289306640625, 1.5516282958984375, 1.5522191162109376, 1.5513795166015625, 1.551456298828125, 1.5522652587890624, 1.5525396728515626, 1.5528642578125, 1.5518760986328124, 1.5516190185546874, 1.5520552978515625, 1.5518668212890625, 1.5532369384765625, 1.552636962890625, 3.202765869140625, 1.5513466796875, 1.553580078125, 1.5520194091796875, 1.5519456787109376, 1.5510753173828125, 1.5518084716796876, 1.55133544921875, 1.5519498291015625, 1.5519241943359374, 1.5514061279296876, 1.551531005859375, 1.5519334716796875, 1.5514398193359376, 1.5528131103515626, 1.55154638671875, 1.55287451171875, 1.552290771484375, 1.5517265625, 1.550793701171875, 1.5515279541015625, 1.5525499267578124, 1.5527659912109375, 1.5533721923828125, 1.5519078369140624, 1.55173583984375, 1.5521668701171876, 1.5522508544921876, 1.5519713134765625, 1.5531519775390625, 1.5524608154296875, 1.551698974609375, 1.552278564453125, 1.55198974609375, 1.5515361328125, 1.55272705078125, 1.551920166015625, 1.552759765625, 1.5514500732421874, 1.5516590576171876, 1.551510498046875, 1.5519405517578124, 1.552752685546875, 1.55184130859375, 1.55072509765625, 1.5517808837890625, 1.5512965087890624, 1.552079833984375, 1.552720947265625, 1.5514542236328126, 1.553217529296875, 1.5522928466796875, 1.5522979736328124, 1.551899658203125, 1.551836181640625, 1.5530526123046875, 1.55264306640625, 1.552206787109375, 1.55278955078125, 1.5506513671875, 1.5514705810546876, 1.552828369140625, 1.5512606201171875, 3.20153076171875, 1.5501854248046876, 1.5515657958984375, 1.5514951171875, 1.552109619140625, 1.552099365234375, 1.552288818359375, 1.5523583984375, 1.5526461181640625, 1.5514920654296875, 1.5523031005859376, 1.5516446533203125, 1.55148291015625, 1.5517808837890625, 1.5515330810546875, 1.5520235595703125, 1.5527147216796875, 1.5511695556640626, 1.552215087890625, 1.5506851806640625, 1.5511387939453125, 1.5506053466796874, 1.55132421875, 1.5518238525390624, 1.551515625, 1.5510947265625, 1.5512073974609375, 1.5515657958984375, 1.5510599365234374, 1.5514132080078125, 1.551810546875, 1.55152490234375, 1.55114697265625, 1.5516077880859376, 1.5506268310546876, 1.5511910400390625, 1.5526318359375, 1.552217041015625, 1.551873046875, 1.5515443115234375, 1.551283203125, 1.551500244140625, 1.55306494140625, 1.5515596923828125, 1.5516978759765625, 1.5517235107421874, 1.551678466796875, 1.553006591796875, 1.5524617919921875, 1.5526624755859375, 1.5517911376953124, 1.5515576171875, 1.5515587158203126, 1.551556640625, 1.5509442138671874, 1.5528365478515624, 1.5521904296875, 1.55300244140625, 1.5519825439453125, 1.5525980224609375, 1.551983642578125, 1.5516201171875, 1.5528509521484375, 3.202553955078125, 1.551009765625, 1.5523553466796876, 1.5506207275390624, 1.5519232177734374, 1.551784912109375, 1.551542236328125, 1.551203369140625, 1.5513343505859376, 1.55032373046875, 1.5522550048828125, 1.5514869384765626, 1.551784912109375, 1.5508438720703126, 1.5520716552734375, 1.5515863037109374, 1.55217919921875, 1.552343017578125, 1.5512073974609375, 1.5511490478515626, 1.5520225830078125, 1.551941650390625, 1.5513641357421875, 1.5520562744140625, 1.5527935791015626, 1.553005615234375, 1.5519703369140625, 1.5519405517578124, 1.55236962890625, 1.5524075927734375, 1.552400390625, 1.5521392822265625, 1.5526983642578125, 1.5521812744140624, 1.5518033447265625, 1.5511234130859375, 1.5530147705078126, 1.552041015625, 1.5515872802734374, 1.5513333740234374, 1.5513599853515625, 1.5520286865234374, 1.5513743896484375, 1.55210546875, 1.5510108642578124, 1.551393798828125, 1.55116845703125, 1.551873046875, 1.5514449462890625, 1.5518966064453126, 1.55174609375, 1.5518658447265625, 1.5518095703125, 1.5519150390625, 1.5517449951171876, 1.5523460693359374, 1.552300048828125, 1.552015380859375, 1.551193115234375, 1.5512535400390626, 1.5524495849609374, 1.55131494140625, 1.553142822265625, 3.20293701171875, 1.5507025146484374, 1.5517491455078125, 1.5505745849609376, 1.551394775390625, 1.5516416015625, 1.5522529296875, 1.5509012451171875, 1.551172607421875, 1.5507313232421875, 1.551066162109375, 1.5524403076171875, 1.552574462890625, 1.5519908447265625, 1.5519549560546875, 1.5521126708984374, 1.5522078857421875, 1.553606689453125, 1.552162841796875, 1.5526041259765626, 1.552289794921875, 1.5522764892578125, 1.55186376953125, 1.5506749267578126, 1.551815673828125, 1.5521044921875, 1.550876708984375, 1.5515504150390624, 1.550856201171875, 1.5514798583984375, 1.552759765625, 1.551753173828125, 1.5519580078125, 1.5519119873046876, 1.551141845703125, 1.551688720703125, 1.5519180908203125, 1.5516416015625, 1.5518433837890624, 1.5516488037109375, 1.55226416015625, 1.5521085205078125, 1.5519918212890624, 1.5524515380859376, 1.55057666015625, 1.5518023681640625, 1.5514920654296875, 1.5522559814453125, 1.551362060546875, 1.5525518798828124, 1.5522672119140626, 1.551783935546875, 1.551810546875, 1.5519447021484376, 1.5518751220703124, 1.5525919189453126, 1.5539844970703125, 1.5537889404296874, 1.5529072265625, 1.55158935546875, 1.55196826171875, 1.5523502197265624, 1.552722900390625]",tokens/s,0.6346973091049438,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-2.7b,facebook/opt-2.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,2119.368704,2816.999424,0.0,2170.55232,1927.744512,s,10,2.4524682617187503,0.245246826171875,0.001090999616645648,0.2451712341308594,0.24625067138671874,0.24669122314453126,0.24704366455078125,"[0.24615277099609376, 0.24713177490234375, 0.24369766235351562, 0.24442691040039063, 0.244733154296875, 0.243721435546875, 0.24478985595703126, 0.2455526123046875, 0.24614064025878907, 0.24612144470214845]",tokens/s,1043.8463322684916,kWh,2.8755634550064334e-06,1.5756288392757208e-06,1.2476319504855795e-05,1.6927511799137947e-05,tokens/kWh,15123309.499807118,MB,2119.368704,2816.999424,0.0,2170.55232,2031.983104,s,10,142.79050195312502,14.279050195312502,0.0031195370707275648,14.27847802734375,14.28149794921875,14.284002392578124,14.286005947265625,"[14.2787724609375, 14.2797041015625, 14.28094140625, 14.27818359375, 14.2772841796875, 14.2808662109375, 14.2747421875, 14.275935546875, 14.2775654296875, 14.2865068359375]",tokens/s,4.412058164812777,kWh,0.00016860610004691852,9.240989681349414e-05,0.0007205406280193562,0.000981556624879769,tokens/kWh,64183.76525930624,,s,629,144.7668543243408,0.23015398143774374,0.029166744118704197,0.22656614685058593,0.22702059936523436,0.22731673583984374,0.47147623168945313,"[0.228210693359375, 0.2264954833984375, 0.2263234558105469, 0.2265917510986328, 0.22652210998535155, 0.2266122283935547, 0.22653952026367188, 0.22631117248535157, 0.22642892456054686, 0.22647091674804687, 0.22641664123535157, 0.22648934936523438, 0.2270392303466797, 0.22645555114746094, 0.22661325073242186, 0.22698086547851562, 0.22665011596679688, 0.22659788513183593, 0.2262640686035156, 0.22631219482421874, 0.22640025329589844, 0.22692658996582032, 0.22699212646484376, 0.22696038818359374, 0.22650674438476562, 0.22637158203125, 0.2266552276611328, 0.22633882141113282, 0.22646173095703126, 0.2265814666748047, 0.22644224548339845, 0.22677912902832031, 0.22628044128417968, 0.22633573913574218, 0.2265681915283203, 0.226693115234375, 0.22634597778320312, 0.2265016326904297, 0.2263531494140625, 0.22655078125, 0.22668800354003907, 0.2267904052734375, 0.22690406799316407, 0.2266306610107422, 0.2264575958251953, 0.22657638549804687, 0.22662553405761718, 0.22665113830566405, 0.22651187133789064, 0.22652723693847657, 0.22627635192871093, 0.22631219482421874, 0.22654464721679687, 0.22648934936523438, 0.2266378173828125, 0.22687333679199218, 0.22802943420410157, 0.2267176971435547, 0.2265917510986328, 0.22732595825195312, 0.2264842224121094, 0.22659481811523438, 0.47421746826171873, 0.22658047485351562, 0.22655282592773437, 0.22653030395507812, 0.22670950317382813, 0.22652517700195313, 0.22650469970703124, 0.22647705078125, 0.22634803771972656, 0.2266787872314453, 0.22657125854492188, 0.22652621459960937, 0.22686515808105467, 0.2263900146484375, 0.22639718627929686, 0.22662553405761718, 0.22654975891113283, 0.22652928161621094, 0.2263838653564453, 0.22636647033691407, 0.2263582763671875, 0.22654975891113283, 0.22638899230957032, 0.2263961639404297, 0.22621388244628907, 0.22636749267578124, 0.22646170043945313, 0.22642381286621094, 0.22654156494140626, 0.2262988739013672, 0.22632652282714844, 0.2261411895751953, 0.2261012420654297, 0.22648838806152344, 0.22639097595214844, 0.2262917175292969, 0.22649856567382812, 0.2264145965576172, 0.22670130920410156, 0.2264453125, 0.227240966796875, 0.22686003112792968, 0.22684364318847655, 0.22670335388183593, 0.2264524841308594, 0.2268170166015625, 0.22824140930175782, 0.2270392303466797, 0.22767205810546876, 0.22691123962402343, 0.22711602783203125, 0.22689286804199219, 0.22666029357910156, 0.2267361297607422, 0.2266439666748047, 0.22683135986328126, 0.22709965515136718, 0.22691226196289063, 0.22731365966796874, 0.22672998046875, 0.22711911010742186, 0.22663475036621095, 0.22664601135253906, 0.47127346801757813, 0.2265681915283203, 0.226840576171875, 0.22668389892578125, 0.22649650573730468, 0.22704742431640626, 0.22672998046875, 0.22683135986328126, 0.22644326782226562, 0.2265743408203125, 0.226555908203125, 0.22665933227539062, 0.22649754333496094, 0.22686720275878905, 0.22685696411132814, 0.2270627899169922, 0.22701158142089845, 0.22654360961914063, 0.227631103515625, 0.22625177001953126, 0.2269306945800781, 0.226946044921875, 0.2268231658935547, 0.2272542724609375, 0.2273054656982422, 0.22650982666015626, 0.22637158203125, 0.22674432373046874, 0.22774681091308593, 0.22656512451171876, 0.22653132629394532, 0.22656101989746094, 0.22630706787109375, 0.22627430725097655, 0.22650367736816407, 0.2263654327392578, 0.2271293487548828, 0.2264596405029297, 0.2263173065185547, 0.2265753631591797, 0.22677912902832031, 0.22655282592773437, 0.22645350646972656, 0.2264954833984375, 0.22634597778320312, 0.22660403442382812, 0.226482177734375, 0.22736895751953126, 0.22658149719238282, 0.22649037170410155, 0.22647296142578124, 0.2265016326904297, 0.22649650573730468, 0.22651187133789064, 0.22654464721679687, 0.22644224548339845, 0.22651084899902343, 0.22658662414550781, 0.22652210998535155, 0.22644940185546875, 0.2268712921142578, 0.226334716796875, 0.22724812316894533, 0.4715550842285156, 0.22658047485351562, 0.22659788513183593, 0.22659075927734376, 0.2265425567626953, 0.22664601135253906, 0.22673306274414062, 0.22659481811523438, 0.2266234893798828, 0.22643096923828124, 0.2263951416015625, 0.22653439331054687, 0.22656716918945313, 0.22648626708984376, 0.2264842224121094, 0.2265323486328125, 0.22739251708984376, 0.22662757873535155, 0.22650778198242189, 0.22635621643066406, 0.22640025329589844, 0.22646783447265625, 0.2263951416015625, 0.2265374755859375, 0.22659071350097656, 0.22637158203125, 0.22640229797363282, 0.22639820861816407, 0.2262640686035156, 0.22646578979492188, 0.22665728759765624, 0.22639411926269531, 0.2265692138671875, 0.22644940185546875, 0.22637055969238282, 0.22653132629394532, 0.22651084899902343, 0.22662042236328125, 0.226440185546875, 0.22653952026367188, 0.2264248352050781, 0.22653132629394532, 0.22682112121582032, 0.22637464904785157, 0.22807859802246094, 0.22658047485351562, 0.22653543090820313, 0.22662144470214843, 0.2265364532470703, 0.22637158203125, 0.2266941375732422, 0.2265016326904297, 0.22764134216308593, 0.2268037109375, 0.22737408447265625, 0.2267689666748047, 0.22669818115234375, 0.22641868591308595, 0.2270064697265625, 0.226808837890625, 0.22760447692871094, 0.2266941375732422, 0.22663475036621095, 0.4716912536621094, 0.226555908203125, 0.22654669189453125, 0.22630400085449218, 0.22651904296875, 0.22646885681152343, 0.2263726043701172, 0.2264627227783203, 0.226376708984375, 0.22640956115722657, 0.22651280212402344, 0.2264944610595703, 0.22653030395507812, 0.22637055969238282, 0.22640538024902343, 0.22671359252929688, 0.22652825927734374, 0.22724607849121095, 0.2267904052734375, 0.22637464904785157, 0.2265006103515625, 0.2266173400878906, 0.2275594177246094, 0.22663168334960937, 0.22695526123046875, 0.22644326782226562, 0.2264453125, 0.22670541381835937, 0.22647193908691407, 0.2264842224121094, 0.22652928161621094, 0.22634495544433594, 0.22692658996582032, 0.2264575958251953, 0.22655078125, 0.22647705078125, 0.22659584045410155, 0.22641151428222656, 0.22638490295410157, 0.2264320068359375, 0.22686924743652342, 0.22668185424804688, 0.226555908203125, 0.22660096740722657, 0.2265364532470703, 0.22648013305664064, 0.22652006530761717, 0.22684774780273437, 0.22667674255371092, 0.2264524841308594, 0.2264145965576172, 0.22648320007324219, 0.2265518035888672, 0.22661631774902344, 0.22665216064453125, 0.22664909362792968, 0.22723379516601563, 0.22827008056640624, 0.22659890747070313, 0.22667776489257813, 0.22666752624511718, 0.22652517700195313, 0.22685696411132814, 0.47040716552734374, 0.2265364532470703, 0.22670335388183593, 0.22675152587890626, 0.22674327087402343, 0.22674124145507812, 0.22666854858398439, 0.2265016326904297, 0.22641664123535157, 0.226408447265625, 0.22644837951660157, 0.22657331848144532, 0.22669926452636718, 0.22673817443847658, 0.22666648864746095, 0.22652006530761717, 0.22712115478515624, 0.2267709503173828, 0.22682829284667969, 0.2264514617919922, 0.22717543029785156, 0.22672691345214843, 0.22699417114257814, 0.22659379577636718, 0.22670541381835937, 0.22654464721679687, 0.22634701538085938, 0.22648320007324219, 0.226376708984375, 0.22674330139160156, 0.2265518035888672, 0.22656716918945313, 0.22633882141113282, 0.22662144470214843, 0.22681292724609375, 0.22643609619140626, 0.22689791870117187, 0.22637055969238282, 0.22802841186523437, 0.22646885681152343, 0.22660301208496095, 0.22652006530761717, 0.22687744140625, 0.226482177734375, 0.22646476745605468, 0.22653132629394532, 0.2266787872314453, 0.22718771362304688, 0.22725325012207032, 0.22660914611816407, 0.22652517700195313, 0.2265364532470703, 0.2266623992919922, 0.2268590087890625, 0.2265999298095703, 0.2266439666748047, 0.22733311462402345, 0.2267484130859375, 0.22656101989746094, 0.22658047485351562, 0.22690713500976561, 0.22647296142578124, 0.2265333709716797, 0.4723138427734375, 0.22668698120117187, 0.22651699829101563, 0.22642994689941406, 0.22652006530761717, 0.22658355712890624, 0.22648832702636718, 0.22646067810058593, 0.22636134338378905, 0.2262917175292969, 0.22641766357421875, 0.22637362670898437, 0.226440185546875, 0.22636851501464844, 0.2264944610595703, 0.2264596405029297, 0.22644122314453125, 0.22647398376464845, 0.22640127563476561, 0.22637977600097656, 0.22642892456054686, 0.22646067810058593, 0.22642994689941406, 0.22729216003417968, 0.2265856018066406, 0.22656614685058593, 0.22651187133789064, 0.22686924743652342, 0.22699212646484376, 0.22667776489257813, 0.22668083190917968, 0.22647091674804687, 0.226440185546875, 0.2262794189453125, 0.22660198974609375, 0.22650778198242189, 0.227240966796875, 0.22670130920410156, 0.2265927734375, 0.22639820861816407, 0.22648524475097656, 0.22650265502929687, 0.22666444396972657, 0.2264637451171875, 0.22673306274414062, 0.2266306610107422, 0.2264524841308594, 0.22670541381835937, 0.22680677795410156, 0.22637773132324218, 0.22640333557128905, 0.22649754333496094, 0.22653030395507812, 0.22647602844238282, 0.2265180206298828, 0.2267105255126953, 0.2264811553955078, 0.2263408660888672, 0.22684159851074218, 0.226840576171875, 0.227162109375, 0.22650572204589844, 0.22680064392089844, 0.47229953002929687, 0.22655386352539061, 0.22669209289550782, 0.22669107055664062, 0.22653849792480468, 0.226914306640625, 0.22678323364257813, 0.22689485168457033, 0.2268078155517578, 0.22660812377929687, 0.22647296142578124, 0.22665318298339843, 0.22659584045410155, 0.2267166748046875, 0.22655897521972657, 0.22654566955566408, 0.2271068115234375, 0.22682009887695312, 0.22680575561523436, 0.22656204223632812, 0.2264217529296875, 0.22652108764648438, 0.22662553405761718, 0.22673408508300782, 0.22695730590820312, 0.22662757873535155, 0.2265364532470703, 0.22667570495605469, 0.2265743408203125, 0.22657125854492188, 0.22663679504394532, 0.2263592987060547, 0.22638490295410157, 0.22639411926269531, 0.2264514617919922, 0.22640640258789063, 0.22657125854492188, 0.22636134338378905, 0.22635008239746093, 0.22646681213378905, 0.2263756866455078, 0.22643507385253905, 0.226661376953125, 0.22654054260253906, 0.2264842224121094, 0.2266480712890625, 0.22652621459960937, 0.22654054260253906, 0.22649958801269532, 0.22644326782226562, 0.22654156494140626, 0.2265886688232422, 0.22666546630859374, 0.22688870239257813, 0.2266972198486328, 0.22649856567382812, 0.22649958801269532, 0.22640847778320314, 0.22657020568847655, 0.22655078125, 0.2271825866699219, 0.22659686279296876, 0.22644224548339845, 0.4732682189941406, 0.2267166748046875, 0.22669004821777344, 0.2266112060546875, 0.22709043884277344, 0.2265323486328125, 0.2266306610107422, 0.2265927734375, 0.22674636840820311, 0.22685285949707032, 0.22652723693847657, 0.2264954833984375, 0.22642381286621094, 0.22641253662109376, 0.22645452880859376, 0.2264248352050781, 0.22657125854492188, 0.22701670837402343, 0.22656716918945313, 0.2264268798828125, 0.22636441040039063, 0.22659379577636718, 0.22651596069335939, 0.22646476745605468, 0.22653543090820313, 0.22654873657226562, 0.22635110473632813, 0.2264453125, 0.22657740783691407, 0.22701568603515626, 0.22657331848144532, 0.226808837890625, 0.22729318237304688, 0.226840576171875, 0.22669004821777344, 0.22642994689941406, 0.22665216064453125, 0.226408447265625, 0.2264637451171875, 0.22659071350097656, 0.22760140991210936, 0.2266787872314453, 0.22653952026367188, 0.2264842224121094, 0.22630911254882813, 0.22652825927734374, 0.22639820861816407, 0.22657023620605468, 0.22674330139160156, 0.2271262664794922, 0.22665728759765624, 0.2265333709716797, 0.22666444396972657, 0.22674227905273436, 0.22675149536132813, 0.2265364532470703, 0.22639820861816407, 0.22652210998535155, 0.2266112060546875, 0.22665113830566405, 0.22656614685058593, 0.2264627227783203, 0.22642073059082032, 0.4729661560058594, 0.22653132629394532, 0.22665420532226563, 0.22658566284179688, 0.22666950988769533, 0.22666035461425782, 0.2266480712890625, 0.2267525177001953, 0.22648934936523438, 0.2266399688720703, 0.2264431610107422, 0.22651187133789064, 0.22709657287597657, 0.22754713439941407, 0.22637055969238282, 0.22647091674804687, 0.22653439331054687, 0.22654975891113283, 0.2265999298095703, 0.22654360961914063, 0.22677197265625, 0.22703718566894532, 0.2268784637451172, 0.22653543090820313, 0.22652210998535155, 0.22646681213378905, 0.2276433868408203, 0.2268784637451172, 0.2265886688232422, 0.22731878662109375, 0.2269306945800781, 0.22686822509765625, 0.22654464721679687, 0.22657945251464845, 0.2267484130859375, 0.22667263793945314, 0.22669209289550782, 0.22645350646972656, 0.22654566955566408, 0.22655282592773437, 0.22689996337890625, 0.2264698944091797, 0.22683544921875, 0.22653952026367188, 0.22666035461425782, 0.22665420532226563, 0.22662655639648438, 0.2269420166015625, 0.2271200714111328, 0.2266972198486328, 0.22657331848144532, 0.22654368591308593, 0.22663877868652343, 0.22703616333007812, 0.22663372802734374, 0.22663270568847657, 0.22767308044433593, 0.22783692932128907, 0.22670130920410156, 0.2265927734375, 0.2266378173828125, 0.22650469970703124, 0.22659584045410155]",tokens/s,4.344917232163973,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-125m,facebook/opt-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,1231.42144,879.230976,0.0,232.783872,169.719808,s,10,0.35891270065307623,0.035891270065307616,0.001166922312453486,0.03606051254272461,0.03654228057861328,0.037586180877685546,0.03842130111694336,"[0.03863008117675781, 0.03461849594116211, 0.03453500747680664, 0.034383518218994144, 0.03608425521850586, 0.03613267135620117, 0.036015392303466794, 0.036310302734375, 0.036166206359863284, 0.03603676986694336]",tokens/s,7132.653693619183,kWh,4.2057203039815654e-07,2.3039066429929688e-07,8.883686426833978e-07,1.5393313373808513e-06,tokens/kWh,166305975.7073354,MB,1231.42144,879.230976,0.0,232.783872,199.793152,s,10,21.796639404296876,2.179663940429687,0.03379298018028798,2.1997751464843747,2.2025270751953125,2.2029048950195316,2.2032071508789066,"[2.15886328125, 2.11872216796875, 2.1151240234375, 2.1993291015625, 2.20328271484375, 2.20231494140625, 2.20022119140625, 2.201608642578125, 2.194730224609375, 2.202443115234375]",tokens/s,28.903538215886858,kWh,2.6054930089725813e-05,1.4278879354016576e-05,5.1935423616310185e-05,9.22692330600526e-05,tokens/kWh,682784.4765870875,,s,629,22.081034248352047,0.035104982906760014,0.004313579122612795,0.03483135986328125,0.03513838195800781,0.035482418823242184,0.06834089904785158,"[0.035579902648925785, 0.0354150390625, 0.035542015075683595, 0.0358737907409668, 0.03517337417602539, 0.035922943115234376, 0.03607551956176758, 0.03511808013916016, 0.03547443389892578, 0.035765247344970705, 0.035937278747558594, 0.03575091171264649, 0.03546214294433594, 0.04169113540649414, 0.03495731353759766, 0.03391795349121094, 0.033565696716308595, 0.03336908721923828, 0.03351347351074219, 0.03344179153442383, 0.03361177444458008, 0.03342745590209961, 0.03343564987182617, 0.03342745590209961, 0.03340185546875, 0.03341926574707031, 0.03338649749755859, 0.03336294555664063, 0.033552383422851564, 0.03352883148193359, 0.03450368118286133, 0.03436032104492188, 0.034111488342285154, 0.03450265502929688, 0.03355136108398438, 0.034164737701416016, 0.03474431991577148, 0.034351104736328124, 0.03521331024169922, 0.035111934661865234, 0.03381760025024414, 0.034427902221679685, 0.033756160736083986, 0.0334510383605957, 0.0331734733581543, 0.03338854217529297, 0.03346944046020508, 0.03350425720214844, 0.03346636962890625, 0.03356671905517578, 0.03343769454956055, 0.03313971328735352, 0.03326259231567383, 0.03314790344238281, 0.03320729446411133, 0.03347967910766601, 0.03352678298950195, 0.033538047790527346, 0.03310182571411133, 0.03322163009643555, 0.033478656768798826, 0.03340800094604492, 0.06847795104980468, 0.03334348678588867, 0.03356576156616211, 0.03346118545532226, 0.033667072296142575, 0.03376025772094727, 0.033555454254150394, 0.03353497695922852, 0.03349913787841797, 0.03347455978393555, 0.03362713623046875, 0.033650688171386715, 0.03344998550415039, 0.03351244735717773, 0.033532928466796875, 0.033718273162841796, 0.033614849090576174, 0.03364659118652344, 0.033478656768798826, 0.03355136108398438, 0.03344998550415039, 0.034282497406005856, 0.03367833709716797, 0.033516544342041016, 0.03338854217529297, 0.033584129333496096, 0.033539070129394534, 0.033326080322265625, 0.03335987091064453, 0.03366604614257813, 0.03334143829345703, 0.033258495330810545, 0.03334041595458984, 0.03343155288696289, 0.03324620819091797, 0.03349401473999023, 0.03338547134399414, 0.03338137435913086, 0.0335175666809082, 0.03442892837524414, 0.03411558532714844, 0.034249729156494144, 0.03337011337280273, 0.03331174468994141, 0.033413120269775394, 0.033301502227783206, 0.03349606323242187, 0.03334143829345703, 0.03332403182983398, 0.033527809143066405, 0.0335206413269043, 0.03582156753540039, 0.035253246307373046, 0.03405209732055664, 0.03366604614257813, 0.033827838897705076, 0.03394559860229492, 0.03346636962890625, 0.03348582458496094, 0.03372544097900391, 0.03336703872680664, 0.03363942337036133, 0.033410049438476565, 0.06798847961425782, 0.033957889556884766, 0.03349401473999023, 0.03372236633300781, 0.033659934997558594, 0.033262561798095704, 0.033339393615722655, 0.03347148895263672, 0.03481190490722656, 0.03399270248413086, 0.033448993682861326, 0.033379295349121096, 0.03344179153442383, 0.03343155288696289, 0.0330618896484375, 0.03346124649047852, 0.033410049438476565, 0.033430526733398434, 0.033495040893554685, 0.03334656143188477, 0.03343564987182617, 0.033446910858154294, 0.033513534545898435, 0.03330553436279297, 0.03338547134399414, 0.033484798431396484, 0.03334147262573242, 0.03339465713500977, 0.03335782241821289, 0.03341516876220703, 0.033355777740478515, 0.03362713623046875, 0.03338854217529297, 0.033360897064208986, 0.03363020706176758, 0.03340800094604492, 0.03336505508422852, 0.03361990356445312, 0.03336191940307617, 0.035092479705810545, 0.034958335876464845, 0.03446783828735352, 0.033719329833984374, 0.0333260498046875, 0.03360768127441406, 0.03371417617797851, 0.03365785598754883, 0.033653759002685545, 0.03359027099609375, 0.03361177444458008, 0.033465343475341795, 0.033562625885009766, 0.03357183837890625, 0.03344076919555664, 0.0335093765258789, 0.03359231948852539, 0.033463294982910154, 0.03351551818847656, 0.03339059066772461, 0.0335022087097168, 0.03354009628295898, 0.033432575225830076, 0.033476608276367184, 0.06785228729248047, 0.038763519287109374, 0.03499212646484375, 0.03476172637939453, 0.03466035079956055, 0.03442892837524414, 0.03445862579345703, 0.0349194221496582, 0.03481087875366211, 0.03482009506225586, 0.03474227142333984, 0.03470336151123047, 0.03492768096923828, 0.0348221435546875, 0.034755519866943356, 0.03479654312133789, 0.035053569793701174, 0.03477503967285156, 0.03513651275634765, 0.03483647918701172, 0.03467161560058594, 0.03479244613647461, 0.034738174438476564, 0.03485184097290039, 0.03481292724609375, 0.03513651275634765, 0.03499728012084961, 0.034848735809326174, 0.03482009506225586, 0.03604684829711914, 0.03580108642578125, 0.034900993347167966, 0.03499622344970703, 0.03482009506225586, 0.03486617660522461, 0.034716670989990234, 0.034871295928955076, 0.03480883026123047, 0.03496857452392578, 0.034961406707763674, 0.03484262466430664, 0.03478732681274414, 0.0348671989440918, 0.03487539291381836, 0.03496038436889649, 0.03478121566772461, 0.03482006454467773, 0.03547750473022461, 0.03485184097290039, 0.034953216552734374, 0.034854911804199216, 0.03493478393554687, 0.03474943923950195, 0.03476889419555664, 0.03471974563598633, 0.034825214385986326, 0.035095550537109374, 0.0350300178527832, 0.034802688598632815, 0.034781185150146485, 0.034677761077880856, 0.034490367889404294, 0.034372608184814454, 0.07120281219482422, 0.03474227142333984, 0.03492659378051758, 0.03486515045166016, 0.035027999877929684, 0.035214336395263675, 0.035108863830566404, 0.03492655944824219, 0.03491635131835937, 0.03499212646484375, 0.0349409294128418, 0.03500236892700195, 0.03487641525268555, 0.03486105728149414, 0.03492659378051758, 0.03494911956787109, 0.03502489471435547, 0.034909183502197266, 0.03505152130126953, 0.03499724960327148, 0.03479142379760742, 0.03497983932495117, 0.03498291015625, 0.03495116806030273, 0.03492454528808594, 0.03508633422851563, 0.034934814453125, 0.03494499206542969, 0.034936832427978515, 0.034841598510742186, 0.03491839981079101, 0.035023872375488284, 0.03488665771484375, 0.03486617660522461, 0.03508428955078125, 0.03497881698608398, 0.03496448135375976, 0.03487334442138672, 0.034991104125976565, 0.035092479705810545, 0.03487641525268555, 0.0350013427734375, 0.035125247955322264, 0.03501567840576172, 0.035350528717041016, 0.03521843338012695, 0.03506687927246094, 0.03503104019165039, 0.034985984802246094, 0.03507308959960938, 0.034981822967529295, 0.03523174285888672, 0.0352174072265625, 0.034923519134521484, 0.034361343383789066, 0.03486822509765625, 0.03495731353759766, 0.03501875305175781, 0.034956287384033204, 0.03489484786987305, 0.034835456848144535, 0.035156993865966796, 0.03473408126831055, 0.07153561401367188, 0.034855934143066404, 0.03496345520019531, 0.034987071990966796, 0.035133377075195316, 0.03502592086791992, 0.035043327331542966, 0.035209217071533204, 0.03504844665527344, 0.03502592086791992, 0.034936832427978515, 0.03493478393554687, 0.0358205451965332, 0.035659774780273434, 0.03496038436889649, 0.03518668746948242, 0.03511296081542969, 0.03490304183959961, 0.03482624053955078, 0.03487539291381836, 0.03492761611938477, 0.03479142379760742, 0.035125247955322264, 0.03506585693359375, 0.03505254364013672, 0.034825214385986326, 0.03476995086669922, 0.03494089508056641, 0.03478732681274414, 0.03488256072998047, 0.035244033813476565, 0.035007488250732424, 0.03475046539306641, 0.03497062301635742, 0.0347770881652832, 0.03480166244506836, 0.03479964828491211, 0.03495734405517578, 0.03496441650390625, 0.03494911956787109, 0.034885631561279294, 0.034953216552734374, 0.034909183502197266, 0.03480575942993164, 0.03474943923950195, 0.034948097229003904, 0.03509862518310547, 0.03496857452392578, 0.03502284622192383, 0.0344719352722168, 0.03450271987915039, 0.03489888000488281, 0.034909183502197266, 0.03487846374511719, 0.03488460922241211, 0.03492147064208984, 0.03486515045166016, 0.03475251388549805, 0.03486310577392578, 0.03487027359008789, 0.03497369766235352, 0.03508230209350586, 0.034826175689697265, 0.07138098907470704, 0.035064830780029296, 0.035089408874511716, 0.03489894485473633, 0.03474537658691406, 0.034952159881591796, 0.034969600677490234, 0.03484985733032227, 0.0348732795715332, 0.03523481750488281, 0.034993152618408206, 0.03489593505859375, 0.03493983840942383, 0.03493580627441406, 0.034854911804199216, 0.03482316970825195, 0.03493273544311523, 0.034797569274902344, 0.03540787124633789, 0.035068992614746095, 0.03575494384765625, 0.03548672103881836, 0.03483955383300781, 0.03533107376098633, 0.03497062301635742, 0.034800640106201174, 0.034786304473876956, 0.03497574234008789, 0.0349306869506836, 0.03490508651733398, 0.034917377471923826, 0.03463065719604492, 0.034928638458251955, 0.03481292724609375, 0.03505254364013672, 0.03483955383300781, 0.03466854476928711, 0.03482112121582031, 0.03525734329223633, 0.03482726287841797, 0.034948097229003904, 0.034776065826416014, 0.03475558471679688, 0.034283519744873044, 0.03471974563598633, 0.034735103607177735, 0.03482931137084961, 0.03508224105834961, 0.034678783416748044, 0.03488870239257812, 0.034802688598632815, 0.03478015899658203, 0.03476070404052734, 0.03488051223754883, 0.034976768493652347, 0.0347586555480957, 0.03504435348510742, 0.034929695129394533, 0.03516617584228516, 0.034799617767333986, 0.034723838806152346, 0.03487334442138672, 0.03482931137084961, 0.07187558746337891, 0.03538739013671875, 0.03526348876953125, 0.03513241577148438, 0.034925567626953126, 0.03487744140625, 0.0354856948852539, 0.035422206878662106, 0.0351723518371582, 0.034976768493652347, 0.03488460922241211, 0.034890750885009765, 0.0349409294128418, 0.03503206253051758, 0.035076095581054685, 0.03489279937744141, 0.035119102478027346, 0.034729984283447264, 0.03479244613647461, 0.034864158630371095, 0.034802654266357425, 0.0348221435546875, 0.034678783416748044, 0.03501055908203125, 0.03484467315673828, 0.0348487663269043, 0.03479859161376953, 0.03482931137084961, 0.03487539291381836, 0.034783233642578126, 0.03495116806030273, 0.03474534225463867, 0.03493580627441406, 0.03485388946533203, 0.03487948989868164, 0.034926624298095704, 0.034881504058837894, 0.03505254364013672, 0.03485081481933594, 0.03487846374511719, 0.03444224166870117, 0.03474431991577148, 0.03492761611938477, 0.034948097229003904, 0.03510067367553711, 0.0347770881652832, 0.03488972854614258, 0.03482112121582031, 0.034977790832519534, 0.034405376434326174, 0.03482422256469726, 0.034968544006347656, 0.03485081481933594, 0.0349409294128418, 0.034840576171875, 0.034907135009765625, 0.035122177124023435, 0.03496755218505859, 0.03485388946533203, 0.035350528717041016, 0.03549593734741211, 0.03502182388305664, 0.03480473709106445, 0.07150796508789062, 0.0350013427734375, 0.03506995010375977, 0.034994174957275394, 0.03483647918701172, 0.03483135986328125, 0.034991104125976565, 0.03488153457641602, 0.03486310577392578, 0.034756607055664065, 0.034781185150146485, 0.03496038436889649, 0.034710529327392575, 0.03470848083496094, 0.034977790832519534, 0.03480780792236328, 0.03479040145874023, 0.034776065826416014, 0.03490304183959961, 0.03465219116210937, 0.0350074577331543, 0.03491123199462891, 0.03498905563354492, 0.03513139343261719, 0.03480883026123047, 0.03476582336425781, 0.03472281646728516, 0.034713600158691404, 0.034830337524414064, 0.03484364700317383, 0.03500646209716797, 0.034683902740478514, 0.03486310577392578, 0.03489894485473633, 0.034664447784423826, 0.03480575942993164, 0.034835456848144535, 0.03486822509765625, 0.0347770881652832, 0.03489382553100586, 0.03503615951538086, 0.03473715209960938, 0.03480473709106445, 0.034802688598632815, 0.034677761077880856, 0.03475251388549805, 0.03466652679443359, 0.03473097610473633, 0.0347883529663086, 0.035141632080078124, 0.03476070404052734, 0.03492966461181641, 0.03492147064208984, 0.034854911804199216, 0.034776065826416014, 0.03474431991577148, 0.0348037109375, 0.034800640106201174, 0.0348037109375, 0.03454054260253906, 0.03448934555053711, 0.03489791870117188, 0.03465830230712891, 0.07124582672119141, 0.03520102310180664, 0.03487744140625, 0.035156993865966796, 0.03470848083496094, 0.03489996719360351, 0.03482726287841797, 0.03494297790527344, 0.03487744140625, 0.034769920349121096, 0.034885631561279294, 0.03489996719360351, 0.03513756942749023, 0.03485385513305664, 0.03532185745239258, 0.03501875305175781, 0.0368721923828125, 0.03486310577392578, 0.03483443069458008, 0.03476582336425781, 0.03483647918701172, 0.0348221435546875, 0.03477196884155274, 0.034669567108154296, 0.03483647918701172, 0.034947071075439456, 0.0348671989440918, 0.03482316970825195, 0.03477814483642578, 0.03476988983154297, 0.03473612976074219, 0.03470745468139649, 0.03570380783081055, 0.03584921646118164, 0.03489484786987305, 0.03489689636230469, 0.03458969497680664, 0.034830337524414064, 0.03489689636230469, 0.03477811050415039, 0.034705406188964845, 0.03480985641479492, 0.03458662414550781, 0.03588614273071289, 0.03546003341674805, 0.035043327331542966, 0.034993152618408206, 0.03488051223754883, 0.03490611267089844, 0.035004417419433595, 0.03475763320922851, 0.034797569274902344, 0.03473715209960938, 0.03501772689819336, 0.034772991180419925, 0.034885631561279294, 0.03487539291381836, 0.03483443069458008, 0.03482316970825195, 0.0350013427734375, 0.03475763320922851, 0.03494604873657227, 0.0350300178527832]",tokens/s,28.485984529775525,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/gemma-7b,google/gemma-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/gemma-7b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aa9e8-1f032d6052fa589103a25bc9;d692eb93-6125-4421-8698-bce84340ff7d) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/gemma-7b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/gemma-7b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,huggyllama/llama-13b,huggyllama/llama-13b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,3015.59808,9299.296256,0.0,8652.849152,8210.185216,s,10,11.00354541015625,1.100354541015625,0.0021995629105345756,1.10037060546875,1.1026337646484374,1.103404345703125,1.104020810546875,"[1.1041749267578125, 1.1024625244140625, 1.0971700439453125, 1.0986126708984374, 1.0993076171875, 1.097432861328125, 1.099380126953125, 1.101361083984375, 1.1021866455078124, 1.1014569091796875]",tokens/s,232.65228656548507,kWh,1.2965490321318309e-05,7.10465502763327e-06,5.986921456200278e-05,7.993935991095436e-05,tokens/kWh,3202427.443566751,MB,3016.822784,9299.296256,0.0,8652.849152,8503.104,s,10,641.11232421875,64.11123242187502,0.005190298699693196,64.11128124999999,64.116784765625,64.1185056640625,64.11988238281249,"[64.1162265625, 64.11091796875, 64.10715234375, 64.11164453125, 64.1140703125, 64.10447265625, 64.1202265625, 64.106796875, 64.1044140625, 64.11640234375]",tokens/s,0.9826671180712498,kWh,0.0007569437074826825,0.00041487128047605436,0.0034860815944185997,0.004657896582377336,tokens/kWh,13525.418369818235,,s,629,649.9943001708987,1.0333772657724936,0.13070199113792613,1.0176091918945311,1.0182221801757814,1.0184349731445312,2.1165633984375,"[1.0176849975585938, 1.0179225463867188, 1.0177464599609376, 1.0181263427734375, 1.01766552734375, 1.0177433471679687, 1.0174985961914063, 1.0180587768554688, 1.0172999877929687, 1.0173552856445311, 1.017459716796875, 1.0184939575195313, 1.017143310546875, 1.0182799072265625, 1.0176931762695312, 1.0179122924804687, 1.0178191528320313, 1.0179092407226562, 1.0176112670898438, 1.0177853393554688, 1.0181171264648436, 1.0180086059570312, 1.0171443481445313, 1.0179840087890626, 1.0173839111328125, 1.0177720336914062, 1.0175538940429687, 1.0174218139648437, 1.0173317260742187, 1.0178887939453125, 1.0179092407226562, 1.0181570434570313, 1.0174976196289063, 1.0177576904296874, 1.017449462890625, 1.017692138671875, 1.0180843505859376, 1.0179747924804687, 1.0176112670898438, 1.0182471923828125, 1.0175887451171874, 1.017427978515625, 1.0171924438476563, 1.0172886962890626, 1.0171688842773436, 1.0173245239257813, 1.0170091552734375, 1.0176368408203125, 1.0171146240234374, 1.0184765625, 1.0177085571289062, 1.017554931640625, 1.0172211303710939, 1.0182522583007811, 1.017290771484375, 1.0179625244140624, 1.0176573486328124, 1.0178734130859375, 1.0176276245117188, 1.0185001220703125, 1.0183004150390624, 1.0180914916992188, 2.120390625, 1.0169886474609375, 1.0174678955078125, 1.0169763793945312, 1.017365478515625, 1.0172262573242188, 1.0176307373046876, 1.017291748046875, 1.0174207763671874, 1.017438232421875, 1.0176419677734374, 1.01682275390625, 1.0177566528320312, 1.0169026489257813, 1.0171187133789064, 1.0169517822265626, 1.0179747924804687, 1.0175006713867187, 1.0174310302734375, 1.0172119140625, 1.018013671875, 1.0172262573242188, 1.0173501586914062, 1.0170921020507813, 1.0172692260742187, 1.0172661743164062, 1.01758154296875, 1.0180413208007812, 1.017764892578125, 1.0175068359375, 1.0178314208984376, 1.0169129028320312, 1.0174238891601564, 1.01768701171875, 1.0177402954101562, 1.0183987426757812, 1.01777099609375, 1.0177843017578125, 1.0177607421875, 1.0173163452148437, 1.0173081665039063, 1.017702392578125, 1.0177484741210938, 1.0175764770507814, 1.01732763671875, 1.018228759765625, 1.018945556640625, 1.0175508422851562, 1.0177587280273437, 1.0177536010742188, 1.0183505859375, 1.0178427124023437, 1.01798095703125, 1.0178457641601562, 1.0182246704101563, 1.0184437866210938, 1.0184017944335937, 1.0174044189453124, 1.01778125, 1.0175907592773437, 1.0179645385742186, 1.0172713012695314, 1.0177116088867189, 2.116588623046875, 1.0172682495117187, 1.017439208984375, 1.0171781005859375, 1.0176378784179687, 1.0172579956054688, 1.0172876586914064, 1.0171351318359374, 1.017439208984375, 1.0174095458984376, 1.0178488159179688, 1.017796630859375, 1.0178846435546876, 1.0169200439453125, 1.0174443359375, 1.0173378295898436, 1.0173532104492187, 1.0172938232421875, 1.0175784912109376, 1.0172467041015625, 1.0177372436523437, 1.01789697265625, 1.0177136840820313, 1.0170101928710937, 1.01732763671875, 1.0172815551757812, 1.0176215209960937, 1.0174166870117187, 1.0176153564453125, 1.017275390625, 1.0179368896484375, 1.0172160034179687, 1.0177505493164063, 1.0171678466796874, 1.01737060546875, 1.01718017578125, 1.0172098388671875, 1.0171494140625, 1.0173306884765625, 1.0171054077148438, 1.0177413330078124, 1.017786376953125, 1.0182041625976563, 1.0170777587890625, 1.0175405883789062, 1.0175836181640625, 1.0175068359375, 1.017169921875, 1.0173778076171875, 1.0175529174804687, 1.0178037719726563, 1.0177454223632814, 1.0187315063476563, 1.01823486328125, 1.0180054931640625, 1.0178140258789063, 1.0181683349609374, 1.017670654296875, 1.0181611328125, 1.0182564086914063, 1.0183587646484376, 1.0175396118164062, 1.0181222534179688, 2.11649853515625, 1.0170214233398438, 1.0172620849609375, 1.0174719848632812, 1.0172548828125, 1.0174484252929688, 1.017628662109375, 1.0173849487304687, 1.0173992919921875, 1.0177689819335938, 1.0176215209960937, 1.01728564453125, 1.0174402465820314, 1.0171627807617187, 1.0172057495117188, 1.0171904296875, 1.0177679443359375, 1.0173214721679686, 1.0170664672851562, 1.0171207885742188, 1.0171893920898438, 1.0169364624023438, 1.0172098388671875, 1.0170009765625, 1.0171842651367187, 1.01760205078125, 1.0179368896484375, 1.0174985961914063, 1.0172743530273438, 1.0170787963867187, 1.0172272338867188, 1.0170316772460937, 1.0173040771484374, 1.0175529174804687, 1.0179287109375, 1.0178242797851562, 1.0183321533203125, 1.01816015625, 1.0178682861328125, 1.0174054565429687, 1.0174771118164063, 1.0178682861328125, 1.0177188110351563, 1.017955322265625, 1.0186710815429687, 1.0180361938476563, 1.018166259765625, 1.017849853515625, 1.0181478271484374, 1.0177095947265624, 1.017891845703125, 1.0177259521484374, 1.0180966186523437, 1.0177996826171876, 1.0185390014648437, 1.0180700073242188, 1.0183905029296876, 1.0182389526367188, 1.0185400390625, 1.0179686279296876, 1.0179194946289063, 1.0175518798828125, 1.0181058349609375, 2.11787255859375, 1.017354248046875, 1.0174566650390624, 1.017512939453125, 1.0176153564453125, 1.0173112182617188, 1.0177638549804688, 1.0170715942382813, 1.017206787109375, 1.0171371459960938, 1.0176399536132812, 1.0181683349609374, 1.0179246215820312, 1.017238525390625, 1.0177177734375, 1.01699072265625, 1.01747509765625, 1.0175672607421875, 1.0181283569335937, 1.01778125, 1.01810791015625, 1.018156005859375, 1.018429443359375, 1.0175344848632812, 1.0176460571289063, 1.0172333984375, 1.0177321166992188, 1.01758154296875, 1.0172559204101563, 1.0175887451171874, 1.0181652221679687, 1.0181017456054688, 1.0174576416015626, 1.0177464599609376, 1.017849853515625, 1.0172507934570312, 1.0176378784179687, 1.0174627685546875, 1.0175211791992187, 1.017702392578125, 1.0177321166992188, 1.01766552734375, 1.0176245727539062, 1.017238525390625, 1.0175139770507813, 1.0173101806640625, 1.0175518798828125, 1.01707568359375, 1.0175723266601562, 1.0176614379882813, 1.018265625, 1.0177587280273437, 1.0177955932617186, 1.017280517578125, 1.0177669067382813, 1.01758056640625, 1.0177003784179688, 1.0202439575195312, 1.0180403442382813, 1.0180946044921875, 1.0176676025390625, 1.0172211303710939, 1.0179778442382812, 2.115203125, 1.0167890014648437, 1.0172354736328124, 1.0170439453125, 1.0171873168945313, 1.0175703125, 1.0176470947265626, 1.0175744018554687, 1.0177515258789063, 1.0173716430664062, 1.0174197998046874, 1.0172938232421875, 1.0174003295898437, 1.0167725830078125, 1.0175344848632812, 1.0169354248046876, 1.017470947265625, 1.017322509765625, 1.017417724609375, 1.0173368530273437, 1.0175313720703125, 1.0170009765625, 1.0170460205078125, 1.0174044189453124, 1.017523193359375, 1.0177791748046876, 1.0174197998046874, 1.0170818481445312, 1.0176419677734374, 1.0173101806640625, 1.0176409301757812, 1.0174095458984376, 1.0175293579101563, 1.0173173828125, 1.0182144165039062, 1.0182215576171876, 1.0176747436523437, 1.01743310546875, 1.0186895141601562, 1.0174781494140626, 1.0176266479492186, 1.0174013671875, 1.0175191040039062, 1.0173880615234374, 1.017828369140625, 1.0177105712890624, 1.0176378784179687, 1.0176849975585938, 1.017650146484375, 1.0173193969726562, 1.0175958862304688, 1.017544677734375, 1.01745458984375, 1.0172119140625, 1.017970703125, 1.0179348754882813, 1.017543701171875, 1.0179287109375, 1.0183065795898438, 1.017802734375, 1.0182492065429687, 1.0176123046875, 1.0174044189453124, 2.1172890625, 1.0174003295898437, 1.0177330932617188, 1.0177269897460937, 1.0172897338867188, 1.0172108764648438, 1.0175191040039062, 1.0172661743164062, 1.0171361083984376, 1.0175949096679688, 1.0182318115234374, 1.0174924926757813, 1.0179317626953126, 1.0172047119140626, 1.017650146484375, 1.017218017578125, 1.0178754272460937, 1.017364501953125, 1.017511962890625, 1.0177156982421875, 1.0180003662109376, 1.0175949096679688, 1.01768603515625, 1.0176327514648438, 1.0180106201171875, 1.0180044555664063, 1.0175570068359374, 1.0175354614257812, 1.0178191528320313, 1.0182748413085938, 1.0180044555664063, 1.0177034301757812, 1.0174884033203124, 1.0172713012695314, 1.0173092041015626, 1.01715966796875, 1.0175641479492188, 1.0176747436523437, 1.0178734130859375, 1.0186076049804687, 1.0178191528320313, 1.017248779296875, 1.0178734130859375, 1.0178088989257812, 1.0177566528320312, 1.0173737182617189, 1.0184867553710937, 1.0182072143554688, 1.018197998046875, 1.0183782348632813, 1.0181672973632812, 1.0175877075195312, 1.0179911499023437, 1.0176522216796875, 1.0181212158203126, 1.017575439453125, 1.0183485717773437, 1.01770751953125, 1.0183301391601562, 1.0184693603515624, 1.0186875, 1.0184386596679686, 1.0182625122070312, 2.120072265625, 1.017439208984375, 1.0174781494140626, 1.0172333984375, 1.0177269897460937, 1.0175928344726564, 1.0170654907226562, 1.0174505004882812, 1.0172559204101563, 1.0168995971679688, 1.0167019653320313, 1.0168186645507813, 1.0172682495117187, 1.0168565673828125, 1.0178058471679687, 1.0173900756835939, 1.0174361572265624, 1.0177945556640624, 1.0173880615234374, 1.0177044677734375, 1.0172129516601562, 1.0170511474609376, 1.0175641479492188, 1.0171637573242187, 1.0182041625976563, 1.0174515380859375, 1.0173992919921875, 1.017302001953125, 1.01718017578125, 1.0170521850585938, 1.017697265625, 1.0175518798828125, 1.0177699584960938, 1.0171688842773436, 1.0180577392578125, 1.0177750854492187, 1.0180413208007812, 1.0171555786132813, 1.0174617309570313, 1.0173552856445311, 1.0172088623046875, 1.0174453735351563, 1.0177587280273437, 1.0179358520507813, 1.017807861328125, 1.0180464477539062, 1.01760205078125, 1.01793994140625, 1.01802392578125, 1.0177402954101562, 1.01783349609375, 1.01783447265625, 1.0175006713867187, 1.017565185546875, 1.0181806030273437, 1.01722314453125, 1.0179891357421875, 1.0179143676757811, 1.0177699584960938, 1.018265625, 1.018239990234375, 1.0177269897460937, 1.01764404296875, 2.12052685546875, 1.0178191528320313, 1.0170664672851562, 1.0168914184570312, 1.016933349609375, 1.0168248291015625, 1.0170194091796876, 1.0169968872070312, 1.0170706176757813, 1.01739111328125, 1.0174085083007813, 1.0172446899414063, 1.01743408203125, 1.0172498168945312, 1.0173173828125, 1.0169630737304687, 1.0172764282226563, 1.017670654296875, 1.0172395629882813, 1.0175570068359374, 1.01774951171875, 1.0171514892578124, 1.0176829223632813, 1.0178232421875, 1.0177474365234376, 1.0177638549804688, 1.0182195434570311, 1.0177802124023438, 1.01798193359375, 1.0181058349609375, 1.0178447265625, 1.0170623779296875, 1.017565185546875, 1.01713818359375, 1.017654296875, 1.0177515258789063, 1.0174115600585938, 1.0176266479492186, 1.01768603515625, 1.0180618286132812, 1.0176409301757812, 1.01743408203125, 1.0178948974609374, 1.0173562622070313, 1.0175979614257813, 1.0175078125, 1.0172713012695314, 1.0172344360351562, 1.0179164428710938, 1.0178744506835937, 1.0179573974609375, 1.01726513671875, 1.017871337890625, 1.0172272338867188, 1.017660400390625, 1.01722314453125, 1.0177474365234376, 1.0172640991210937, 1.0182471923828125, 1.0183259887695313, 1.0176481323242188, 1.0172948608398438, 1.0174535522460937, 2.119248779296875, 1.0171259155273438, 1.0173378295898436, 1.0171361083984376, 1.0176768188476562, 1.0173480834960937, 1.0173060913085938, 1.0175313720703125, 1.0172876586914064, 1.0169313354492187, 1.0169517822265626, 1.0170767211914062, 1.016911865234375, 1.0168883056640625, 1.0176091918945311, 1.0169405517578125, 1.0172006225585937, 1.0173839111328125, 1.0174832763671875, 1.0169149169921874, 1.0174115600585938, 1.017101318359375, 1.0171627807617187, 1.0175191040039062, 1.0177177734375, 1.0175150146484375, 1.0177362060546875, 1.0174299926757813, 1.0178037719726563, 1.0174238891601564, 1.0178887939453125, 1.0173532104492187, 1.0172507934570312, 1.01740234375, 1.0190120849609374, 1.0181201782226563, 1.018503173828125, 1.0183167724609374, 1.0181693725585939, 1.0181580810546875, 1.0182164306640624, 1.0177146606445313, 1.0178099365234374, 1.0184202270507812, 1.0194851684570312, 1.019615234375, 1.0187745361328124, 1.018102783203125, 1.0178682861328125, 1.01852978515625, 1.018186767578125, 1.017744384765625, 1.0179573974609375, 1.0179799194335937, 1.0177484741210938, 1.01732861328125, 1.0178109741210937, 1.0174668579101562, 1.017786376953125, 1.017491455078125, 1.0179041137695313, 1.0180321044921874, 1.0186588134765624]",tokens/s,0.9677007934294523,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-7B,Qwen/Qwen-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 550, in from_pretrained + model_class = get_class_from_dynamic_module( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 498, in get_class_from_dynamic_module + final_module = get_cached_module_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 323, in get_cached_module_file + modules_needed = check_imports(resolved_module_file) + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 181, in check_imports + raise ImportError( +ImportError: This modeling file requires the following packages that were not found in your environment: transformers_stream_generator. Run `pip install transformers_stream_generator` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-72B,Qwen/Qwen1.5-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 96.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciLM-7B,,cuda,0,42,,,,,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,,MB,1674.596352,5516.034048,0.0,4869.586944,4743.593472,s,10,6.137113647460938,0.6137113647460938,0.0032895717697753717,0.6129547424316406,0.6153997436523437,0.6190885131835938,0.6220395288085938,"[0.6227772827148438, 0.6144583129882812, 0.61261669921875, 0.61060546875, 0.612577392578125, 0.6105811157226563, 0.6121093139648438, 0.6145800170898438, 0.6132927856445313, 0.6135152587890625]",tokens/s,417.13420136176387,kWh,7.2172448039054885e-06,3.954746732982612e-06,3.355963469082604e-05,4.473162622771414e-05,tokens/kWh,5723020.189268045,MB,1674.596352,5516.034048,0.0,4869.586944,4769.651712,s,10,361.12618359375006,36.112618359375006,0.009999043540991457,36.112980468749996,36.125964453125,36.127720507812505,36.1291253515625,"[36.1294765625, 36.0964296875, 36.12557421875, 36.1174140625, 36.11404296875, 36.1114296875, 36.11191796875, 36.09769140625, 36.10714453125, 36.1150625]",tokens/s,1.7445425688343892,kWh,0.00042614284929302004,0.0002335616249413365,0.0019425026683799653,0.0026022071426143214,tokens/kWh,24210.217153084406,,s,629,366.0647033081055,0.5819788605852233,0.07282065333686359,0.5730969848632812,0.5741784912109374,0.5745491943359375,1.1854891552734375,"[0.5736980590820312, 0.573138916015625, 0.5735465087890625, 0.574129150390625, 0.5737092895507813, 0.57447216796875, 0.5745264892578125, 0.5731635131835937, 0.5737205810546875, 0.5735956420898437, 0.5749248046875, 0.5725880126953125, 0.5736376342773437, 0.57371337890625, 0.573749267578125, 0.5733990478515625, 0.5727344360351563, 0.57394482421875, 0.57497705078125, 0.5740851440429687, 0.5741229858398438, 0.5742008056640625, 0.5734738159179688, 0.574509033203125, 0.5738157958984375, 0.573897705078125, 0.5732402954101562, 0.5745704956054688, 0.5727620849609375, 0.5734061889648437, 0.5746319580078125, 0.5745643310546875, 0.5739735107421875, 0.5731901245117188, 0.5725511474609375, 0.5742970581054687, 0.5729157104492187, 0.572837890625, 0.57303857421875, 0.5725552368164063, 0.5740175170898437, 0.5737943115234375, 0.57314404296875, 0.5743441772460938, 0.5747333374023438, 0.5747230834960938, 0.5726239013671875, 0.5742520141601563, 0.5729638671875, 0.5725716552734375, 0.572348388671875, 0.572031005859375, 0.5730672607421875, 0.5724190673828125, 0.572590087890625, 0.5723648071289062, 0.5724487915039063, 0.572526611328125, 0.5727549438476562, 0.57301708984375, 0.5722869873046875, 0.5721016235351563, 1.18763623046875, 0.5728788452148438, 0.5735679931640625, 0.5728031005859375, 0.5727078247070313, 0.5732781982421875, 0.5730969848632812, 0.5728460693359375, 0.57259521484375, 0.5722409057617187, 0.57280615234375, 0.57242724609375, 0.57225830078125, 0.5726760864257813, 0.572142578125, 0.5723013305664062, 0.5727958984375, 0.5730846557617187, 0.572316650390625, 0.5732341918945313, 0.57354443359375, 0.5724467163085938, 0.5723648071289062, 0.5720924072265625, 0.5729249267578125, 0.5723299560546875, 0.5724815063476563, 0.572590087890625, 0.5723525390625, 0.5726453857421875, 0.572458984375, 0.5725296630859374, 0.5720729370117188, 0.572706787109375, 0.5740257568359375, 0.572758056640625, 0.5724251708984375, 0.5737236328125, 0.57303857421875, 0.5752218017578125, 0.5730856323242187, 0.57284814453125, 0.5734194946289063, 0.5729658813476562, 0.5726095581054688, 0.5726730346679687, 0.5732310791015625, 0.5731287231445312, 0.5729985961914063, 0.572822509765625, 0.5728573608398437, 0.5731840209960938, 0.5744937133789062, 0.5729187622070312, 0.5740841064453125, 0.5739561157226563, 0.5734850463867187, 0.5729812622070313, 0.5727211303710937, 0.5733560180664062, 0.5732577514648437, 0.5737369384765625, 0.5736038208007812, 1.18655078125, 0.5729157104492187, 0.5731143798828126, 0.5728522338867188, 0.5734451293945313, 0.5739530029296875, 0.572958740234375, 0.572759033203125, 0.5748234252929687, 0.57375537109375, 0.573359130859375, 0.5733130493164063, 0.5738352661132813, 0.57318603515625, 0.5733314819335937, 0.5734747924804687, 0.5750180053710937, 0.5730693359375, 0.5736345825195313, 0.573718505859375, 0.57335498046875, 0.5729044189453125, 0.5738772583007813, 0.5738352661132813, 0.5730682983398437, 0.5731051635742187, 0.5734297485351563, 0.5730928344726562, 0.5731317749023438, 0.5730283813476562, 0.573849609375, 0.57297509765625, 0.5731461181640625, 0.5727354736328125, 0.5730130004882813, 0.5731461181640625, 0.5728123168945313, 0.5729003295898437, 0.5728778076171875, 0.5738741455078125, 0.5740830688476563, 0.572737548828125, 0.5726536254882812, 0.5742734985351563, 0.5730897827148438, 0.5729822998046875, 0.5732136840820312, 0.5756242065429688, 0.5729976196289063, 0.5738700561523438, 0.573106201171875, 0.573149169921875, 0.5729013671875, 0.573991943359375, 0.573201416015625, 0.5733877563476563, 0.5732608032226563, 0.5743902587890625, 0.5742141723632812, 0.5728829345703125, 0.5743646850585937, 0.5734010620117187, 0.573259765625, 1.1840142822265625, 0.5737728271484375, 0.5731215209960937, 0.5732106323242188, 0.5732505493164063, 0.572969970703125, 0.573033447265625, 0.57314306640625, 0.5731942138671875, 0.5728460693359375, 0.5744957275390625, 0.5742643432617187, 0.5740144653320313, 0.5731768798828125, 0.5741864624023437, 0.5734174194335937, 0.573322265625, 0.5733375854492188, 0.5739192504882813, 0.5734788818359375, 0.5737677001953125, 0.5734850463867187, 0.5734430541992187, 0.5732044677734375, 0.5726986083984374, 0.572821533203125, 0.5730499267578125, 0.5735310668945313, 0.574656494140625, 0.5733201904296875, 0.5725225219726563, 0.5739570922851562, 0.5737728271484375, 0.57303857421875, 0.5733232421875, 0.5734522705078124, 0.5738311767578125, 0.575151123046875, 0.5733201904296875, 0.5735167846679687, 0.5731133422851562, 0.5729003295898437, 0.5728235473632812, 0.573318115234375, 0.5730816040039063, 0.5742418212890625, 0.573048828125, 0.5728880615234375, 0.572990478515625, 0.5729679565429687, 0.5725675659179688, 0.5726566162109376, 0.5728235473632812, 0.5745910034179688, 0.5722357788085938, 0.5722327270507812, 0.572821533203125, 0.5726668701171875, 0.5724385375976563, 0.5728890991210938, 0.5728051147460937, 0.5724405517578125, 0.5733775634765625, 1.1864791259765626, 0.5730631713867187, 0.5741793212890625, 0.574244873046875, 0.5729197998046875, 0.5730785522460937, 0.5742643432617187, 0.5736028442382812, 0.5729208374023438, 0.5731768188476563, 0.573612060546875, 0.5728307495117188, 0.5726996459960938, 0.57314306640625, 0.5741906127929688, 0.572669921875, 0.5738731689453125, 0.5743124389648437, 0.5728778076171875, 0.5729352416992187, 0.574119873046875, 0.5733079223632812, 0.5724866333007812, 0.572231689453125, 0.5732946166992188, 0.5724682006835937, 0.5730529174804687, 0.5733519287109375, 0.573106201171875, 0.5727999877929687, 0.57236376953125, 0.5724620971679687, 0.5725665893554688, 0.5737697143554688, 0.5733406372070312, 0.573254638671875, 0.5726945190429688, 0.5743114013671875, 0.5735679931640625, 0.5726617431640625, 0.5724610595703125, 0.57396533203125, 0.5732372436523437, 0.5741332397460938, 0.5733570556640625, 0.5733375854492188, 0.572717041015625, 0.572859375, 0.5731901245117188, 0.572558349609375, 0.5723750610351562, 0.5745213623046875, 0.573633544921875, 0.5729638671875, 0.573412353515625, 0.5735505981445312, 0.5726494750976563, 0.5753558959960937, 0.5728399658203125, 0.573432861328125, 0.5723391723632812, 0.5729290161132813, 0.572626953125, 1.1842232666015624, 0.5728092041015626, 0.572779541015625, 0.572632080078125, 0.5737963256835937, 0.5736775512695312, 0.5726033935546875, 0.572564453125, 0.5733416748046875, 0.5725654907226563, 0.5723832397460937, 0.5723812255859375, 0.5731624755859375, 0.5723832397460937, 0.5729197998046875, 0.572788818359375, 0.5723237915039062, 0.5723801879882813, 0.5726771850585938, 0.5727251586914063, 0.5726986083984374, 0.57333349609375, 0.5739939575195312, 0.5731481323242188, 0.5727344360351563, 0.5737195434570312, 0.5727918090820312, 0.5725234985351563, 0.5729740600585937, 0.5731522827148438, 0.5736734619140625, 0.5729310913085938, 0.5734471435546875, 0.573127685546875, 0.5729720458984375, 0.5730723876953125, 0.5730529174804687, 0.5730682983398437, 0.5760235595703125, 0.5742151489257813, 0.573750244140625, 0.5731143798828126, 0.573844482421875, 0.5748480224609375, 0.5746339721679687, 0.5739223022460938, 0.5732044677734375, 0.5739735107421875, 0.5730263061523437, 0.5728583984375, 0.5729924926757812, 0.5730785522460937, 0.5734502563476562, 0.5732157592773437, 0.573191162109375, 0.5728983154296875, 0.5738905639648437, 0.5733457641601563, 0.57269873046875, 0.572650390625, 0.5734747924804687, 0.5728338012695312, 0.5733191528320313, 1.1859814453125, 0.57253271484375, 0.5722654418945312, 0.5727242431640625, 0.5725839233398438, 0.5722265625, 0.572601318359375, 0.5729863891601562, 0.572706787109375, 0.5726239013671875, 0.5739888916015625, 0.5729249267578125, 0.573317138671875, 0.5733673095703125, 0.5743882446289063, 0.5732106323242188, 0.5734819946289063, 0.5727047729492187, 0.5730140380859375, 0.572675048828125, 0.5729136352539063, 0.5729924926757812, 0.5723709716796875, 0.5744793701171875, 0.57339599609375, 0.5728604125976563, 0.5722327270507812, 0.5737615356445313, 0.5753026733398438, 0.5738475341796875, 0.5734512939453125, 0.5752913818359375, 0.5733673095703125, 0.5729085693359375, 0.5732177734375, 0.573886474609375, 0.5727651977539062, 0.57282763671875, 0.57284814453125, 0.5731399536132813, 0.5727467651367187, 0.5733314819335937, 0.5736110229492187, 0.5734830322265625, 0.5729003295898437, 0.574867431640625, 0.5740697631835937, 0.572969970703125, 0.5728942260742188, 0.5735751342773437, 0.5731563720703124, 0.5731113891601562, 0.5729085083007812, 0.5745828247070313, 0.5727620849609375, 0.5729269409179687, 0.5724876708984376, 0.573665283203125, 0.5732260131835938, 0.5726268920898437, 0.57246923828125, 0.572621826171875, 0.5729013671875, 1.1888896484375, 0.57364990234375, 0.5738916015625, 0.5726515502929688, 0.5726239013671875, 0.5722838745117188, 0.5730723876953125, 0.573032470703125, 0.5727662353515625, 0.5732413330078125, 0.5728880615234375, 0.5727396240234375, 0.5731368408203125, 0.573233154296875, 0.5731963500976562, 0.5735249633789062, 0.5744219970703125, 0.5727999877929687, 0.5726064453125, 0.5729290771484375, 0.5737103271484375, 0.572811279296875, 0.5732679443359375, 0.5726730346679687, 0.5734287109375, 0.5722675170898438, 0.5725828857421875, 0.5725849609375, 0.5723627319335938, 0.5727416381835938, 0.5728348388671874, 0.5723678588867187, 0.5728604125976563, 0.573675537109375, 0.57333349609375, 0.5725634765625, 0.5726546020507812, 0.5731399536132813, 0.5728031005859375, 0.5724630737304688, 0.57310205078125, 0.5734573974609375, 0.5732689819335938, 0.5729290161132813, 0.5729782104492187, 0.5726064453125, 0.572527587890625, 0.5728818969726562, 0.5733007202148438, 0.5730426635742187, 0.5736365966796875, 0.5730549926757813, 0.573111328125, 0.5725552368164063, 0.5726505126953125, 0.5732802734375, 0.5726494750976563, 0.5724876708984376, 0.5733058471679687, 0.5727211303710937, 0.5725931396484375, 0.5726648559570312, 0.5728655395507812, 1.1863818359375, 0.5727938842773438, 0.5733508911132813, 0.5740676879882812, 0.574234619140625, 0.5727047729492187, 0.5733447875976563, 0.5732515869140625, 0.5743012084960938, 0.5728123168945313, 0.5730969848632812, 0.5729976196289063, 0.5739694213867188, 0.5728798828125, 0.5730140380859375, 0.5724334106445312, 0.5727744140625, 0.5740390625, 0.5736099853515625, 0.5726300048828125, 0.572969970703125, 0.5736683349609375, 0.572948486328125, 0.5725542602539062, 0.5725962524414062, 0.5727139892578125, 0.573559814453125, 0.572568603515625, 0.5725419311523438, 0.5730426635742187, 0.5727252197265625, 0.5726781616210938, 0.5734676513671875, 0.573179931640625, 0.5726351318359375, 0.573233154296875, 0.5737062377929687, 0.5729423217773437, 0.5729710083007813, 0.5735885009765626, 0.5732741088867187, 0.5730344848632812, 0.5727262573242188, 0.5732567138671875, 0.5732689819335938, 0.5727232055664062, 0.5727559814453125, 0.5730703125, 0.5724334716796875, 0.5731378784179687, 0.5730191650390625, 0.572896240234375, 0.5732177734375, 0.5728727416992188, 0.5729361572265625, 0.5730979614257813, 0.574118896484375, 0.573497314453125, 0.5732976684570312, 0.5737840576171875, 0.5730130004882813, 0.5730549926757813, 0.572958740234375, 1.187092529296875, 0.5724129028320313, 0.5726473999023437, 0.57280615234375, 0.572416015625, 0.5723361206054688, 0.572821533203125, 0.5732761840820313, 0.5727733764648437, 0.573095947265625, 0.5734052124023438, 0.573053955078125, 0.572674072265625, 0.5729341430664062, 0.5734830322265625, 0.573849609375, 0.5724999389648437, 0.5727334594726563, 0.5741107177734375, 0.5736663208007813, 0.57402978515625, 0.5733611450195313, 0.572958740234375, 0.5730099487304687, 0.5733345336914063, 0.5739038696289063, 0.575910888671875, 0.574392333984375, 0.5739898681640625, 0.5732730712890625, 0.5733109741210938, 0.5730928344726562, 0.5741782836914062, 0.5731993408203125, 0.57302734375, 0.5732771606445313, 0.572506103515625, 0.5729556274414063, 0.5733088989257813, 0.572968994140625, 0.5736283569335937, 0.5738291015625, 0.5733499145507812, 0.5729085693359375, 0.573539306640625, 0.5743226928710937, 0.5736959838867187, 0.5733365478515625, 0.57354443359375, 0.5733949584960938, 0.5731727294921874, 0.5731215209960937, 0.5734061889648437, 0.5725133056640626, 0.5726607055664062, 0.5729464111328125, 0.5725101928710937, 0.57212109375, 0.5734563598632813, 0.5726597290039063, 0.5727313842773437, 0.572416015625, 0.5735731201171875]",tokens/s,1.7182754696526694,,,main,False,False,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-j-6b,,cuda,0,42,,,,,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,,MB,2082.578432,5566.365696,0.0,4919.918592,4635.53792,s,10,5.0974686889648435,0.5097468688964844,0.002523113409812242,0.5091204681396484,0.5118887878417968,0.5140431793212891,0.5157666925048828,"[0.5161975708007812, 0.508827880859375, 0.5094269104003907, 0.508350830078125, 0.5081865539550782, 0.5073575439453125, 0.5094130554199219, 0.5110750427246094, 0.5072232666015625, 0.5114100341796874]",tokens/s,502.2100489880333,kWh,5.994651640454928e-06,3.28480666631549e-06,2.7987258500894452e-05,3.726671680766487e-05,tokens/kWh,6869400.417569034,MB,2082.578432,5566.365696,0.0,4919.918592,4794.464768,s,10,295.3985,29.53985,0.004666201106872227,29.538681640625,29.5469361328125,29.54772978515625,29.548364707031247,"[29.534837890625, 29.546759765625, 29.542654296875, 29.540826171875, 29.5363359375, 29.535224609375, 29.54083984375, 29.5359609375, 29.5485234375, 29.536537109375]",tokens/s,2.132712251416307,kWh,0.0003485870901164082,0.0001910528620806053,0.0016025825737317144,0.002142222525928728,tokens/kWh,29408.709523623045,,s,629,299.5096441345215,0.47616795569876236,0.06052059742525611,0.46878207397460936,0.46948126220703124,0.4697759826660156,0.977483603515625,"[0.4688783264160156, 0.46847384643554685, 0.46878515625, 0.4694343566894531, 0.4695541687011719, 0.46879034423828125, 0.4688465270996094, 0.46848306274414064, 0.46828851318359377, 0.4686806945800781, 0.46836428833007815, 0.4685475769042969, 0.4686540832519531, 0.46948455810546874, 0.4689162292480469, 0.46845745849609377, 0.4684011535644531, 0.4683714599609375, 0.4684134521484375, 0.46870220947265623, 0.468790283203125, 0.4682076110839844, 0.4684062805175781, 0.46837042236328125, 0.469607421875, 0.46908721923828123, 0.4688885803222656, 0.4687656860351562, 0.4693012390136719, 0.4686929931640625, 0.46864382934570314, 0.468430908203125, 0.4682413330078125, 0.4685291442871094, 0.46904525756835935, 0.4706498413085938, 0.46837042236328125, 0.4685875244140625, 0.4687718505859375, 0.4685537414550781, 0.46882406616210937, 0.46856805419921876, 0.4686806945800781, 0.46866943359375, 0.4686581726074219, 0.4682403869628906, 0.4695777893066406, 0.4685884704589844, 0.4686663818359375, 0.4685537414550781, 0.46855987548828126, 0.4685066223144531, 0.4687831115722656, 0.46864794921875, 0.4684666748046875, 0.4691046447753906, 0.46919476318359377, 0.4684984436035156, 0.4691128234863281, 0.4691221008300781, 0.46927557373046874, 0.46959002685546875, 0.977623046875, 0.46814208984375, 0.4683407287597656, 0.46863873291015623, 0.46877490234375, 0.46880459594726565, 0.4687513732910156, 0.46857113647460935, 0.468421630859375, 0.46894284057617186, 0.4688281555175781, 0.46850765991210935, 0.4687431640625, 0.4691404724121094, 0.4699504699707031, 0.46978561401367186, 0.46937701416015626, 0.4690975341796875, 0.4691311950683594, 0.47030169677734374, 0.4690933837890625, 0.4689039306640625, 0.4690298767089844, 0.46860595703125, 0.4687359924316406, 0.4692275085449219, 0.46936166381835936, 0.46901248168945314, 0.46843902587890623, 0.46878106689453125, 0.46938113403320314, 0.46973849487304686, 0.46935861206054685, 0.4687390441894531, 0.4685845031738281, 0.46890591430664064, 0.46899813842773436, 0.4688281555175781, 0.4687083435058594, 0.4687564697265625, 0.4687308654785156, 0.4687615966796875, 0.4689858703613281, 0.4688762817382813, 0.46893264770507814, 0.46929302978515625, 0.46915994262695315, 0.4696995849609375, 0.4688701477050781, 0.4687912902832031, 0.4691885986328125, 0.46969558715820314, 0.4696114196777344, 0.4692490234375, 0.46899301147460937, 0.46896435546875, 0.4694558715820312, 0.46925927734375, 0.4685393981933594, 0.4686530456542969, 0.46877902221679685, 0.4687155151367188, 0.46934527587890623, 0.9787330322265625, 0.468917236328125, 0.46886502075195313, 0.46889984130859375, 0.4691875915527344, 0.4684267578125, 0.46861004638671877, 0.46856500244140625, 0.46833560180664063, 0.4684933166503906, 0.4683735046386719, 0.4683263854980469, 0.4683735046386719, 0.468236328125, 0.46872265625, 0.4684236755371094, 0.46842572021484374, 0.46854452514648437, 0.46859982299804687, 0.4687155151367188, 0.46844732666015626, 0.4685188293457031, 0.4686991577148438, 0.46980194091796873, 0.46855987548828126, 0.4690032653808594, 0.46884454345703125, 0.4687575073242187, 0.4688609313964844, 0.4688558044433594, 0.4689459228515625, 0.46866329956054686, 0.4685557861328125, 0.46875238037109374, 0.4686827392578125, 0.46909747314453126, 0.46889984130859375, 0.4689756164550781, 0.46924798583984373, 0.46901144409179685, 0.4689469299316406, 0.46897048950195314, 0.46984909057617186, 0.4700190734863281, 0.4696781005859375, 0.469550048828125, 0.4697763977050781, 0.47159500122070314, 0.46917837524414063, 0.4684482421875, 0.4692193298339844, 0.46906060791015625, 0.4689141845703125, 0.46894284057617186, 0.468885498046875, 0.469064697265625, 0.46948043823242186, 0.4687861633300781, 0.4690513916015625, 0.46873907470703124, 0.46875341796875, 0.4690708618164062, 0.46917938232421874, 0.9775615844726563, 0.46874624633789064, 0.46858853149414065, 0.4686796875, 0.46987161254882814, 0.46889370727539065, 0.4687656860351562, 0.468790283203125, 0.4687278137207031, 0.46867864990234376, 0.4685322265625, 0.46837454223632813, 0.4686530456542969, 0.46875955200195313, 0.46893466186523436, 0.468706298828125, 0.4689858703613281, 0.46853018188476564, 0.46881585693359373, 0.46894796752929685, 0.4691353454589844, 0.46871038818359373, 0.46851071166992186, 0.468632568359375, 0.46843902587890623, 0.46874725341796875, 0.468890625, 0.4688670654296875, 0.46844723510742187, 0.468490234375, 0.4689090576171875, 0.46941900634765626, 0.46909030151367187, 0.4689100952148437, 0.4687175598144531, 0.46938113403320314, 0.4690616455078125, 0.4687718505859375, 0.46904730224609376, 0.4689336242675781, 0.4687933349609375, 0.46920501708984375, 0.46855471801757814, 0.4687718505859375, 0.47131646728515625, 0.4688424987792969, 0.46912005615234376, 0.4692090148925781, 0.4688189392089844, 0.4685926513671875, 0.4688619384765625, 0.46930743408203124, 0.4693247680664063, 0.4690636901855469, 0.46892135620117187, 0.4687145690917969, 0.4691373291015625, 0.4687145080566406, 0.4686673889160156, 0.46867352294921877, 0.46872677612304686, 0.4689254455566406, 0.46927871704101565, 0.9772830810546875, 0.46864077758789063, 0.4684922790527344, 0.46857421875, 0.46906878662109375, 0.4690033264160156, 0.46841543579101563, 0.4685895690917969, 0.46849639892578127, 0.468279296875, 0.46869912719726564, 0.4685619201660156, 0.4687575073242187, 0.46905242919921875, 0.46849432373046873, 0.46846875, 0.46872576904296875, 0.46877490234375, 0.4686592102050781, 0.4684892272949219, 0.4687974548339844, 0.46863565063476564, 0.4685844421386719, 0.46846157836914065, 0.46825778198242185, 0.4684994506835938, 0.4686448669433594, 0.46830081176757815, 0.46856805419921876, 0.4685209655761719, 0.46874929809570315, 0.4689264526367187, 0.46845541381835937, 0.4686520385742188, 0.46867770385742186, 0.46937490844726565, 0.46933709716796873, 0.4687554626464844, 0.4690616455078125, 0.46915789794921875, 0.46891213989257813, 0.46888754272460936, 0.4687278137207031, 0.46915277099609376, 0.4688332824707031, 0.4693790588378906, 0.46926437377929686, 0.46878515625, 0.4690831298828125, 0.46895718383789065, 0.4689776611328125, 0.4690616455078125, 0.46907391357421874, 0.46851788330078126, 0.46869195556640625, 0.4686376953125, 0.46900634765625, 0.4722012023925781, 0.4687718505859375, 0.4688670654296875, 0.46880459594726565, 0.468642822265625, 0.4693155822753906, 0.9763594360351563, 0.46846771240234375, 0.46886502075195313, 0.4687083740234375, 0.4689837646484375, 0.4684646301269531, 0.4687247314453125, 0.46836224365234375, 0.4683929748535156, 0.4685619201660156, 0.46836737060546874, 0.4687503356933594, 0.46857830810546874, 0.46878411865234376, 0.4683345947265625, 0.468358154296875, 0.46845849609375, 0.4687032470703125, 0.4686520385742188, 0.4686458740234375, 0.46885784912109374, 0.46872677612304686, 0.4685823974609375, 0.46873095703125, 0.4685475158691406, 0.468790283203125, 0.46927053833007815, 0.46861822509765627, 0.46847589111328125, 0.46836224365234375, 0.470181884765625, 0.4690575256347656, 0.4688353271484375, 0.4692449340820313, 0.4685137939453125, 0.46857217407226565, 0.46875238037109374, 0.46873190307617185, 0.4689664001464844, 0.4691271667480469, 0.4686090087890625, 0.4692101135253906, 0.4687974548339844, 0.469317626953125, 0.4686315612792969, 0.46899917602539065, 0.4692490234375, 0.4690309143066406, 0.4687646789550781, 0.46850253295898436, 0.4687083435058594, 0.46883941650390626, 0.46927871704101565, 0.4686152038574219, 0.4694292297363281, 0.4691435546875, 0.46974566650390626, 0.4686940307617187, 0.468969482421875, 0.46890188598632815, 0.46895001220703125, 0.4687564697265625, 0.46904730224609376, 0.9786060791015625, 0.46841650390625, 0.4683786315917969, 0.46915994262695315, 0.46917938232421874, 0.4692777099609375, 0.46883636474609375, 0.468716552734375, 0.46857318115234375, 0.4691128234863281, 0.468600830078125, 0.4694783935546875, 0.46952960205078126, 0.4696739807128906, 0.4687196044921875, 0.46852505493164065, 0.46841854858398435, 0.4683591613769531, 0.4683100280761719, 0.4684431457519531, 0.4690411376953125, 0.4687923278808594, 0.4687216491699219, 0.46836224365234375, 0.4686090087890625, 0.4691363830566406, 0.46857113647460935, 0.46862335205078126, 0.46869094848632814, 0.4686581726074219, 0.4694435729980469, 0.4694640502929687, 0.46918450927734373, 0.46924288940429687, 0.4694343566894531, 0.4690462646484375, 0.469317626953125, 0.46890188598632815, 0.46938323974609375, 0.4693267822265625, 0.46866329956054686, 0.4690370483398438, 0.468864013671875, 0.4687575073242187, 0.4686438598632813, 0.46876776123046876, 0.46984698486328125, 0.4690380859375, 0.46935552978515627, 0.46853836059570314, 0.46839910888671876, 0.46874008178710935, 0.46875955200195313, 0.4687667236328125, 0.4688209838867187, 0.4691302490234375, 0.46921624755859376, 0.4689674377441406, 0.46906777954101564, 0.46867864990234376, 0.46859161376953123, 0.46864794921875, 0.4688332824707031, 0.980326416015625, 0.468701171875, 0.46863360595703124, 0.468738037109375, 0.46885989379882814, 0.4684267578125, 0.468389892578125, 0.46853839111328127, 0.46828131103515624, 0.4683601989746094, 0.46820352172851565, 0.467962890625, 0.4681553955078125, 0.4686499938964844, 0.4686612548828125, 0.4687503356933594, 0.46850253295898436, 0.4680570983886719, 0.4683458557128906, 0.46834994506835936, 0.46889471435546876, 0.46862130737304686, 0.4687196044921875, 0.46831512451171875, 0.4685035400390625, 0.4694077453613281, 0.4687421569824219, 0.4688087158203125, 0.4701829528808594, 0.4684830322265625, 0.4688619384765625, 0.4692244567871094, 0.46869094848632814, 0.4689715270996094, 0.46905548095703126, 0.4686315612792969, 0.468864013671875, 0.4689029235839844, 0.46895001220703125, 0.46892340087890627, 0.46898687744140627, 0.469212158203125, 0.46885272216796875, 0.4688230285644531, 0.46921624755859376, 0.46866943359375, 0.46926849365234374, 0.46906777954101564, 0.469248046875, 0.4691763000488281, 0.46952960205078126, 0.4694057006835938, 0.46893157958984377, 0.4688035888671875, 0.4690667419433594, 0.4687615966796875, 0.4690288696289063, 0.46915890502929686, 0.46848818969726563, 0.4687585144042969, 0.4686315612792969, 0.4686049194335937, 0.46867352294921877, 0.9793843383789063, 0.4697733154296875, 0.46955825805664064, 0.46915890502929686, 0.46897869873046877, 0.4683345947265625, 0.46859982299804687, 0.468316162109375, 0.46864694213867186, 0.4686110534667969, 0.46845745849609377, 0.46816357421875, 0.4684892272949219, 0.46949169921875, 0.46866842651367185, 0.46861312866210936, 0.4686090087890625, 0.4685424499511719, 0.46886605834960937, 0.4688752746582031, 0.46871353149414063, 0.4691127624511719, 0.46852615356445315, 0.46842361450195313, 0.46897665405273437, 0.4689141845703125, 0.4685926513671875, 0.46885989379882814, 0.468421630859375, 0.4683888549804687, 0.46856600952148436, 0.4689182739257812, 0.46858547973632814, 0.46870220947265623, 0.4691896362304688, 0.46975283813476565, 0.46981631469726565, 0.4697753601074219, 0.470096923828125, 0.469834716796875, 0.46959820556640625, 0.4698542175292969, 0.46952243041992187, 0.4697272338867188, 0.4697907104492188, 0.4687656860351562, 0.4691937255859375, 0.46966680908203123, 0.46944461059570314, 0.46914764404296877, 0.469760009765625, 0.4693544921875, 0.46949993896484377, 0.4696063537597656, 0.46877694702148437, 0.4688281555175781, 0.4692244567871094, 0.468706298828125, 0.4684646301269531, 0.46868582153320315, 0.4686725158691406, 0.46858547973632814, 0.4686612548828125, 0.9803509521484375, 0.46829159545898436, 0.4681697387695313, 0.4690083923339844, 0.46880972290039064, 0.46866329956054686, 0.4685547485351563, 0.46837042236328125, 0.468105224609375, 0.4683243713378906, 0.46842364501953127, 0.4681779174804688, 0.4683816833496094, 0.4685599365234375, 0.4682454528808594, 0.46819635009765626, 0.46841036987304685, 0.46834994506835936, 0.468537353515625, 0.46845849609375, 0.46887115478515623, 0.46867352294921877, 0.46852197265625, 0.46868377685546875, 0.46900018310546876, 0.46878207397460936, 0.4686499938964844, 0.4685537414550781, 0.46879437255859374, 0.4684031982421875, 0.4687779846191406, 0.4690032653808594, 0.4684912719726563, 0.46856298828125, 0.46899917602539065, 0.4696033020019531, 0.4696708984375, 0.469855224609375, 0.4690411376953125, 0.468674560546875, 0.46875442504882814, 0.469359619140625, 0.4701552734375, 0.469064697265625, 0.46861312866210936, 0.4686315612792969, 0.4697047119140625, 0.4690083923339844, 0.46888754272460936, 0.46868582153320315, 0.469073974609375, 0.4693411254882813, 0.4695132141113281, 0.46920089721679686, 0.46941696166992186, 0.46919271850585936, 0.46985626220703125, 0.46913946533203127, 0.46915789794921875, 0.4686253967285156, 0.46875955200195313, 0.468642822265625, 0.4691517333984375]",tokens/s,2.1000993200656053,,,main,False,False,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-0.5B,Qwen/Qwen1.5-0.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,1460.260864,1709.703168,0.0,1063.256064,942.605312,s,10,0.8792122802734375,0.08792122802734374,0.0017793873831683257,0.08834756851196289,0.08902708129882812,0.09042998733520508,0.09155231216430663,"[0.09183289337158203, 0.08871532440185546, 0.08646198272705079, 0.08839430236816406, 0.08854621124267578, 0.08836022186279296, 0.08833491516113282, 0.08542150115966797, 0.08774205017089844, 0.08540287780761718]",tokens/s,2911.6972742962976,kWh,1.0152025072210082e-06,5.56282725791299e-07,2.6184398915211728e-06,4.1899251245334795e-06,tokens/kWh,61098943.869194776,MB,1460.588544,1709.703168,0.0,1063.256064,942.607872,s,10,54.0859736328125,5.40859736328125,0.07705683066534309,5.45104150390625,5.477568017578125,5.481690747070313,5.484988930664063,"[5.4858134765625, 5.457708984375, 5.47665185546875, 5.46595458984375, 5.472109375, 5.4443740234375, 5.39297412109375, 5.29275341796875, 5.29816357421875, 5.29947021484375]",tokens/s,11.648121641981426,kWh,6.466018401433034e-05,3.543631727276219e-05,0.00015379821301947998,0.0002538947143065725,tokens/kWh,248134.35038245353,,s,629,54.78888442230225,0.08710474470954252,0.010581359575427846,0.08675225830078125,0.08761917724609375,0.0884242431640625,0.17260260864257815,"[0.08388607788085937, 0.08398233795166016, 0.08383897399902343, 0.08367718505859376, 0.08453632354736328, 0.0890890884399414, 0.08862201690673828, 0.08834764862060547, 0.08846745300292969, 0.08839679718017578, 0.08905522918701173, 0.08837939453125, 0.08853196716308594, 0.08820326232910156, 0.08852684783935547, 0.08853094482421875, 0.08833126068115234, 0.0883599395751953, 0.08833229064941406, 0.08829747009277343, 0.08771071624755859, 0.08838553619384766, 0.08845721435546874, 0.08875212860107422, 0.08825856018066407, 0.08850534057617188, 0.08816230773925782, 0.08822886657714844, 0.08858009338378907, 0.0882339859008789, 0.08867430114746094, 0.08864665222167968, 0.08844902038574219, 0.08828108978271484, 0.0884295654296875, 0.08401407623291016, 0.08382259368896484, 0.0838635482788086, 0.08400077056884765, 0.08771788787841797, 0.0872069091796875, 0.08849203491210937, 0.08865280151367187, 0.08827187347412109, 0.08736563110351563, 0.08729804992675781, 0.08718950653076171, 0.08671231842041016, 0.08691506958007812, 0.08681779479980468, 0.0866529312133789, 0.086940673828125, 0.08705535888671875, 0.08424960327148437, 0.08400383758544921, 0.08397414398193359, 0.08412364959716796, 0.08525926208496094, 0.08683622741699219, 0.08680038452148438, 0.08671949005126953, 0.08692940521240235, 0.1763768310546875, 0.08659661102294922, 0.08731238555908204, 0.0868884506225586, 0.08696832275390624, 0.08851148986816407, 0.08714854431152344, 0.08598323059082032, 0.08636518096923829, 0.08685465240478515, 0.08704512023925781, 0.08686284637451172, 0.08745779418945313, 0.08453529357910156, 0.0869775390625, 0.08696627044677735, 0.08664575958251954, 0.0876072998046875, 0.08689766693115235, 0.08705535888671875, 0.08679936218261719, 0.0866550064086914, 0.08538006591796875, 0.08647270202636718, 0.08609996795654297, 0.08713215637207031, 0.08740147399902344, 0.08702365112304687, 0.08689250946044921, 0.08694169616699218, 0.08711577606201172, 0.08716902160644531, 0.08723149108886719, 0.0867747802734375, 0.08696012878417969, 0.08699494171142579, 0.0842239990234375, 0.08377855682373046, 0.08404684448242188, 0.08385945892333985, 0.08593510437011719, 0.08742301177978516, 0.08683209228515625, 0.08660889434814453, 0.0867952651977539, 0.08660377502441406, 0.08784178924560547, 0.08544153594970703, 0.08774246215820312, 0.08707891082763672, 0.08705331420898438, 0.08723865509033203, 0.08705843353271485, 0.08711577606201172, 0.08695193481445312, 0.08490188598632813, 0.08705741119384766, 0.08451789093017578, 0.08685260772705078, 0.08724787139892579, 0.08672563171386719, 0.08705638122558594, 0.08634572601318359, 0.1737533416748047, 0.08698675537109375, 0.08697344207763671, 0.08704000091552734, 0.08702982330322266, 0.08781715393066407, 0.08694579315185547, 0.08745369720458984, 0.08681785583496093, 0.08706249237060547, 0.08565961456298828, 0.0864000015258789, 0.0841502685546875, 0.08501248168945312, 0.08692018890380859, 0.08674201965332032, 0.0869222412109375, 0.08693862152099609, 0.0867041244506836, 0.08703385925292968, 0.08675225830078125, 0.086761474609375, 0.08619315338134766, 0.08686489868164063, 0.08680652618408204, 0.08746598052978516, 0.08694783782958984, 0.08730828857421875, 0.08752537536621094, 0.08693350219726563, 0.08705023956298828, 0.08706867218017578, 0.08747212982177735, 0.0880005111694336, 0.08732262420654296, 0.08604876708984376, 0.08707180786132812, 0.08670201873779297, 0.0873164825439453, 0.08682189178466797, 0.08709529876708984, 0.08716492462158203, 0.08606719970703125, 0.08672358703613281, 0.08690790557861328, 0.0867962875366211, 0.08702361297607422, 0.08684031677246094, 0.08695507049560547, 0.08671123504638673, 0.08706867218017578, 0.08863744354248047, 0.08740863800048829, 0.08729503631591796, 0.08682080078125, 0.08690380859375, 0.08818892669677734, 0.08729702758789062, 0.08700415802001953, 0.08678809356689453, 0.0868106231689453, 0.08693657684326171, 0.08690995025634765, 0.17623654174804687, 0.08711167907714844, 0.08703897857666015, 0.08722227478027343, 0.08726322937011718, 0.08703590393066406, 0.08675942230224609, 0.08699088287353515, 0.08680445098876953, 0.08698777770996094, 0.08604876708984376, 0.08687308502197266, 0.08715878295898437, 0.08707379150390625, 0.087108642578125, 0.08752022552490234, 0.08703385925292968, 0.08690995025634765, 0.0869939193725586, 0.08686182403564453, 0.08672665405273437, 0.08708710479736329, 0.086866943359375, 0.08725094604492188, 0.08706047821044922, 0.08675945281982422, 0.08477897644042968, 0.08724479675292969, 0.08646656036376953, 0.086866943359375, 0.08677273559570313, 0.08715058898925782, 0.08593714904785156, 0.08755712127685547, 0.08769945526123046, 0.08737586975097657, 0.0873512954711914, 0.08723763275146484, 0.08681779479980468, 0.08708710479736329, 0.08671129608154297, 0.08673894500732422, 0.08689766693115235, 0.0878828125, 0.08707987213134766, 0.08682710266113282, 0.08714435577392578, 0.08673996734619141, 0.08677581024169922, 0.08577126312255859, 0.08668160247802735, 0.08681574249267578, 0.0868136978149414, 0.08693555450439452, 0.08668364715576173, 0.0859504623413086, 0.08378470611572265, 0.08481996917724609, 0.08730931091308594, 0.08659455871582031, 0.08389631652832032, 0.08560025787353516, 0.08680038452148438, 0.17275801086425782, 0.08369971466064453, 0.08384204864501953, 0.08377139282226563, 0.08545689392089843, 0.08574976348876953, 0.08727859497070313, 0.08688025665283203, 0.0867799072265625, 0.08705741119384766, 0.08711580657958984, 0.08679933166503906, 0.0867962875366211, 0.08704307556152344, 0.08686386871337891, 0.08696320343017579, 0.08698880004882813, 0.08742403411865235, 0.0870973129272461, 0.08734207916259766, 0.08837529754638672, 0.08715980529785156, 0.08705023956298828, 0.086793212890625, 0.0869744644165039, 0.08696217346191407, 0.08680550384521485, 0.08711167907714844, 0.08749362945556641, 0.08701235198974609, 0.08706150054931641, 0.08685772705078125, 0.08689766693115235, 0.08696115112304688, 0.0865955810546875, 0.08719564819335937, 0.08694989013671875, 0.08730931091308594, 0.08871324920654297, 0.08721711730957031, 0.08766668701171874, 0.08723865509033203, 0.08547942352294922, 0.088416259765625, 0.08738614654541016, 0.08706454467773438, 0.08716083526611328, 0.08708815765380859, 0.08697238159179688, 0.08681676483154296, 0.08706559753417968, 0.0872499237060547, 0.08688639831542969, 0.08803123474121094, 0.0873512954711914, 0.08714649963378907, 0.08679936218261719, 0.08706047821044922, 0.08689254760742188, 0.08732466888427734, 0.0865843505859375, 0.08680239868164062, 0.08697548675537109, 0.1747763214111328, 0.08670310211181641, 0.08710451507568359, 0.08699903869628907, 0.08693657684326171, 0.08607334136962891, 0.0869969940185547, 0.08681574249267578, 0.08696627044677735, 0.08683827209472657, 0.08736153411865234, 0.08679219055175781, 0.08696012878417969, 0.086830078125, 0.0871352310180664, 0.08736870574951172, 0.08679424285888672, 0.08719462585449218, 0.08674508666992188, 0.08693862152099609, 0.08799231719970703, 0.08773529815673828, 0.08741171264648437, 0.08712397003173829, 0.08704512023925781, 0.08727654266357422, 0.08699903869628907, 0.08696729278564454, 0.0868853759765625, 0.08697344207763671, 0.08712499237060548, 0.08712499237060548, 0.0865771484375, 0.08749158477783203, 0.08688127899169922, 0.08696524810791016, 0.08687923431396484, 0.08711475372314453, 0.08697856140136719, 0.08686489868164063, 0.08696832275390624, 0.0843724822998047, 0.08361881256103515, 0.08333516693115234, 0.08410316467285156, 0.08653314971923828, 0.0870778579711914, 0.08714342498779297, 0.08421171569824219, 0.08394239807128906, 0.08393215942382813, 0.08404377746582031, 0.08362393951416015, 0.08348060607910156, 0.08413180541992188, 0.08405094146728516, 0.08691302490234375, 0.08714854431152344, 0.08712191772460938, 0.08708812713623047, 0.08714649963378907, 0.08697241973876953, 0.08708812713623047, 0.175857666015625, 0.08723865509033203, 0.08733695983886719, 0.08694886779785156, 0.08750080108642579, 0.08700927734375, 0.0869959716796875, 0.08687718200683593, 0.08966963195800781, 0.0872069091796875, 0.08702873229980469, 0.0869713897705078, 0.08688333129882812, 0.08693247985839844, 0.08705228424072266, 0.08724275207519532, 0.0872847671508789, 0.08704819488525391, 0.08684848022460938, 0.08682803344726563, 0.08715570831298829, 0.08688742065429687, 0.08504934692382812, 0.08777011108398437, 0.08640819549560547, 0.08897232055664063, 0.08616751861572265, 0.0871905288696289, 0.0890224609375, 0.0864686050415039, 0.08654438018798828, 0.08633036804199219, 0.08477593231201172, 0.08412876892089843, 0.08437452697753907, 0.08408985900878906, 0.08387686157226562, 0.08403353881835937, 0.08393318176269532, 0.08396185302734376, 0.08435302734375, 0.08403558349609375, 0.08378470611572265, 0.0841338882446289, 0.08378470611572265, 0.08367820739746094, 0.08367922973632813, 0.0835563507080078, 0.08376831817626954, 0.08383078765869141, 0.08380723571777343, 0.0837734375, 0.08563814544677735, 0.08487833404541016, 0.08407552337646484, 0.0837580795288086, 0.0836485137939453, 0.08376012420654297, 0.08358809661865234, 0.08362290954589843, 0.08371814727783203, 0.08357478332519531, 0.08497869110107421, 0.1696030731201172, 0.08372121429443359, 0.08371916961669922, 0.0837918701171875, 0.08386457824707032, 0.08376319885253906, 0.08395263671875, 0.08374784088134765, 0.08342527770996094, 0.08687206268310547, 0.08412262725830078, 0.08380416107177735, 0.08415436553955079, 0.0838440933227539, 0.08380316925048828, 0.08431715393066407, 0.08392499542236329, 0.08376319885253906, 0.08381747436523437, 0.08540467071533203, 0.08532991790771484, 0.08402022552490235, 0.084421630859375, 0.0841195526123047, 0.08383897399902343, 0.08380623626708984, 0.08381231689453125, 0.08393830108642578, 0.08379705810546875, 0.08394233703613281, 0.08370175933837891, 0.08405811309814454, 0.08380210876464844, 0.0837734375, 0.08362290954589843, 0.08380006408691407, 0.08357478332519531, 0.08359935760498047, 0.08378470611572265, 0.0834549789428711, 0.08362188720703125, 0.08369152069091797, 0.08387686157226562, 0.08376627349853516, 0.08377961730957031, 0.0838430404663086, 0.08387686157226562, 0.08553778839111328, 0.0837232666015625, 0.08371916961669922, 0.08374476623535156, 0.08370073699951172, 0.08382259368896484, 0.08384921264648437, 0.08401203155517578, 0.08357990264892579, 0.08365055847167968, 0.08378575897216797, 0.08367919921875, 0.08369055938720703, 0.0851127700805664, 0.08426700592041016, 0.08618905639648437, 0.1722030029296875, 0.08384819030761718, 0.08397721862792969, 0.0837570571899414, 0.08369664001464844, 0.0834549789428711, 0.08403865814208984, 0.08388505554199219, 0.08392908477783204, 0.08388198089599609, 0.08392704010009766, 0.08387379455566406, 0.0837027816772461, 0.08390860748291015, 0.08369254302978515, 0.08369766235351563, 0.08587161254882812, 0.0840478744506836, 0.08396288299560548, 0.08394649505615234, 0.08418099212646485, 0.08524288177490234, 0.08649318695068359, 0.0840816650390625, 0.08387174224853515, 0.08375603485107422, 0.08405811309814454, 0.08390758514404296, 0.0839393310546875, 0.08409497833251953, 0.08389836883544922, 0.08448000335693359, 0.08434893035888671, 0.08451583862304687, 0.0842639389038086, 0.08398745727539063, 0.08408370971679688, 0.08378880310058594, 0.08385740661621094, 0.08418406677246094, 0.08394035339355468, 0.08402124786376954, 0.08522137451171875, 0.08547840118408204, 0.08503705596923829, 0.08401510620117188, 0.08397516632080078, 0.08380723571777343, 0.08466022491455077, 0.0840079345703125, 0.08379193878173828, 0.08372525024414063, 0.08389631652832032, 0.0839393310546875, 0.08357071685791016, 0.08390652465820313, 0.08391577911376953, 0.0837550048828125, 0.08391474914550781, 0.08372633361816406, 0.08374476623535156, 0.08390758514404296, 0.0837550048828125, 0.1732474822998047, 0.08425472259521484, 0.0840273895263672, 0.08414924621582032, 0.08400179290771484, 0.08393830108642578, 0.08408268737792969, 0.08393727874755859, 0.08398745727539063, 0.08395468902587891, 0.08501862335205078, 0.08602214050292968, 0.08572313690185547, 0.08382463836669922, 0.08390962982177734, 0.08375296020507812, 0.08380928039550781, 0.0840263671875, 0.08383385467529297, 0.08391986846923828, 0.08388505554199219, 0.083666015625, 0.08356137847900391, 0.08396697235107421, 0.08387481689453125, 0.08371916961669922, 0.08389119720458985, 0.08397926330566406, 0.0838656005859375, 0.08394751739501953, 0.08407244873046875, 0.0837232666015625, 0.08388813018798828, 0.08374578857421874, 0.08396185302734376, 0.08385330963134766, 0.08368434906005859, 0.08367206573486329, 0.08419328308105468, 0.08365875244140625, 0.08380825805664062, 0.08445235443115234, 0.0841533432006836, 0.08358707427978515, 0.0837027816772461, 0.08361472320556641, 0.08509645080566407, 0.08371302032470704, 0.08371097564697266, 0.08373248291015625, 0.08388198089599609, 0.08400691223144531, 0.08375910186767578, 0.08609996795654297, 0.08662732696533203, 0.08376422119140625, 0.08392396545410157, 0.08380006408691407, 0.08389119720458985, 0.0837754898071289, 0.0837734375, 0.0838175048828125, 0.08397615814208985]",tokens/s,11.480430869002337,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-160m,EleutherAI/pythia-160m,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,1303.977984,1030.22592,0.0,383.778816,312.280064,s,10,0.30782016181945804,0.030782016181945804,0.001617235684294077,0.030563695907592774,0.031164259910583494,0.03323989057540893,0.03490039510726929,"[0.03531552124023438, 0.03054502487182617, 0.030670495986938478, 0.029116479873657227, 0.030544319152832032, 0.02917731285095215, 0.030620607376098632, 0.030559423446655274, 0.030567968368530273, 0.030703008651733397]",tokens/s,8316.544260351227,kWh,3.557098667864181e-07,1.9491237936146868e-07,8.339715507108577e-07,1.3845937968587447e-06,tokens/kWh,184891771.5656334,MB,1303.977984,1030.22592,0.0,383.778816,347.090432,s,10,18.765849975585937,1.8765849975585938,0.01629298714952931,1.8788928833007814,1.8852921142578125,1.8978253173828123,1.9078518798828124,"[1.9103585205078124, 1.8825069580078124, 1.879499267578125, 1.8580986328125, 1.8786011962890625, 1.8434554443359374, 1.8807445068359374, 1.8754886474609376, 1.8791845703125, 1.8779122314453125]",tokens/s,33.5716208335684,kWh,2.201072483317281e-05,1.2059537590589184e-05,4.825437327548859e-05,8.232463569925058e-05,tokens/kWh,765263.0280703872,,s,629,19.00581683158875,0.030215925010474952,0.003670631507539437,0.02977996826171875,0.03002470397949219,0.03029995498657227,0.05998026794433601,"[0.03094118309020996, 0.03138150405883789, 0.03138355255126953, 0.031524864196777344, 0.030980096817016602, 0.03207884979248047, 0.032363521575927735, 0.03149004745483398, 0.03120844841003418, 0.031268863677978515, 0.03084492874145508, 0.032194561004638675, 0.031160320281982422, 0.03057868766784668, 0.030236671447753907, 0.03517747116088867, 0.030717952728271485, 0.030108671188354492, 0.03002470397949219, 0.029833215713500977, 0.029899776458740233, 0.02977177619934082, 0.02977791976928711, 0.02977382469177246, 0.030176256179809572, 0.0299683837890625, 0.029740032196044923, 0.02977894401550293, 0.029844480514526366, 0.029732864379882814, 0.02974412727355957, 0.029797376632690428, 0.029791231155395507, 0.029726720809936522, 0.029648895263671874, 0.029749248504638674, 0.02977791976928711, 0.029949951171875, 0.02977996826171875, 0.02976255989074707, 0.02981990432739258, 0.029797376632690428, 0.0297728328704834, 0.02997039985656738, 0.029839359283447265, 0.029833215713500977, 0.029870080947875976, 0.02978508758544922, 0.02968780708312988, 0.029784063339233398, 0.02979327964782715, 0.029755392074584962, 0.02972774314880371, 0.029976640701293945, 0.02982700729370117, 0.029852672576904295, 0.029896703720092774, 0.029731840133666993, 0.029747200012207032, 0.02983628845214844, 0.029772800445556642, 0.029739007949829102, 0.06072115325927734, 0.029829120635986327, 0.02977689552307129, 0.029834239959716798, 0.030591999053955078, 0.03021414375305176, 0.029755392074584962, 0.02983628845214844, 0.02980352020263672, 0.029842432022094727, 0.029707263946533204, 0.029665279388427734, 0.029698047637939453, 0.02979430389404297, 0.029593599319458007, 0.02993152046203613, 0.0301527042388916, 0.029792255401611328, 0.029764608383178712, 0.029861888885498046, 0.03102822494506836, 0.030027776718139648, 0.02980147171020508, 0.030087167739868165, 0.02975129508972168, 0.030086143493652344, 0.029826047897338868, 0.02974617576599121, 0.030042112350463866, 0.029730815887451172, 0.029899776458740233, 0.029838336944580077, 0.02994790458679199, 0.02980659294128418, 0.02976972770690918, 0.029682687759399414, 0.02978816032409668, 0.029817855834960938, 0.029734912872314452, 0.029831167221069335, 0.02974208068847656, 0.02978508758544922, 0.029848575592041016, 0.029800447463989257, 0.029785120010375976, 0.029689823150634766, 0.03014553642272949, 0.029886463165283202, 0.03018035125732422, 0.029938688278198244, 0.02973695945739746, 0.030068735122680663, 0.02973695945739746, 0.029897727966308595, 0.0297891845703125, 0.0298024959564209, 0.02995814323425293, 0.02976051139831543, 0.02976870346069336, 0.02997555160522461, 0.029975584030151367, 0.029692895889282228, 0.029708288192749024, 0.061895679473876954, 0.030073856353759764, 0.02974412727355957, 0.029680639266967773, 0.029848575592041016, 0.029882368087768556, 0.029543424606323244, 0.02978713607788086, 0.029755392074584962, 0.029875200271606447, 0.029856767654418945, 0.029831167221069335, 0.029770751953125, 0.02977791976928711, 0.02973695945739746, 0.029897727966308595, 0.030079999923706056, 0.029894655227661132, 0.029703168869018554, 0.0297574405670166, 0.029837312698364257, 0.029813760757446288, 0.029814783096313476, 0.029743167877197267, 0.029764543533325194, 0.029707263946533204, 0.02977791976928711, 0.029730815887451172, 0.030044160842895507, 0.029833215713500977, 0.02974617576599121, 0.029815807342529296, 0.029740032196044923, 0.029815807342529296, 0.029816831588745117, 0.029799423217773437, 0.029713407516479492, 0.0297891845703125, 0.02983628845214844, 0.029896703720092774, 0.029834239959716798, 0.0297574405670166, 0.029868032455444334, 0.029748224258422853, 0.029849599838256836, 0.02977382469177246, 0.02986604881286621, 0.029807552337646485, 0.030026752471923827, 0.02976255989074707, 0.02976972770690918, 0.029892608642578124, 0.029701120376586915, 0.02976870346069336, 0.02976255989074707, 0.029797376632690428, 0.02977177619934082, 0.029867008209228517, 0.029808639526367187, 0.02974515151977539, 0.029740032196044923, 0.029821952819824218, 0.030018560409545897, 0.05799935913085937, 0.028490751266479493, 0.028519424438476562, 0.028368896484375, 0.028404735565185548, 0.02834022331237793, 0.028285951614379884, 0.02834739112854004, 0.028308479309082032, 0.028387327194213868, 0.028321792602539062, 0.028298240661621094, 0.028923904418945313, 0.031113216400146484, 0.030136320114135744, 0.02981990432739258, 0.029868032455444334, 0.029728832244873046, 0.02970515251159668, 0.029386751174926756, 0.0297523193359375, 0.029822975158691405, 0.02971238327026367, 0.02978508758544922, 0.02980352020263672, 0.02977484893798828, 0.029677600860595704, 0.029713375091552734, 0.029651968002319336, 0.02973798370361328, 0.029881343841552735, 0.029772800445556642, 0.02979532814025879, 0.029672447204589843, 0.029861888885498046, 0.029646848678588866, 0.029551616668701174, 0.029764608383178712, 0.02973695945739746, 0.029662208557128908, 0.02970419120788574, 0.029772800445556642, 0.02975436782836914, 0.02972876739501953, 0.02979430389404297, 0.029699071884155274, 0.02981888008117676, 0.029753376007080078, 0.02983011245727539, 0.029718528747558592, 0.029489152908325194, 0.029640703201293944, 0.029736991882324218, 0.02938057518005371, 0.029748224258422853, 0.029775871276855468, 0.02980659294128418, 0.029814783096313476, 0.029716512680053712, 0.029680608749389648, 0.02979635238647461, 0.029783039093017577, 0.02978816032409668, 0.06087680053710937, 0.029916160583496092, 0.030342144012451173, 0.029867008209228517, 0.029744159698486327, 0.029726688385009765, 0.02975846481323242, 0.029857791900634766, 0.0297891845703125, 0.029784063339233398, 0.02979020881652832, 0.029849599838256836, 0.029449216842651366, 0.02939084815979004, 0.029541376113891602, 0.029874176025390626, 0.029816831588745117, 0.029724672317504884, 0.029915136337280275, 0.029921279907226563, 0.029829120635986327, 0.029734912872314452, 0.02981888008117676, 0.02980147171020508, 0.029860864639282225, 0.029874176025390626, 0.029817855834960938, 0.029855743408203125, 0.029718528747558592, 0.029854719161987304, 0.029833215713500977, 0.02995199966430664, 0.02972979164123535, 0.02980147171020508, 0.029864959716796875, 0.030071807861328126, 0.029874208450317383, 0.02984239959716797, 0.02975027275085449, 0.02978508758544922, 0.029832191467285156, 0.029860864639282225, 0.02976255989074707, 0.029702144622802733, 0.029920255661010742, 0.02975948715209961, 0.02994790458679199, 0.02977382469177246, 0.02973388862609863, 0.029643775939941407, 0.029703168869018554, 0.029852672576904295, 0.029839359283447265, 0.02983628845214844, 0.029784063339233398, 0.029769792556762695, 0.029781951904296874, 0.02998784065246582, 0.02981068801879883, 0.02998784065246582, 0.0297574405670166, 0.029809696197509766, 0.029711328506469726, 0.05807513427734375, 0.02834329605102539, 0.028271615982055662, 0.02834329605102539, 0.028402687072753906, 0.028421119689941408, 0.028279808044433592, 0.028318719863891603, 0.028318719863891603, 0.02834022331237793, 0.02835148811340332, 0.02838118362426758, 0.02836172866821289, 0.028310527801513673, 0.02832793617248535, 0.02815795135498047, 0.028197887420654297, 0.02831667137145996, 0.028222463607788087, 0.028233728408813476, 0.028296192169189452, 0.028251136779785156, 0.028318719863891603, 0.02973798370361328, 0.030136320114135744, 0.029817855834960938, 0.029731840133666993, 0.02978099250793457, 0.02972979164123535, 0.029930496215820314, 0.029886463165283202, 0.029709312438964845, 0.029739007949829102, 0.029848575592041016, 0.02972774314880371, 0.02976870346069336, 0.029648895263671874, 0.029732864379882814, 0.029813760757446288, 0.029820928573608397, 0.029755392074584962, 0.029713407516479492, 0.029643775939941407, 0.02977996826171875, 0.029856767654418945, 0.029802528381347657, 0.029723615646362306, 0.029940736770629882, 0.029823007583618163, 0.029761503219604492, 0.02976972770690918, 0.029691904067993165, 0.02968783950805664, 0.02980246353149414, 0.029711360931396483, 0.029895679473876953, 0.029684736251831056, 0.029894655227661132, 0.029791231155395507, 0.029692928314208986, 0.029740032196044923, 0.029820928573608397, 0.030217216491699218, 0.06100377655029297, 0.029852672576904295, 0.02978508758544922, 0.029661184310913087, 0.029660160064697266, 0.02972368049621582, 0.029649887084960937, 0.02981171226501465, 0.02997555160522461, 0.02975846481323242, 0.02975846481323242, 0.029682687759399414, 0.029800447463989257, 0.02975129508972168, 0.030038015365600586, 0.02983526420593262, 0.02981888008117676, 0.029861888885498046, 0.029864959716796875, 0.029840383529663086, 0.030003200531005858, 0.0307906551361084, 0.03020697593688965, 0.029899776458740233, 0.02975846481323242, 0.02978508758544922, 0.029917184829711913, 0.02975129508972168, 0.029702144622802733, 0.02978713607788086, 0.02976972770690918, 0.02976665687561035, 0.029708288192749024, 0.029823999404907226, 0.029826047897338868, 0.02983628845214844, 0.03013222312927246, 0.03099545669555664, 0.030038015365600586, 0.029878271102905272, 0.029697023391723632, 0.02977791976928711, 0.029929471969604493, 0.029862911224365234, 0.029932544708251952, 0.029714431762695313, 0.02996940803527832, 0.02980659294128418, 0.02976051139831543, 0.02951372718811035, 0.029684736251831056, 0.029820928573608397, 0.02977996826171875, 0.029778976440429688, 0.029754335403442383, 0.029816831588745117, 0.02976563262939453, 0.029755456924438477, 0.02987615966796875, 0.029709312438964845, 0.029723743438720703, 0.029749151229858398, 0.02977382469177246, 0.06091059112548828, 0.03012403106689453, 0.0297574405670166, 0.029799423217773437, 0.02974515151977539, 0.02982809638977051, 0.029861888885498046, 0.029743104934692382, 0.02978201675415039, 0.029713407516479492, 0.029759519577026366, 0.029883359909057616, 0.02977484893798828, 0.02979532814025879, 0.029874176025390626, 0.029875200271606447, 0.029850624084472657, 0.02984351921081543, 0.02972358322143555, 0.029660160064697266, 0.02992742347717285, 0.02974515151977539, 0.02968780708312988, 0.029831167221069335, 0.029741056442260744, 0.029732864379882814, 0.029642751693725586, 0.029642751693725586, 0.029895679473876953, 0.029628416061401368, 0.029723648071289063, 0.029618175506591796, 0.029715456008911133, 0.029864959716796875, 0.029732864379882814, 0.02978201675415039, 0.029748224258422853, 0.029813760757446288, 0.02982707214355469, 0.0297574405670166, 0.029854719161987304, 0.02972876739501953, 0.029657087326049804, 0.029668352127075196, 0.029611007690429687, 0.029412351608276367, 0.02972876739501953, 0.029885440826416015, 0.029694976806640624, 0.02976870346069336, 0.029775871276855468, 0.02993152046203613, 0.02977894401550293, 0.029700096130371095, 0.02977382469177246, 0.029710336685180663, 0.029688831329345702, 0.029684736251831056, 0.029700096130371095, 0.02970419120788574, 0.029667327880859375, 0.029730815887451172, 0.029863935470581054, 0.060827648162841794, 0.02976563262939453, 0.029716480255126954, 0.029770751953125, 0.02973388862609863, 0.02977689552307129, 0.02975846481323242, 0.029849599838256836, 0.0298024959564209, 0.029850624084472657, 0.029688831329345702, 0.029731840133666993, 0.029740032196044923, 0.029731840133666993, 0.029859840393066408, 0.02975027275085449, 0.029707263946533204, 0.029718528747558592, 0.02973388862609863, 0.029772800445556642, 0.029853696823120116, 0.02976563262939453, 0.029611007690429687, 0.02977996826171875, 0.02979430389404297, 0.029807615280151366, 0.029343807220458984, 0.029495231628417967, 0.029894655227661132, 0.03171123123168945, 0.030046207427978516, 0.03002470397949219, 0.029706239700317383, 0.02974412727355957, 0.029679616928100585, 0.029911039352416992, 0.02976051139831543, 0.029823999404907226, 0.029944831848144532, 0.02991926383972168, 0.029790176391601562, 0.02977382469177246, 0.029692928314208986, 0.029726720809936522, 0.029773855209350587, 0.029737951278686524, 0.02981990432739258, 0.029906944274902345, 0.029869056701660155, 0.029907968521118163, 0.029767679214477538, 0.02978201675415039, 0.02968780708312988, 0.02976563262939453, 0.029872127532958984, 0.02996428871154785, 0.02977689552307129, 0.02996326446533203, 0.029944831848144532, 0.029889568328857422, 0.029706207275390625, 0.029854719161987304, 0.029902847290039062, 0.06100787353515625, 0.029770816802978516, 0.029812671661376952, 0.02975027275085449, 0.029656063079833983, 0.029666303634643554, 0.02972979164123535, 0.0297256965637207, 0.029929471969604493, 0.029775871276855468, 0.029690879821777344, 0.029693952560424806, 0.02958028793334961, 0.029640703201293944, 0.02977996826171875, 0.029739007949829102, 0.03000115203857422, 0.029897727966308595, 0.029840383529663086, 0.029814783096313476, 0.0297523193359375, 0.029708288192749024, 0.02975027275085449, 0.02981068801879883, 0.029732864379882814, 0.029838336944580077, 0.029718528747558592, 0.02970419120788574, 0.0297205753326416, 0.029740032196044923, 0.02974617576599121, 0.02969599914550781, 0.029757503509521485, 0.0298853759765625, 0.02979840087890625, 0.029700096130371095, 0.029752351760864257, 0.02980656051635742, 0.02973695945739746, 0.029663232803344725, 0.02972159957885742, 0.030027776718139648, 0.029971519470214845, 0.029904832839965822, 0.029829120635986327, 0.02974412727355957, 0.030027776718139648, 0.029853696823120116, 0.029837312698364257, 0.029840383529663086, 0.029850624084472657, 0.02986911964416504, 0.030047168731689455, 0.030050304412841795, 0.029815807342529296, 0.029775871276855468, 0.02998784065246582, 0.029890560150146486, 0.02975334358215332, 0.029883392333984377, 0.02983526420593262, 0.029847551345825195, 0.02976972770690918]",tokens/s,33.09513111557333,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Salesforce/codegen-16B-nl,Salesforce/codegen-16B-nl,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,1736.183808,12352.749568,0.0,11706.302464,11035.465216,s,10,13.024343627929689,1.3024343627929689,0.002338564901261307,1.30185107421875,1.3038882202148436,1.3064520446777343,1.3085031042480468,"[1.309015869140625, 1.3033184814453125, 1.3003494873046875, 1.3011973876953125, 1.3011253662109374, 1.300955810546875, 1.301638427734375, 1.302063720703125, 1.3022261962890624, 1.302452880859375]",tokens/s,196.55501061184225,kWh,1.5360654360718197e-05,8.415124671664671e-06,7.215477994602094e-05,9.593055897840379e-05,tokens/kWh,2668596.9802139024,MB,1736.183808,12352.749568,0.0,11706.302464,11329.172992,s,10,763.3166562499999,76.331665625,0.01404827179512581,76.33404687500001,76.34705,76.34801328125,76.34878390625,"[76.3116171875, 76.3329609375, 76.3489765625, 76.3411171875, 76.3468359375, 76.331546875, 76.3441953125, 76.3351328125, 76.31521875, 76.3090546875]",tokens/s,0.8253455428249735,kWh,0.0009011750282347202,0.0004939259724664952,0.0042084087556131695,0.005603509756314386,tokens/kWh,11242.953566558468,,s,629,773.821041748047,1.2302401299650985,0.15462762199250651,1.21159375,1.2122087890625,1.2124819091796875,2.513270859375,"[1.2114217529296876, 1.2109813232421875, 1.2112762451171875, 1.2113695068359376, 1.2109967041015626, 1.21078369140625, 1.210820556640625, 1.2114083251953125, 1.211473876953125, 1.2109915771484374, 1.2110264892578124, 1.211177001953125, 1.2111329345703126, 1.2108482666015625, 1.210894287109375, 1.2108328857421875, 1.2114595947265625, 1.21109912109375, 1.2110377197265625, 1.2114605712890625, 1.21109814453125, 1.2109107666015626, 1.2113837890625, 1.211439208984375, 1.2113089599609375, 1.2114852294921874, 1.2108953857421876, 1.211968505859375, 1.2109854736328125, 1.2114114990234375, 1.211503662109375, 1.2115701904296876, 1.2111954345703124, 1.212129150390625, 1.211504638671875, 1.2116378173828124, 1.21143701171875, 1.211441162109375, 1.2110633544921876, 1.21100390625, 1.210900390625, 1.2119766845703126, 1.21135302734375, 1.211661376953125, 1.21140625, 1.211376708984375, 1.21117578125, 1.211125732421875, 1.2112117919921874, 1.2114166259765624, 1.2111329345703126, 1.2116644287109375, 1.2111021728515625, 1.211261962890625, 1.2113489990234374, 1.2112978515625, 1.211167724609375, 1.211284423828125, 1.2117083740234376, 1.2115958251953125, 1.2114329833984374, 1.2118804931640625, 2.514330810546875, 1.210652587890625, 1.2109178466796875, 1.21115234375, 1.210639404296875, 1.2111728515625, 1.2111964111328124, 1.2109495849609375, 1.210982421875, 1.210829833984375, 1.211577392578125, 1.2116376953125, 1.2123822021484374, 1.2119183349609375, 1.2111278076171874, 1.2120401611328124, 1.2116695556640624, 1.2114871826171876, 1.2117001953125, 1.2118865966796875, 1.211429931640625, 1.2113817138671874, 1.21173095703125, 1.2116612548828125, 1.211536376953125, 1.211513916015625, 1.2119930419921876, 1.2118671875, 1.2117279052734375, 1.211217041015625, 1.2116253662109375, 1.211358154296875, 1.211610107421875, 1.21132958984375, 1.2115968017578125, 1.2115753173828125, 1.212295166015625, 1.211431884765625, 1.211931640625, 1.21170947265625, 1.21185791015625, 1.2116807861328125, 1.211569091796875, 1.2115169677734374, 1.2117945556640626, 1.211475830078125, 1.2115762939453125, 1.2120863037109375, 1.211937744140625, 1.2119654541015625, 1.2116817626953125, 1.2118035888671874, 1.212337158203125, 1.2117073974609376, 1.2118385009765624, 1.2118773193359376, 1.21191015625, 1.2120863037109375, 1.21194091796875, 1.212105712890625, 1.2121497802734376, 1.2118917236328124, 1.211799560546875, 2.513431640625, 1.2114862060546876, 1.2116961669921875, 1.2118785400390626, 1.2121865234375, 1.2122286376953124, 1.2120013427734375, 1.2121025390625, 1.21226953125, 1.211442138671875, 1.2113643798828124, 1.2111483154296876, 1.2115557861328126, 1.2120247802734374, 1.2112076416015625, 1.2115435791015625, 1.2116663818359374, 1.2110797119140626, 1.2113746337890625, 1.2114166259765624, 1.2112281494140624, 1.2118333740234375, 1.211620361328125, 1.2116275634765625, 1.211684814453125, 1.2115281982421875, 1.212507080078125, 1.2123299560546874, 1.211989990234375, 1.2121068115234375, 1.212232666015625, 1.2115784912109375, 1.21185888671875, 1.211282470703125, 1.211768798828125, 1.2113284912109374, 1.2114544677734376, 1.2116142578125, 1.2119736328125, 1.21160400390625, 1.2118660888671875, 1.211783203125, 1.211763671875, 1.21185888671875, 1.211821044921875, 1.211894775390625, 1.212494873046875, 1.2119080810546874, 1.2119388427734374, 1.2123709716796875, 1.211740234375, 1.2130374755859374, 1.212662841796875, 1.212389404296875, 1.21267919921875, 1.2124569091796875, 1.212505126953125, 1.212078125, 1.2119132080078125, 1.2119111328125, 1.2119111328125, 1.2122039794921875, 1.212859375, 2.513799072265625, 1.2114248046875, 1.211298828125, 1.2112598876953125, 1.2111002197265626, 1.211157470703125, 1.2119388427734374, 1.2117012939453125, 1.211177001953125, 1.21225927734375, 1.2121456298828126, 1.212099609375, 1.2122501220703126, 1.212095458984375, 1.2122705078125, 1.212464111328125, 1.21236474609375, 1.21200439453125, 1.211461669921875, 1.21135107421875, 1.2123013916015626, 1.2120933837890624, 1.2120391845703125, 1.2125921630859375, 1.2116397705078126, 1.21157421875, 1.2111861572265625, 1.21097216796875, 1.2113848876953126, 1.2114390869140625, 1.21153125, 1.211658203125, 1.2115947265625, 1.2115711669921876, 1.2116275634765625, 1.211242431640625, 1.21172998046875, 1.2113121337890624, 1.2114923095703125, 1.211609130859375, 1.2118292236328125, 1.2115927734375, 1.21164794921875, 1.2114503173828124, 1.2116920166015626, 1.2116182861328124, 1.2114801025390625, 1.211564208984375, 1.21261865234375, 1.2118896484375, 1.211937744140625, 1.2124937744140625, 1.2117872314453124, 1.2115753173828125, 1.2119429931640624, 1.211613037109375, 1.212316650390625, 1.211663330078125, 1.2121661376953126, 1.2119859619140625, 1.2116695556640624, 1.2118455810546875, 1.2117862548828124, 2.5132001953125, 1.21116162109375, 1.2113070068359375, 1.2113653564453124, 1.211356201171875, 1.2111585693359375, 1.21143505859375, 1.211683837890625, 1.2116121826171875, 1.2118189697265624, 1.2116448974609375, 1.21178515625, 1.2120279541015626, 1.21154052734375, 1.211658203125, 1.211451416015625, 1.21213134765625, 1.2125296630859375, 1.2122440185546874, 1.212705810546875, 1.2123677978515626, 1.2122071533203125, 1.2123084716796875, 1.2123751220703125, 1.2121220703125, 1.2126556396484376, 1.2121026611328125, 1.21187841796875, 1.212506103515625, 1.21231982421875, 1.212422119140625, 1.2116009521484374, 1.2117115478515625, 1.2120238037109374, 1.2117667236328125, 1.2115660400390624, 1.2122071533203125, 1.211494384765625, 1.211684814453125, 1.21158251953125, 1.211788330078125, 1.2116644287109375, 1.211957275390625, 1.2114073486328125, 1.211895751953125, 1.2115875244140626, 1.2115179443359374, 1.211845703125, 1.2115126953125, 1.2115343017578124, 1.2121968994140624, 1.2115262451171875, 1.211950927734375, 1.2120555419921875, 1.211916259765625, 1.2119879150390624, 1.2116162109375, 1.2115435791015625, 1.212336181640625, 1.2118609619140626, 1.21176171875, 1.211826171875, 1.2119696044921875, 2.513016845703125, 1.2115302734375, 1.21221533203125, 1.211273193359375, 1.211552734375, 1.21154248046875, 1.211378662109375, 1.2118077392578126, 1.2109425048828124, 1.2108125, 1.211199462890625, 1.2112393798828125, 1.2109864501953125, 1.2110008544921875, 1.2113499755859376, 1.21132958984375, 1.211335693359375, 1.2113797607421875, 1.211218994140625, 1.21196240234375, 1.211410400390625, 1.2120770263671874, 1.2116243896484375, 1.2119141845703125, 1.2118671875, 1.211895751953125, 1.2114534912109376, 1.2118670654296875, 1.211398193359375, 1.2112353515625, 1.21159375, 1.2112230224609375, 1.211525146484375, 1.211252685546875, 1.2116796875, 1.21173095703125, 1.2115252685546876, 1.21129150390625, 1.211832275390625, 1.211515869140625, 1.21154052734375, 1.21152001953125, 1.2113541259765626, 1.211916259765625, 1.211999267578125, 1.211557861328125, 1.2116612548828125, 1.2116644287109375, 1.211683837890625, 1.2118814697265625, 1.2116859130859374, 1.21154248046875, 1.212590087890625, 1.211631591796875, 1.21187841796875, 1.2120863037109375, 1.211619384765625, 1.2120166015625, 1.2119869384765625, 1.2120177001953125, 1.2121640625, 1.211704345703125, 1.211842529296875, 2.51329833984375, 1.2113704833984376, 1.2111278076171874, 1.21133056640625, 1.2119141845703125, 1.2111728515625, 1.2112230224609375, 1.2113807373046874, 1.2113182373046876, 1.211325439453125, 1.2114288330078125, 1.211557861328125, 1.211810791015625, 1.2117626953125, 1.21128759765625, 1.2117484130859375, 1.21141650390625, 1.2113756103515625, 1.2116080322265625, 1.211427978515625, 1.2113477783203126, 1.2119552001953124, 1.211189208984375, 1.2114248046875, 1.211610107421875, 1.21135205078125, 1.2119481201171876, 1.2117523193359374, 1.211484130859375, 1.2121180419921875, 1.21152099609375, 1.2116326904296875, 1.211673583984375, 1.2113121337890624, 1.2119869384765625, 1.2122685546875, 1.2123853759765626, 1.2128603515625, 1.2124027099609376, 1.2125235595703125, 1.2127314453125, 1.2127191162109374, 1.212885986328125, 1.2127242431640626, 1.2118333740234375, 1.211953125, 1.2121129150390626, 1.2117801513671875, 1.212015625, 1.211763671875, 1.2117012939453125, 1.2120667724609375, 1.21187939453125, 1.2117197265625, 1.2119981689453125, 1.21179443359375, 1.2120340576171875, 1.212147705078125, 1.21187841796875, 1.2120791015625, 1.21209033203125, 1.211905029296875, 1.2124200439453126, 2.514125732421875, 1.2114227294921875, 1.211292724609375, 1.211658203125, 1.2115589599609375, 1.2113983154296875, 1.2119888916015624, 1.2113704833984376, 1.211610107421875, 1.211758544921875, 1.2114852294921874, 1.21145654296875, 1.211821044921875, 1.2112169189453126, 1.2116602783203125, 1.211989990234375, 1.2113223876953125, 1.2115517578125, 1.2113489990234374, 1.2112762451171875, 1.2118035888671874, 1.211737060546875, 1.2115977783203125, 1.2129249267578126, 1.21261767578125, 1.2122420654296875, 1.211623291015625, 1.2114923095703125, 1.212148681640625, 1.2119019775390625, 1.21175146484375, 1.2118385009765624, 1.2116920166015626, 1.2118294677734376, 1.21165185546875, 1.2111922607421874, 1.2113212890625, 1.2114442138671875, 1.211694091796875, 1.21236474609375, 1.212018798828125, 1.211768798828125, 1.211747314453125, 1.2114288330078125, 1.2113223876953125, 1.2116746826171876, 1.21150048828125, 1.21159375, 1.21177294921875, 1.21147802734375, 1.212080078125, 1.21192138671875, 1.2113746337890625, 1.2114554443359375, 1.2115548095703126, 1.21162646484375, 1.211575439453125, 1.2113038330078125, 1.211806640625, 1.2116695556640624, 1.211400146484375, 1.2115343017578124, 1.2114677734375, 2.5139150390625, 1.2104947509765625, 1.2104796142578125, 1.2110601806640624, 1.210818603515625, 1.21080419921875, 1.2113817138671874, 1.2107110595703126, 1.2115599365234375, 1.210962890625, 1.21103466796875, 1.2109844970703125, 1.2109957275390626, 1.210639404296875, 1.210967041015625, 1.210735595703125, 1.211298828125, 1.2114892578125, 1.210882080078125, 1.2117001953125, 1.211292724609375, 1.2112998046875, 1.211736083984375, 1.2114698486328126, 1.2116868896484374, 1.212020751953125, 1.2116572265625, 1.2118538818359375, 1.211252685546875, 1.2111072998046875, 1.211619384765625, 1.2111688232421876, 1.211368408203125, 1.2118763427734376, 1.2112978515625, 1.211556884765625, 1.211440185546875, 1.2111011962890625, 1.21163671875, 1.2113223876953125, 1.2111298828125, 1.2116162109375, 1.2110653076171876, 1.2111217041015625, 1.211494384765625, 1.211451416015625, 1.2114759521484375, 1.21181689453125, 1.2115538330078126, 1.2116080322265625, 1.211831298828125, 1.211515869140625, 1.21173193359375, 1.211282470703125, 1.2113018798828126, 1.21145654296875, 1.2112281494140624, 1.2112547607421875, 1.2116612548828125, 1.21142578125, 1.21210986328125, 1.2116920166015626, 1.2119910888671874, 2.51358935546875, 1.21063427734375, 1.211430908203125, 1.210544189453125, 1.2111871337890625, 1.21103564453125, 1.2112589111328125, 1.2109700927734375, 1.2114503173828124, 1.2111072998046875, 1.21147802734375, 1.210945556640625, 1.2111922607421874, 1.211684814453125, 1.2114554443359375, 1.2112486572265626, 1.2119346923828125, 1.2112281494140624, 1.2114483642578124, 1.2110079345703124, 1.2108524169921875, 1.2109935302734376, 1.2109833984375, 1.2113858642578126, 1.2111492919921876, 1.210892333984375, 1.2111124267578126, 1.2112025146484375, 1.2109700927734375, 1.211509765625, 1.21159375, 1.2114063720703125, 1.211040771484375, 1.21109814453125, 1.210799072265625, 1.210892333984375, 1.210966064453125, 1.210841064453125, 1.2113079833984375, 1.21124560546875, 1.2112608642578124, 1.2112496337890626, 1.2110306396484376, 1.2115035400390626, 1.211241455078125, 1.2117279052734375, 1.2114945068359375, 1.2111419677734374, 1.2113212890625, 1.211431884765625, 1.2111728515625, 1.2113961181640625, 1.211167724609375, 1.210883056640625, 1.21160498046875, 1.211494384765625, 1.2115814208984375, 1.2117698974609374, 1.2115025634765626, 1.2116275634765625, 1.2118814697265625, 1.211230224609375, 1.21170947265625]",tokens/s,0.8128494394247809,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-MoE-A2.7B,Qwen/Qwen1.5-MoE-A2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3640, in from_pretrained + hf_quantizer.preprocess_model( + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 182, in preprocess_model + return self._process_model_before_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 85, in _process_model_before_weight_loading + model, has_been_replaced = replace_with_awq_linear( + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 178, in replace_with_awq_linear + _, has_been_replaced = replace_with_awq_linear( + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 178, in replace_with_awq_linear + _, has_been_replaced = replace_with_awq_linear( + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 178, in replace_with_awq_linear + _, has_been_replaced = replace_with_awq_linear( + [Previous line repeated 1 more time] + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 165, in replace_with_awq_linear + model._modules[name] = target_cls( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemm.py"", line 102, in __init__ + assert out_features % (32 // self.w_bit) == 0 +AssertionError + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-1.3b,EleutherAI/pythia-1.3b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,,,MB,1688.977408,2250.768384,0.0,1604.32128,1463.693312,s,10,1.2614907455444335,0.12614907455444335,0.001156305474696835,0.12596041870117186,0.12672372131347656,0.12802366485595704,0.12906361968994143,"[0.1293236083984375, 0.12533795166015624, 0.12527750396728515, 0.1264024658203125, 0.12550204467773438, 0.12507218933105468, 0.1257669143676758, 0.12615392303466796, 0.12621929931640624, 0.12643484497070312]",tokens/s,2029.3450499275416,kWh,1.4795928945144019e-06,8.107423014348568e-07,6.482848936275459e-06,8.773184132224718e-06,tokens/kWh,29179827.545131337,MB,1688.977408,2250.768384,0.0,1604.32128,1560.975872,s,10,72.8006865234375,7.280068652343751,0.00707557452467299,7.2794865722656255,7.288760400390625,7.290637524414063,7.292139223632812,"[7.2925146484375, 7.2867998046875, 7.28834326171875, 7.272900390625, 7.28120458984375, 7.28177783203125, 7.2702080078125, 7.2777685546875, 7.277201171875, 7.27196826171875]",tokens/s,8.653764546535935,kWh,8.591771873335045e-05,4.70891264273655e-05,0.0003703246191761246,0.0005033314643368406,tokens/kWh,125166.02768516574,,s,629,73.8150656280517,0.1173530455135958,0.015000349665490152,0.11545600128173829,0.11595755310058595,0.11627356262207031,0.24130564147949218,"[0.1175910415649414, 0.11764224243164062, 0.11673804473876953, 0.11585126495361328, 0.115281982421875, 0.11545388793945313, 0.11590860748291015, 0.11528498840332031, 0.11539968109130859, 0.11538841247558594, 0.11548159790039063, 0.11523072052001954, 0.11521842956542969, 0.11532492828369141, 0.11577958679199218, 0.11560755157470703, 0.1152696304321289, 0.11566079711914062, 0.1159925765991211, 0.11540889739990234, 0.11546623992919922, 0.11522150421142578, 0.11538944244384766, 0.1153617935180664, 0.11544371032714844, 0.1166714859008789, 0.11574886322021484, 0.1154672622680664, 0.11605299377441407, 0.11584102630615234, 0.1156485137939453, 0.11544882965087891, 0.11554099273681641, 0.115557373046875, 0.11551026916503906, 0.1155788803100586, 0.115725341796875, 0.11555939483642579, 0.11551129913330078, 0.11544371032714844, 0.11543244934082031, 0.11539968109130859, 0.11540684509277344, 0.11523072052001954, 0.11638988494873047, 0.11598438262939453, 0.11537510681152344, 0.11538534545898438, 0.11558604431152343, 0.11559219360351562, 0.11591474914550781, 0.11588198089599609, 0.1159362564086914, 0.11570175933837891, 0.11595673370361329, 0.11578368377685547, 0.11646463775634766, 0.11617485046386719, 0.11598336029052735, 0.11553279876708984, 0.11578470611572265, 0.11579801940917969, 0.24151040649414063, 0.11573862457275391, 0.11605094146728516, 0.11689881896972656, 0.1161523208618164, 0.1156648941040039, 0.11554099273681641, 0.11535564422607422, 0.11528498840332031, 0.11527065277099609, 0.11566387176513672, 0.11533516693115234, 0.11576012420654297, 0.11564749145507812, 0.1155051498413086, 0.11542221069335938, 0.11543961334228516, 0.11525529479980469, 0.11523993682861328, 0.11541913604736329, 0.1155225601196289, 0.1157734375, 0.11563929748535157, 0.11553587341308594, 0.11531571197509766, 0.115378173828125, 0.11524201965332032, 0.11551340484619141, 0.11532073974609375, 0.11577961730957032, 0.11592704010009766, 0.11556451416015626, 0.1154549789428711, 0.11527680206298828, 0.11526656341552734, 0.11538739013671875, 0.11530854034423828, 0.11538432312011719, 0.11546214294433593, 0.11541401672363281, 0.11528192138671875, 0.11616563415527344, 0.11571199798583984, 0.11578880310058594, 0.11551747131347656, 0.11540067291259766, 0.11685273742675781, 0.11557068634033203, 0.11726950073242187, 0.1159393310546875, 0.11543148803710937, 0.11552044677734374, 0.1161891860961914, 0.11548876953125, 0.11572940826416016, 0.1157396469116211, 0.11657421112060547, 0.11594854736328125, 0.11598851013183593, 0.11593417358398438, 0.11551846313476563, 0.11548365020751954, 0.11548569488525391, 0.2411530303955078, 0.11556454467773437, 0.11577139282226563, 0.11626496124267578, 0.11560652923583985, 0.11565161895751953, 0.11566384124755859, 0.11581747436523437, 0.1154672622680664, 0.11527474975585937, 0.11536998748779297, 0.11562598419189453, 0.11528089904785156, 0.11566902160644531, 0.11589116668701172, 0.11617894744873047, 0.1157201919555664, 0.11589734649658204, 0.11571302032470702, 0.11588607788085938, 0.11542323303222657, 0.11556966400146484, 0.1156833267211914, 0.11643392181396485, 0.11568434906005859, 0.11659468841552735, 0.11596083068847657, 0.11616665649414062, 0.11567922973632813, 0.11553794860839844, 0.11543036651611328, 0.11538636779785157, 0.11580620574951171, 0.11554918670654298, 0.11579296112060547, 0.1156412811279297, 0.11582361602783203, 0.11581132507324218, 0.11555328369140624, 0.11571507263183593, 0.11545600128173829, 0.11533926391601562, 0.1156147232055664, 0.1153966064453125, 0.1154119644165039, 0.11549286651611328, 0.11582566070556641, 0.11649638366699219, 0.1159516830444336, 0.11595462036132813, 0.11600281524658203, 0.11591474914550781, 0.11610931396484375, 0.11565261077880859, 0.1153290252685547, 0.11547238159179687, 0.1153290252685547, 0.115346435546875, 0.11534438323974609, 0.11529523468017579, 0.115378173828125, 0.11543551635742187, 0.11540377807617187, 0.24315084838867188, 0.11575193786621094, 0.11506790161132813, 0.11524915313720703, 0.1155962905883789, 0.11531366729736328, 0.11553177642822265, 0.11558297729492187, 0.11519385528564453, 0.11537305450439453, 0.11545811462402343, 0.11599967956542968, 0.11572838592529297, 0.1153966064453125, 0.11536077117919921, 0.11558092498779297, 0.1153617935180664, 0.1152911376953125, 0.11529011535644532, 0.11535564422607422, 0.11525939178466797, 0.1152143325805664, 0.11515494537353516, 0.11549696350097656, 0.11525017547607422, 0.11560550689697266, 0.11576217651367188, 0.11544268798828125, 0.11537203216552734, 0.11532288360595704, 0.11530239868164062, 0.11551334381103516, 0.115162109375, 0.11524813079833984, 0.1151436767578125, 0.11530035400390624, 0.11520614624023437, 0.11535871887207032, 0.11520716857910156, 0.11560345458984375, 0.11539968109130859, 0.11531676483154298, 0.11529417419433594, 0.1153966064453125, 0.11623017883300782, 0.11542422485351563, 0.11526451110839844, 0.11549388885498046, 0.11539762878417968, 0.11577855682373046, 0.1153955841064453, 0.11553279876708984, 0.11524403381347656, 0.11579084777832031, 0.11539968109130859, 0.11569152069091797, 0.11539353942871093, 0.11553997039794922, 0.11545600128173829, 0.11532492828369141, 0.1152696304321289, 0.11538841247558594, 0.1153433609008789, 0.241364990234375, 0.11547853088378907, 0.11533721923828125, 0.11539049530029297, 0.11531874847412109, 0.11551129913330078, 0.1152542724609375, 0.11560758209228515, 0.11542422485351563, 0.1155389404296875, 0.11542118072509766, 0.11545804595947265, 0.1156178207397461, 0.115463134765625, 0.11520716857910156, 0.11564339447021485, 0.11541913604736329, 0.11587276458740234, 0.1153955841064453, 0.11544882965087891, 0.11543654632568359, 0.11536793518066406, 0.11598540496826172, 0.11562393951416015, 0.11578470611572265, 0.11597004699707031, 0.11561062622070313, 0.11569459533691406, 0.11571302032470702, 0.11590144348144531, 0.11557068634033203, 0.11549900817871094, 0.11544678497314453, 0.11544268798828125, 0.11540889739990234, 0.11547443389892578, 0.11561574554443359, 0.11543449401855468, 0.11550624084472656, 0.11566381072998047, 0.11555225372314454, 0.11549286651611328, 0.11550822448730469, 0.11547955322265625, 0.1153064956665039, 0.11546623992919922, 0.11551744079589844, 0.11544371032714844, 0.11622195434570312, 0.11589017486572266, 0.11552665710449218, 0.1153986587524414, 0.11532390594482422, 0.11578470611572265, 0.11576934051513672, 0.11560550689697266, 0.11554713439941407, 0.11664281463623047, 0.11574784088134765, 0.11553997039794922, 0.11533824157714843, 0.11548569488525391, 0.11580518341064452, 0.2409891815185547, 0.11542425537109376, 0.11530342102050781, 0.11544166564941406, 0.115346435546875, 0.11540684509277344, 0.11648614501953125, 0.1165475845336914, 0.11569561767578125, 0.11536281585693359, 0.11529318237304688, 0.11541401672363281, 0.11542221069335938, 0.11558809661865234, 0.11539456176757812, 0.11545702362060548, 0.11533209228515626, 0.11550822448730469, 0.11541913604736329, 0.11535667419433594, 0.11532185363769532, 0.11544985961914063, 0.11528300476074219, 0.11542829132080078, 0.11535564422607422, 0.11552665710449218, 0.11533004760742188, 0.11546419525146484, 0.11538022613525391, 0.11521539306640625, 0.1153545913696289, 0.11576934051513672, 0.11584614562988281, 0.11546214294433593, 0.11551232147216797, 0.11547647857666016, 0.11558092498779297, 0.1154796142578125, 0.11671033477783203, 0.11560243225097656, 0.11540480041503906, 0.11559120178222657, 0.11558499145507813, 0.11560243225097656, 0.11611443328857422, 0.11554815673828125, 0.11593113708496093, 0.11560857391357422, 0.11550003051757812, 0.1153433609008789, 0.11553488159179688, 0.11636220550537109, 0.11614924621582032, 0.11589631652832032, 0.11544371032714844, 0.11543961334228516, 0.11590144348144531, 0.11554303741455078, 0.11543961334228516, 0.11548467254638672, 0.1157918701171875, 0.11545804595947265, 0.11548159790039063, 0.24169778442382814, 0.11536077117919921, 0.11515801239013672, 0.1152573471069336, 0.11527884674072265, 0.11527986907958984, 0.115346435546875, 0.11558502197265624, 0.1152471694946289, 0.11550918579101563, 0.11519078063964844, 0.11538329315185547, 0.11518669128417969, 0.11536281585693359, 0.11518163299560547, 0.11539347076416015, 0.11518678283691407, 0.11522755432128906, 0.11527577972412109, 0.11537920379638672, 0.11528710174560547, 0.11573241424560547, 0.11531775665283203, 0.11599462127685548, 0.11536895751953125, 0.11559219360351562, 0.1152174072265625, 0.11544473266601563, 0.11515187072753906, 0.11589734649658204, 0.11575091552734375, 0.11540377807617187, 0.11524508666992188, 0.11525116729736327, 0.1152573471069336, 0.11555532836914062, 0.1153259506225586, 0.1153617935180664, 0.1152542724609375, 0.11547135925292969, 0.11532083129882813, 0.11537305450439453, 0.11539968109130859, 0.11537612915039062, 0.11517235565185546, 0.11529216003417969, 0.11535257720947266, 0.11539046478271485, 0.11532390594482422, 0.11542940521240234, 0.11545801544189453, 0.115557373046875, 0.11530239868164062, 0.11545193481445312, 0.11529827117919922, 0.11565363311767578, 0.11541299438476563, 0.11608678436279297, 0.11551641845703126, 0.11544064331054688, 0.11531980895996094, 0.11546214294433593, 0.11526348876953126, 0.2421370849609375, 0.11553485107421875, 0.11565363311767578, 0.11698073577880859, 0.11526451110839844, 0.11523788452148437, 0.11507615661621094, 0.11531769561767578, 0.11525939178466797, 0.11527884674072265, 0.11663155364990234, 0.11554815673828125, 0.11534745788574219, 0.11536281585693359, 0.11533106994628907, 0.11543449401855468, 0.11581439971923828, 0.11606221008300781, 0.11575091552734375, 0.11566182708740234, 0.1154867172241211, 0.1154119644165039, 0.11538432312011719, 0.11553075408935547, 0.11544985961914063, 0.1154119644165039, 0.11534540557861328, 0.11543142700195312, 0.11537612915039062, 0.11558502197265624, 0.11556352233886719, 0.1155072021484375, 0.11545394897460938, 0.11533824157714843, 0.11553485107421875, 0.11541094207763672, 0.1153597412109375, 0.11537407684326172, 0.11529216003417969, 0.11547955322265625, 0.11540787506103516, 0.11534950256347656, 0.11540275573730469, 0.11537305450439453, 0.11569664001464844, 0.11560038757324219, 0.11557785797119141, 0.11525635528564453, 0.115210205078125, 0.11607859039306641, 0.11538022613525391, 0.11546419525146484, 0.11593727874755859, 0.11551436614990235, 0.11531263732910156, 0.11544576263427735, 0.11548569488525391, 0.11550924682617188, 0.11583999633789062, 0.11551641845703126, 0.11535871887207032, 0.11538739013671875, 0.11543142700195312, 0.24210124206542968, 0.11537920379638672, 0.11549593353271484, 0.11537612915039062, 0.11512525177001953, 0.11541709136962891, 0.1153280029296875, 0.11576729583740235, 0.115378173828125, 0.11518669128417969, 0.11519593811035156, 0.115463134765625, 0.11528498840332031, 0.11549491119384765, 0.11540172576904296, 0.11594445037841797, 0.11559526062011719, 0.116279296875, 0.11579698944091797, 0.11580210876464844, 0.11537203216552734, 0.11540991973876953, 0.11529318237304688, 0.11539456176757812, 0.1153597412109375, 0.11614412689208985, 0.11539968109130859, 0.11535155487060547, 0.11532492828369141, 0.11529523468017579, 0.11584108734130859, 0.1154241943359375, 0.11634585571289062, 0.11548159790039063, 0.11571302032470702, 0.1153219223022461, 0.1154579849243164, 0.11549388885498046, 0.1154303970336914, 0.11558297729492187, 0.11536383819580077, 0.11582975769042969, 0.11551436614990235, 0.11530035400390624, 0.11534233856201172, 0.11510886383056641, 0.11527168273925781, 0.11539968109130859, 0.11526860809326171, 0.1152573471069336, 0.11550822448730469, 0.11615952301025391, 0.1159669418334961, 0.11542835235595703, 0.11572223663330078, 0.11554918670654298, 0.11545906829833984, 0.11568742370605468, 0.1153812484741211, 0.11539762878417968, 0.11535155487060547, 0.1154303970336914, 0.11551129913330078, 0.24231832885742188, 0.11538432312011719, 0.1153259506225586, 0.11540480041503906, 0.11546419525146484, 0.11534745788574219, 0.11531878662109375, 0.11547138977050782, 0.115283935546875, 0.11556147003173828, 0.11574169921875, 0.11635916900634766, 0.11543142700195312, 0.11534540557861328, 0.11524813079833984, 0.1152911376953125, 0.11551436614990235, 0.11536589050292968, 0.11532083129882813, 0.11543247985839844, 0.11533615875244141, 0.11537407684326172, 0.11525631713867188, 0.11534438323974609, 0.11512217712402344, 0.11535769653320313, 0.11603763580322266, 0.11543449401855468, 0.1153986587524414, 0.11541506958007812, 0.11541094207763672, 0.11551331329345703, 0.11517440032958984, 0.11537715148925781, 0.11538534545898438, 0.11531263732910156, 0.11526758575439452, 0.11542835235595703, 0.11535052490234375, 0.11555430603027343, 0.11530035400390624, 0.11539046478271485, 0.11545398712158203, 0.1154815673828125, 0.11535667419433594, 0.11539250946044922, 0.1153812484741211, 0.1153986587524414, 0.1155051498413086, 0.11556665802001953, 0.11569554901123047, 0.11560447692871094, 0.11522866821289063, 0.11546419525146484, 0.11527577972412109, 0.11578880310058594, 0.11543142700195312, 0.1154119644165039, 0.115346435546875, 0.11526553344726563, 0.11522560119628907, 0.11540480041503906, 0.11531673431396484]",tokens/s,8.5212956819612,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-2.7b,EleutherAI/pythia-2.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,2265.919488,3330.801664,0.0,2684.35456,2447.59552,s,10,2.4030111236572265,0.24030111236572266,0.0012440053970201484,0.24010930633544922,0.24119976043701172,0.24230707931518555,0.24319293441772463,"[0.2434143981933594, 0.2409536895751953, 0.23907139587402343, 0.23946762084960938, 0.2396122589111328, 0.23890797424316407, 0.23970700073242188, 0.2405301055908203, 0.24051161193847656, 0.24083506774902344]",tokens/s,1065.330066430923,kWh,2.8237744456245787e-06,1.5472998550649208e-06,1.2583984934904673e-05,1.6955059235594172e-05,tokens/kWh,15098738.166751605,MB,2265.919488,3330.801664,0.0,2684.35456,2597.681664,s,10,139.56764257812497,13.9567642578125,0.0037662168729373563,13.95505322265625,13.962588671875,13.96323818359375,13.963757792968751,"[13.95483984375, 13.95481640625, 13.9549306640625, 13.95517578125, 13.9599765625, 13.9638876953125, 13.9624443359375, 13.953060546875, 13.952154296875, 13.9563564453125]",tokens/s,4.513940254076789,kWh,0.00016469605360239274,9.026671937368518e-05,0.0007221266028330957,0.0009770893758091737,tokens/kWh,64477.21320050864,,s,629,141.50236668396,0.22496401698562793,0.028578568615475336,0.22146662902832032,0.22197882690429688,0.222200830078125,0.46145660278320316,"[0.222501953125, 0.22133958435058593, 0.22126080322265626, 0.22115533447265626, 0.22122291564941407, 0.22115737915039063, 0.22111334228515625, 0.2211881561279297, 0.22123001098632813, 0.22108671569824218, 0.22202674865722657, 0.2213621826171875, 0.22130892944335936, 0.221412353515625, 0.22178816223144532, 0.22125669860839844, 0.22154550170898438, 0.22125155639648436, 0.22184141540527344, 0.22194586181640624, 0.22131817626953126, 0.22113481140136718, 0.22139187622070314, 0.22203596496582031, 0.22120550537109376, 0.22182707214355468, 0.22148197937011718, 0.22121574401855468, 0.22116249084472656, 0.22113999938964843, 0.22153315734863283, 0.22137344360351563, 0.22128536987304687, 0.22147686767578126, 0.22139497375488282, 0.2215167694091797, 0.22115020751953124, 0.22119526672363282, 0.2216222686767578, 0.22131507873535156, 0.2215731201171875, 0.22197042846679688, 0.2214871063232422, 0.22137344360351563, 0.22149530029296874, 0.2211829833984375, 0.22195916748046876, 0.22144102478027344, 0.22115122985839844, 0.22134988403320313, 0.2216417236328125, 0.22153318786621093, 0.2213396453857422, 0.22140824890136718, 0.2213949432373047, 0.22160794067382814, 0.22140109252929688, 0.22149017333984375, 0.22158233642578126, 0.2214686737060547, 0.2215290832519531, 0.2215116729736328, 0.4632596435546875, 0.22171238708496094, 0.221233154296875, 0.22143283081054688, 0.2215679931640625, 0.22207183837890626, 0.22199087524414063, 0.2218137664794922, 0.22159461975097655, 0.2211604766845703, 0.22112969970703125, 0.22134375, 0.22146156311035156, 0.2211706237792969, 0.2218250274658203, 0.2215004119873047, 0.2212454376220703, 0.22122291564941407, 0.22121881103515625, 0.22129049682617188, 0.22144000244140624, 0.22122496032714845, 0.22121574401855468, 0.22197247314453125, 0.2214799346923828, 0.22136012268066407, 0.22134066772460936, 0.22136422729492186, 0.22131097412109374, 0.22127923583984374, 0.22128128051757812, 0.22142156982421876, 0.22127308654785155, 0.22122598266601562, 0.22121676635742188, 0.22201344299316406, 0.2224384002685547, 0.22226841735839845, 0.22168576049804686, 0.22131507873535156, 0.2211666259765625, 0.22145021057128905, 0.2213509063720703, 0.2210908203125, 0.2211420135498047, 0.2215116729736328, 0.22274458312988282, 0.22148300170898438, 0.2212454376220703, 0.22115327453613282, 0.22183013916015626, 0.22141746520996095, 0.22183628845214845, 0.2215557098388672, 0.22127104187011717, 0.2220482635498047, 0.2213693389892578, 0.22153421020507813, 0.22129356384277343, 0.22213938903808594, 0.22136114501953125, 0.22106112670898437, 0.22137957763671876, 0.4608409729003906, 0.22114405822753908, 0.2210918426513672, 0.22119833374023437, 0.22196018981933593, 0.2215413818359375, 0.22137139892578125, 0.2214246368408203, 0.22122802734375, 0.22138983154296876, 0.2213478698730469, 0.2211327667236328, 0.22113591003417968, 0.2211645050048828, 0.22161613464355467, 0.22142771911621092, 0.22118502807617188, 0.22131715393066406, 0.22171746826171876, 0.22127410888671875, 0.22118911743164063, 0.22119740295410156, 0.22118185424804687, 0.22219775390625, 0.22114816284179686, 0.22126797485351563, 0.22138265991210937, 0.22202163696289062, 0.221623291015625, 0.22125567626953124, 0.2214297637939453, 0.22170930480957032, 0.22169293212890626, 0.22206874084472655, 0.22159257507324218, 0.2214799346923828, 0.22199090576171876, 0.22157720947265624, 0.22174208068847656, 0.22161613464355467, 0.2212843475341797, 0.22151271057128907, 0.22158233642578126, 0.22143283081054688, 0.22131404113769532, 0.22166323852539063, 0.22163046264648437, 0.22240870666503906, 0.22161407470703126, 0.22181785583496094, 0.22200831604003907, 0.22142874145507813, 0.22144102478027344, 0.22163661193847656, 0.22149530029296874, 0.22142668151855469, 0.22134169006347656, 0.2213683166503906, 0.2215076141357422, 0.2218208923339844, 0.221412353515625, 0.2214072265625, 0.22158233642578126, 0.46169601440429686, 0.2212136993408203, 0.22105702209472655, 0.2211266632080078, 0.2211031036376953, 0.22123622131347656, 0.22111846923828124, 0.22110105895996093, 0.22219468688964844, 0.22128536987304687, 0.2216058807373047, 0.22124237060546875, 0.22226022338867188, 0.22136323547363282, 0.2218741455078125, 0.22145741271972658, 0.22111538696289063, 0.22111231994628905, 0.22129766845703125, 0.22105702209472655, 0.22113792419433595, 0.22142874145507813, 0.2212351989746094, 0.22122496032714845, 0.22117068481445312, 0.22113279724121093, 0.22115122985839844, 0.22112973022460938, 0.2212024383544922, 0.2211102752685547, 0.2212956085205078, 0.22164582824707033, 0.2216058807373047, 0.22137548828125, 0.22140524291992186, 0.22150752258300782, 0.22146560668945312, 0.22119526672363282, 0.22161715698242188, 0.22124032592773438, 0.22137651062011718, 0.221770751953125, 0.22136524963378906, 0.22113690185546875, 0.22289407348632811, 0.22195814514160156, 0.22175027465820313, 0.22173286437988282, 0.2216048583984375, 0.221559814453125, 0.22163661193847656, 0.22160383605957032, 0.22145741271972658, 0.22292991638183593, 0.22169804382324218, 0.22141644287109374, 0.22166732788085938, 0.22217625427246093, 0.2222335968017578, 0.22209843444824218, 0.22188134765625, 0.2213939208984375, 0.22153727722167968, 0.46185061645507813, 0.2211788787841797, 0.22111436462402342, 0.22129049682617188, 0.22150553894042968, 0.22128128051757812, 0.2212024383544922, 0.22144000244140624, 0.22146969604492187, 0.22136114501953125, 0.22112562561035157, 0.2214481964111328, 0.2222335968017578, 0.22209536743164063, 0.22137548828125, 0.2215034942626953, 0.22125567626953124, 0.22210150146484375, 0.22135398864746095, 0.22115122985839844, 0.22146560668945312, 0.22135501098632812, 0.2211973114013672, 0.2217144317626953, 0.22152809143066407, 0.22151266479492188, 0.2218260498046875, 0.22138777160644532, 0.22115327453613282, 0.22133042907714845, 0.2219438018798828, 0.22181068420410155, 0.22151679992675782, 0.22163967895507813, 0.22198374938964843, 0.222202880859375, 0.2216089630126953, 0.22158848571777343, 0.22197964477539062, 0.2219304962158203, 0.2216058807373047, 0.22188236999511718, 0.2217144317626953, 0.2216990966796875, 0.22185162353515625, 0.22161509704589843, 0.22144825744628907, 0.2215761260986328, 0.22152294921875, 0.22147789001464843, 0.22122189331054687, 0.22149017333984375, 0.22142054748535156, 0.2215905303955078, 0.2212833251953125, 0.22145535278320314, 0.2214256591796875, 0.22370611572265625, 0.22159666442871093, 0.22157209777832032, 0.22164889526367187, 0.22127622985839843, 0.2215331268310547, 0.4607232055664062, 0.2214246368408203, 0.22150758361816406, 0.22157212829589842, 0.22139695739746093, 0.22143283081054688, 0.22193766784667968, 0.22149017333984375, 0.22146456909179688, 0.22140824890136718, 0.22149533081054687, 0.22149014282226562, 0.22156288146972655, 0.22154342651367187, 0.22170623779296875, 0.2220062713623047, 0.22157107543945312, 0.22163865661621093, 0.22165196228027345, 0.22187519836425781, 0.22198477172851563, 0.22197862243652344, 0.22181272888183592, 0.22163456726074218, 0.22144921875, 0.2217902069091797, 0.22170623779296875, 0.22131404113769532, 0.22122291564941407, 0.22189158630371095, 0.22152499389648436, 0.2215362548828125, 0.22123417663574219, 0.2213365783691406, 0.2216407012939453, 0.22166015625, 0.22149221801757812, 0.22126182556152343, 0.22392626953125, 0.2219438018798828, 0.22167347717285157, 0.22165606689453124, 0.22170930480957032, 0.22157626342773437, 0.2213846435546875, 0.22152499389648436, 0.2214307861328125, 0.2218014678955078, 0.22188134765625, 0.22133042907714845, 0.22165298461914062, 0.22161407470703126, 0.22176870727539064, 0.22166835021972656, 0.22149017333984375, 0.22150656127929688, 0.22176768493652343, 0.22180863952636717, 0.22148101806640624, 0.221297607421875, 0.22151373291015625, 0.222308349609375, 0.22171034240722656, 0.46220184326171876, 0.22253363037109375, 0.22171034240722656, 0.22214041137695312, 0.22114816284179686, 0.2212351989746094, 0.22153932189941405, 0.22129664611816408, 0.2211778564453125, 0.22154853820800782, 0.22134176635742187, 0.22164781188964844, 0.22147378540039062, 0.22127615356445313, 0.2212843475341797, 0.22165913391113282, 0.2215885467529297, 0.22137234497070313, 0.22157516479492187, 0.22207180786132813, 0.22170623779296875, 0.22145126342773438, 0.22118502807617188, 0.22266777038574218, 0.22152601623535156, 0.2213744659423828, 0.22128947448730468, 0.22164991760253908, 0.2218076171875, 0.22161920166015625, 0.22146969604492187, 0.22127513122558592, 0.22119424438476562, 0.22168269348144531, 0.22124441528320313, 0.22097100830078126, 0.221085693359375, 0.22113381958007813, 0.22170008850097656, 0.22122700500488282, 0.22127923583984374, 0.22145330810546876, 0.22255923461914062, 0.22140007019042968, 0.22144613647460937, 0.22161509704589843, 0.222023681640625, 0.22214041137695312, 0.22231552124023438, 0.2217574462890625, 0.22175949096679687, 0.22195610046386718, 0.22176666259765626, 0.221739013671875, 0.22195001220703126, 0.22145529174804687, 0.2223953857421875, 0.22163661193847656, 0.22160281372070312, 0.22152294921875, 0.22183322143554687, 0.2217840576171875, 0.2217410583496094, 0.46331597900390625, 0.22162431335449218, 0.22123930358886718, 0.2212843475341797, 0.2211409912109375, 0.22146662902832032, 0.22136729431152344, 0.22115020751953124, 0.22153216552734376, 0.22155059814453126, 0.22213427734375, 0.2212034606933594, 0.221412353515625, 0.22128640747070313, 0.22130995178222657, 0.22144102478027344, 0.22194688415527344, 0.22158131408691406, 0.22149632263183594, 0.22149427795410156, 0.2216222686767578, 0.22129664611816408, 0.22110617065429689, 0.22211993408203126, 0.2216058807373047, 0.22146456909179688, 0.22136012268066407, 0.22197760009765624, 0.2215905303955078, 0.22179327392578124, 0.2214993896484375, 0.22128230285644532, 0.2213017578125, 0.22127410888671875, 0.22146456909179688, 0.22121778869628905, 0.22168576049804686, 0.22120037841796875, 0.22118406677246094, 0.22135084533691407, 0.22131301879882812, 0.22129869079589845, 0.22116659545898437, 0.2210508728027344, 0.2215045166015625, 0.22192332458496095, 0.2215854034423828, 0.22149325561523436, 0.22160076904296874, 0.2215669708251953, 0.22166323852539063, 0.22128536987304687, 0.22132333374023438, 0.22139077758789064, 0.22124549865722656, 0.22128941345214845, 0.221665283203125, 0.22161613464355467, 0.22151577758789062, 0.22136524963378906, 0.2216407012939453, 0.22126693725585939, 0.22150860595703126, 0.4624425048828125, 0.22111744689941407, 0.22104473876953126, 0.2211829833984375, 0.22187826538085936, 0.22165196228027345, 0.2212290496826172, 0.22156083679199218, 0.22200729370117187, 0.22158950805664063, 0.22172671508789063, 0.221380615234375, 0.221306884765625, 0.22107449340820312, 0.22128428649902343, 0.22132121276855468, 0.221306884765625, 0.22197042846679688, 0.22143589782714843, 0.2213027801513672, 0.22142771911621092, 0.22136524963378906, 0.22115225219726561, 0.22165811157226561, 0.22134988403320313, 0.22114303588867187, 0.22162124633789063, 0.22177484130859376, 0.22148812866210937, 0.2213570556640625, 0.22117170715332032, 0.22172569274902343, 0.221444091796875, 0.22128640747070313, 0.22114816284179686, 0.22123423767089845, 0.22163040161132813, 0.22124134826660155, 0.22113900756835939, 0.22141433715820313, 0.22133351135253906, 0.22119833374023437, 0.22116659545898437, 0.2211727294921875, 0.2214256591796875, 0.22142771911621092, 0.22107034301757814, 0.2211973114013672, 0.22134579467773438, 0.22134988403320313, 0.2222950439453125, 0.22165298461914062, 0.22142361450195314, 0.22213938903808594, 0.22171749877929686, 0.2223206329345703, 0.2216816711425781, 0.22124032592773438, 0.22122700500488282, 0.22170623779296875, 0.22179840087890626, 0.22177587890625, 0.2214297637939453, 0.4629862365722656, 0.22168269348144531, 0.22135910034179687, 0.22179840087890626, 0.22143487548828125, 0.2214686737060547, 0.22133042907714845, 0.22189260864257812, 0.22169088745117188, 0.22153421020507813, 0.22115327453613282, 0.22121165466308593, 0.22176153564453124, 0.221380615234375, 0.22127308654785155, 0.22127206420898438, 0.2213959655761719, 0.22139085388183594, 0.22160592651367186, 0.2212945556640625, 0.22128128051757812, 0.2216611785888672, 0.22154751586914062, 0.22151373291015625, 0.2215004119873047, 0.22142361450195314, 0.22282957458496094, 0.22140313720703125, 0.22110617065429689, 0.22118502807617188, 0.22146456909179688, 0.2219325408935547, 0.22170828247070312, 0.22170008850097656, 0.22203187561035156, 0.22157005310058595, 0.22241897583007814, 0.22136521911621093, 0.22132940673828125, 0.22154444885253907, 0.22126284790039064, 0.22154444885253907, 0.22126080322265626, 0.22116966247558595, 0.22158335876464844, 0.22131199645996094, 0.2215188446044922, 0.22126797485351563, 0.22128128051757812, 0.22140518188476563, 0.22145126342773438, 0.221559814453125, 0.22204620361328126, 0.22181068420410155, 0.22146456909179688, 0.22160794067382814, 0.22161715698242188, 0.22150553894042968, 0.22128640747070313, 0.22156185913085938, 0.22152806091308594, 0.22130995178222657, 0.2214297637939453]",tokens/s,4.445155333725598,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-14B,Qwen/Qwen1.5-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,3951.190016,12732.33408,0.0,12085.886976,11337.496064,s,10,10.983638916015623,1.0983638916015628,0.002145812483257245,1.0982380981445312,1.1009110717773436,1.1018160705566407,1.1025400695800782,"[1.1027210693359375, 1.1007099609375, 1.0960345458984375, 1.0957255859375, 1.0965640869140625, 1.096655029296875, 1.0976143798828124, 1.09886181640625, 1.0996507568359375, 1.0991016845703125]",tokens/s,233.07394021003134,kWh,1.2942184060811996e-05,7.0910633564744784e-06,6.339727294000231e-05,8.343052035728877e-05,tokens/kWh,3068421.4709879244,MB,3951.190016,12732.33408,0.0,12085.886976,11686.800384,s,10,637.80165234375,63.780165234375,0.009100497531913991,63.777451171875,63.784855859375,63.7953341796875,63.8037168359375,"[63.775953125, 63.7747265625, 63.78252734375, 63.77445703125, 63.77937109375, 63.7818203125, 63.8058125, 63.7783125, 63.77208203125, 63.77658984375]",tokens/s,0.9877679019565393,kWh,0.0007529117455250688,0.00041266306092793453,0.003710265523765599,0.004875840330218602,tokens/kWh,12920.849686063342,,s,629,646.6700090332033,1.028092224218129,0.13044395391192767,1.0122936401367189,1.0129029296874998,1.0132424438476564,2.1090928125,"[1.0118082275390625, 1.0120591430664063, 1.0121963500976563, 1.0122721557617187, 1.0119987182617187, 1.0119086303710938, 1.0127564697265625, 1.0119772338867188, 1.013127197265625, 1.0117877807617188, 1.0121246948242189, 1.0118850708007812, 1.0127083740234375, 1.0122250366210936, 1.0122311401367188, 1.01222705078125, 1.0121298217773438, 1.01207958984375, 1.0122537231445312, 1.0121769409179688, 1.0120580444335938, 1.0120304565429687, 1.0126674194335938, 1.0122168579101563, 1.0119823608398437, 1.0118645629882812, 1.0125404052734375, 1.0122362670898437, 1.01228955078125, 1.012083740234375, 1.0118092651367188, 1.0122362670898437, 1.0126837768554688, 1.0127236938476563, 1.0127257690429687, 1.0131476440429688, 1.0123878784179687, 1.0122587280273438, 1.0124882202148437, 1.0123243408203124, 1.0121123657226563, 1.0124574584960937, 1.012696044921875, 1.0123038940429687, 1.012063232421875, 1.01218408203125, 1.0126653442382811, 1.012890625, 1.0124677124023438, 1.0129090576171875, 1.0120939331054688, 1.01304931640625, 1.0123591918945312, 1.0120345458984374, 1.0119608154296875, 1.0121820068359375, 1.0124882202148437, 1.0120724487304686, 1.01193115234375, 1.0123868408203125, 1.0123858642578125, 1.0128014526367188, 2.113512451171875, 1.0119444580078125, 1.0120970458984375, 1.0116249389648437, 1.0117980346679687, 1.0118410034179688, 1.0118922119140625, 1.0117406616210938, 1.012063232421875, 1.0121236572265624, 1.0118225708007813, 1.0117457885742187, 1.012200439453125, 1.0118594360351563, 1.0123018188476562, 1.0119075927734376, 1.0119393310546876, 1.0118450927734375, 1.0121697387695312, 1.0119772338867188, 1.0117940063476563, 1.0117733764648438, 1.0121502685546875, 1.0117877807617188, 1.0121226196289062, 1.0117345581054686, 1.011778564453125, 1.0116874389648438, 1.0122762451171874, 1.0118328247070312, 1.0120519409179687, 1.0118645629882812, 1.0123099975585939, 1.0122506103515625, 1.012801513671875, 1.0120089721679688, 1.0126571655273438, 1.0122014770507812, 1.0129817504882812, 1.0126458740234374, 1.012490234375, 1.0121615600585938, 1.0127646484375, 1.0119393310546876, 1.0128711547851563, 1.011999755859375, 1.0165934448242187, 1.0120037841796874, 1.0129735717773438, 1.0124287719726563, 1.0122291259765626, 1.0124595336914062, 1.01273291015625, 1.0122619018554688, 1.0127390747070313, 1.0128394165039063, 1.0125383911132813, 1.01269091796875, 1.013369873046875, 1.0124830932617188, 1.0126510009765626, 1.0123509521484375, 1.0127933349609375, 2.1089033203125, 1.0117857055664063, 1.0120253295898438, 1.012158447265625, 1.0124718017578125, 1.0120816650390625, 1.0120447998046875, 1.0125680541992188, 1.0123571166992187, 1.0120724487304686, 1.0127575073242188, 1.01252001953125, 1.0122987060546875, 1.0124451904296874, 1.0127728881835938, 1.011794921875, 1.011989501953125, 1.0122485961914063, 1.0122066040039062, 1.0119823608398437, 1.012210693359375, 1.0119772338867188, 1.01199462890625, 1.0123724975585937, 1.0123673706054688, 1.0120867919921874, 1.0119280395507813, 1.0121994018554688, 1.0121431274414063, 1.0121226196289062, 1.0124666748046875, 1.0119382934570313, 1.012031494140625, 1.0129346313476562, 1.012621337890625, 1.0127984619140624, 1.0124257202148437, 1.012552734375, 1.0124932861328124, 1.0126151733398439, 1.0124769287109374, 1.0124185791015625, 1.0123202514648437, 1.0127390747070313, 1.0145853271484375, 1.0121594848632813, 1.0121994018554688, 1.016111083984375, 1.01210009765625, 1.012337646484375, 1.0126919555664062, 1.0125444946289062, 1.0131015625, 1.0132469482421875, 1.0124186401367188, 1.0119146728515624, 1.0125578002929687, 1.0123294677734376, 1.0121595458984376, 1.0121768798828126, 1.0121963500976563, 1.0120345458984374, 1.0120714111328124, 2.108083251953125, 1.0117611694335937, 1.0117723999023438, 1.0123253784179687, 1.0121656494140625, 1.011751953125, 1.011820556640625, 1.0120601806640626, 1.0119454956054688, 1.01188916015625, 1.0119761962890625, 1.0117877807617188, 1.0130258178710938, 1.0123018188476562, 1.0121328735351562, 1.01246875, 1.0121953125, 1.0118369140625, 1.0122199096679687, 1.0124267578125, 1.0122546997070312, 1.0124503173828125, 1.0120703735351562, 1.0122537231445312, 1.0119536743164061, 1.0123397216796874, 1.0123960571289063, 1.0120325317382812, 1.0122772216796876, 1.012115478515625, 1.012005859375, 1.011937255859375, 1.0124103393554686, 1.0128527221679688, 1.012220947265625, 1.0126182250976563, 1.0130596313476563, 1.012464599609375, 1.0132828369140625, 1.0123468627929688, 1.0119547119140626, 1.0120048828125, 1.0124175415039063, 1.012010986328125, 1.0151219482421876, 1.0123069458007812, 1.0124124145507813, 1.0122076416015624, 1.0128138427734374, 1.0126878662109375, 1.0123939819335936, 1.0121318359375, 1.01235302734375, 1.011895263671875, 1.0121062622070311, 1.0122885131835937, 1.0123724975585937, 1.012242431640625, 1.012421630859375, 1.0123786010742188, 1.0121113891601563, 1.0120017700195312, 1.0122393798828124, 2.10916650390625, 1.0120038452148437, 1.0127247314453125, 1.0125916137695312, 1.0125660400390626, 1.0119761962890625, 1.0121748657226564, 1.0119721069335939, 1.0121370239257812, 1.0121942138671876, 1.0123048706054687, 1.0119403686523438, 1.0117099609375, 1.01187890625, 1.0121195678710937, 1.0117509155273436, 1.0118615112304687, 1.013433349609375, 1.012052978515625, 1.0117959594726562, 1.0120274047851563, 1.0121123657226563, 1.012421630859375, 1.012052978515625, 1.0121298217773438, 1.0119987182617187, 1.0118973388671875, 1.0125066528320312, 1.0120836791992187, 1.0121307983398438, 1.01264697265625, 1.0120714111328124, 1.0120857543945312, 1.0123540649414062, 1.0124544067382812, 1.0121564331054687, 1.0122383422851562, 1.0122454833984376, 1.0119935913085938, 1.0119239501953126, 1.0125363159179688, 1.0117755126953125, 1.011968017578125, 1.0119495239257812, 1.0124840698242188, 1.0122025146484375, 1.0122854614257812, 1.01258447265625, 1.0125885620117188, 1.0125301513671876, 1.012516845703125, 1.012400146484375, 1.0123970336914063, 1.0127892456054688, 1.01247998046875, 1.01256396484375, 1.0124237060546875, 1.0165504150390625, 1.0126458740234374, 1.0126848754882813, 1.0138418579101562, 1.0126428833007812, 1.0132418212890626, 2.10956298828125, 1.0119423828125, 1.0119608154296875, 1.0120847778320312, 1.011984375, 1.0120007934570312, 1.0124031982421875, 1.01182568359375, 1.0117908325195313, 1.0122680053710937, 1.012263916015625, 1.0119035034179686, 1.0124677734375, 1.0120057983398438, 1.0118276977539062, 1.011904541015625, 1.0125209350585938, 1.012252685546875, 1.0120653076171875, 1.0124205932617187, 1.0122495727539063, 1.0123724975585937, 1.0126407470703125, 1.012590576171875, 1.0119659423828125, 1.0125209350585938, 1.0127493286132812, 1.0124779663085937, 1.0127708129882813, 1.0121123657226563, 1.0121646118164063, 1.0116229248046875, 1.0122229614257812, 1.0120335083007812, 1.0120325317382812, 1.0120078735351563, 1.0126233520507812, 1.0124185791015625, 1.01650634765625, 1.0135787353515624, 1.012727783203125, 1.0122936401367189, 1.0128947143554687, 1.0121277465820313, 1.012600830078125, 1.0126612548828124, 1.0123960571289063, 1.012105224609375, 1.0127769775390625, 1.01275341796875, 1.012274169921875, 1.012279296875, 1.0123386840820312, 1.0121942749023438, 1.0125885620117188, 1.012947998046875, 1.0124779663085937, 1.0123930053710937, 1.0128373413085938, 1.0122721557617187, 1.01233251953125, 1.0123386840820312, 1.0127564697265625, 2.110841796875, 1.0128209838867188, 1.0126233520507812, 1.012595703125, 1.0131865844726562, 1.01278515625, 1.0128342895507811, 1.0131220703125, 1.0128527221679688, 1.012337646484375, 1.0124810180664063, 1.0129080200195313, 1.012747314453125, 1.0124318237304688, 1.0134036254882812, 1.0129408569335938, 1.0123612060546876, 1.0120929565429688, 1.0134886474609375, 1.0132265014648438, 1.0129080200195313, 1.0132152099609375, 1.0127984619140624, 1.0145208129882812, 1.013728271484375, 1.0132162475585937, 1.0127083740234375, 1.0127789916992187, 1.012833251953125, 1.0120120239257813, 1.0122465209960938, 1.0125762329101562, 1.0127605590820312, 1.0129448852539062, 1.01275341796875, 1.0117826538085937, 1.0119331665039062, 1.0120591430664063, 1.0131445922851563, 1.0131220703125, 1.0128814086914062, 1.0128875732421876, 1.0127513427734376, 1.013012451171875, 1.0131896362304686, 1.0130308837890625, 1.0130473022460937, 1.0133401489257812, 1.0132930297851563, 1.0126981201171874, 1.0128568115234375, 1.012906982421875, 1.0124503173828125, 1.0124984130859376, 1.0132520751953125, 1.0128076782226563, 1.0122936401367189, 1.0126827392578126, 1.0122035522460937, 1.0120509033203124, 1.0123540649414062, 1.0125946655273437, 1.0123171997070313, 2.112021484375, 1.0123406982421874, 1.0120253295898438, 1.0121513061523437, 1.0124144897460938, 1.012274169921875, 1.012401123046875, 1.0119710693359374, 1.012263916015625, 1.0128281860351562, 1.0122537231445312, 1.0122373046875, 1.0121615600585938, 1.0122373046875, 1.0120325317382812, 1.0120929565429688, 1.01382861328125, 1.0119669799804687, 1.0119239501953126, 1.0123489379882813, 1.0119403686523438, 1.0117539672851563, 1.0119976806640625, 1.0121533203125, 1.0122137451171875, 1.0119065551757813, 1.0129019165039062, 1.0124779052734374, 1.0123939819335936, 1.0126571655273438, 1.0125834350585938, 1.0119721069335939, 1.0120038452148437, 1.012474853515625, 1.0120499267578125, 1.0122383422851562, 1.0124830932617188, 1.0124758911132812, 1.0127083740234375, 1.012632568359375, 1.0122034912109374, 1.0120222778320314, 1.012178955078125, 1.0122977294921875, 1.0122034912109374, 1.01212158203125, 1.0132428588867188, 1.0123274536132814, 1.0125435180664062, 1.0123755493164062, 1.0122383422851562, 1.01226904296875, 1.0125250854492187, 1.0119833374023437, 1.0120601806640626, 1.0121134033203125, 1.012974609375, 1.0124656372070313, 1.01243798828125, 1.012580322265625, 1.0127124633789062, 1.0127247314453125, 1.0126704711914063, 2.11262353515625, 1.0118778686523437, 1.0124564208984375, 1.0118143920898437, 1.0125772705078124, 1.0121441040039063, 1.0118225708007813, 1.0119639892578125, 1.0122833251953125, 1.011894287109375, 1.0119280395507813, 1.0119198608398436, 1.0127401123046875, 1.01235205078125, 1.012570068359375, 1.011726318359375, 1.0117744750976563, 1.013375, 1.0122465209960938, 1.01218505859375, 1.0120325317382812, 1.011989501953125, 1.0127339477539063, 1.0121031494140624, 1.0123878173828125, 1.0119874267578124, 1.011962890625, 1.0119680786132812, 1.0123140258789063, 1.0120407104492188, 1.012220947265625, 1.0121410522460939, 1.0122772216796876, 1.011979248046875, 1.0125343017578126, 1.01197412109375, 1.01228955078125, 1.0118973388671875, 1.0122465209960938, 1.0119423828125, 1.012041748046875, 1.0123131103515626, 1.0124677124023438, 1.0119342041015624, 1.012621337890625, 1.012105224609375, 1.0122291259765626, 1.0119669799804687, 1.0125332641601563, 1.0122034912109374, 1.0122045288085937, 1.0123519897460938, 1.012875244140625, 1.0126397705078125, 1.0126233520507812, 1.0123182373046875, 1.0127708129882813, 1.0124646606445313, 1.0125987548828126, 1.0120775756835938, 1.012442138671875, 1.012552734375, 1.012252685546875, 2.11210546875, 1.011620849609375, 1.0120519409179687, 1.0119721069335939, 1.0122485961914063, 1.0120714111328124, 1.0120621948242188, 1.0127001342773438, 1.0125178833007813, 1.012548583984375, 1.01243701171875, 1.0118604736328125, 1.0117130126953124, 1.0120304565429687, 1.0122587890625, 1.01167822265625, 1.0118010864257811, 1.0119813232421875, 1.011894287109375, 1.0122905883789062, 1.01184716796875, 1.0117959594726562, 1.0116761474609375, 1.0118410034179688, 1.0117744750976563, 1.0119515991210937, 1.0144215087890625, 1.0122987670898438, 1.012021240234375, 1.0126315307617189, 1.0124862060546875, 1.0122158203125, 1.0121727905273437, 1.0125875244140625, 1.0124472045898438, 1.0122403564453124, 1.0122117309570313, 1.0127349853515626, 1.0121431274414063, 1.012495361328125, 1.012738037109375, 1.0124758911132812, 1.0128773193359375, 1.0123079833984374, 1.0123099975585939, 1.0121298217773438, 1.0122557373046874, 1.0126878662109375, 1.0124810180664063, 1.013000244140625, 1.012516845703125, 1.0124298095703126, 1.012358154296875, 1.0124656372070313, 1.0123171997070313, 1.01239501953125, 1.01268994140625, 1.0127656860351562, 1.01236328125, 1.0124308471679688, 1.012527099609375, 1.0123038940429687, 1.0129132080078125]",tokens/s,0.972675384993313,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen2-beta-14B,Qwen/Qwen2-beta-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,3905.671168,12732.33408,0.0,12085.886976,11337.501696,s,10,10.970613281250001,1.097061328125,0.0017840508518984434,1.097439453125,1.0989015869140624,1.0993332397460938,1.0996785620117187,"[1.0981041259765625, 1.099764892578125, 1.0942474365234376, 1.09466943359375, 1.09612890625, 1.0954654541015625, 1.0967747802734376, 1.0988056640625, 1.098486328125, 1.098166259765625]",tokens/s,233.35067369253872,kWh,1.2926500191291174e-05,7.083296092205274e-06,6.255157781898113e-05,8.256137410247758e-05,tokens/kWh,3100723.5863376665,MB,3909.746688,12732.33408,0.0,12085.886976,11686.806016,s,10,636.987296875,63.6987296875,0.008792231338471245,63.69969140625,63.70640390625,63.70996171875,63.71280796875,"[63.682359375, 63.687671875, 63.70561328125, 63.70497265625, 63.69943359375, 63.69994921875, 63.70396484375, 63.71351953125, 63.6979921875, 63.6918203125]",tokens/s,0.9890307123716925,kWh,0.0007521740069488685,0.00041225789324771534,0.003629864737222821,0.0047942966374194046,tokens/kWh,13140.613684244328,,s,629,645.8482131958002,1.0267857125529425,0.13035148276579694,1.0110596313476563,1.01170830078125,1.0120460571289063,2.107026708984375,"[1.0111242065429686, 1.0104995727539063, 1.0103193359375, 1.0103952026367187, 1.0104954223632812, 1.0104678955078126, 1.0100704956054687, 1.0106204223632813, 1.0101810913085938, 1.0110658569335937, 1.0104442749023437, 1.0103756713867187, 1.0101524658203125, 1.0103828735351563, 1.0101749877929687, 1.0103490600585938, 1.0100950927734376, 1.0109429931640626, 1.0103602905273437, 1.0109215698242187, 1.0110667724609375, 1.0112665405273438, 1.0110863647460937, 1.01104638671875, 1.0108724365234374, 1.0105343017578126, 1.0103214111328125, 1.010872314453125, 1.0106912841796876, 1.0113380737304687, 1.01102490234375, 1.011188720703125, 1.0115625, 1.0118072509765625, 1.011420166015625, 1.0109522094726562, 1.010282470703125, 1.0109522094726562, 1.0106019897460938, 1.0113290405273438, 1.0105897216796875, 1.0110146484375, 1.0105538330078125, 1.0109419555664063, 1.0105497436523438, 1.0111878051757812, 1.0105476684570311, 1.0111273193359376, 1.0112327880859375, 1.0111181640625, 1.0113422241210936, 1.0108223266601561, 1.0107902221679688, 1.0107567749023438, 1.0107431030273437, 1.0107125854492187, 1.0108497924804687, 1.0110812377929688, 1.0113054809570312, 1.011472412109375, 1.010951171875, 1.011072998046875, 2.110793701171875, 1.0105712890625, 1.0109173583984374, 1.0111078491210939, 1.0105538940429688, 1.0114539184570313, 1.0104873046875, 1.0103736572265625, 1.01060302734375, 1.0104473876953124, 1.0103705444335938, 1.0103173217773438, 1.0103797607421876, 1.0106593017578125, 1.0105569458007813, 1.0109276123046875, 1.0109112548828125, 1.0105016479492188, 1.0110904541015624, 1.0104453125, 1.0106286010742187, 1.0105743408203125, 1.0106060791015625, 1.0110525512695312, 1.0107782592773438, 1.0118408813476563, 1.0113218383789062, 1.0111395874023437, 1.0107064208984375, 1.010524169921875, 1.0105466918945312, 1.0102415161132812, 1.010682861328125, 1.0110791625976563, 1.0108632202148438, 1.01125830078125, 1.0108221435546876, 1.0107473754882812, 1.0111160278320312, 1.0105282592773437, 1.0108026733398439, 1.0104678344726563, 1.0108098754882813, 1.0106849365234376, 1.0112696533203125, 1.0114283447265624, 1.01136181640625, 1.01125634765625, 1.0113484497070313, 1.0112225341796874, 1.0113013916015625, 1.0110914306640626, 1.0109788208007813, 1.011293212890625, 1.0114918212890625, 1.0115665893554688, 1.0111375122070312, 1.0111918334960937, 1.0110986328125, 1.0106838989257811, 1.011178466796875, 1.01096142578125, 1.0111539306640625, 2.106599365234375, 1.0114857177734375, 1.0112235717773437, 1.011198974609375, 1.01125732421875, 1.010946044921875, 1.0106123657226562, 1.010948974609375, 1.0108016357421874, 1.0105702514648438, 1.011140625, 1.0107391967773438, 1.0105272216796874, 1.010386962890625, 1.0104515991210938, 1.010242431640625, 1.0104063720703125, 1.010171875, 1.0105518188476563, 1.0103695068359375, 1.01096142578125, 1.0108364868164061, 1.0114160766601563, 1.0112081909179687, 1.0110771484375, 1.01035107421875, 1.0108743896484376, 1.01146728515625, 1.0117959594726562, 1.0120355834960937, 1.012463623046875, 1.0119403686523438, 1.01194140625, 1.0119802856445312, 1.011323974609375, 1.0110104370117188, 1.0110330810546875, 1.0113382568359375, 1.0116249389648437, 1.0108795166015625, 1.01194140625, 1.0112030639648437, 1.0113484497070313, 1.011093505859375, 1.0115000610351563, 1.0112604370117189, 1.0109788208007813, 1.0104954833984374, 1.011493896484375, 1.01127783203125, 1.0117058715820313, 1.0108528442382811, 1.0114898071289062, 1.0111221923828124, 1.0114129638671876, 1.011178466796875, 1.0114631958007811, 1.0113863525390625, 1.0115481567382814, 1.01125732421875, 1.0124105834960937, 1.0113810424804688, 1.0117324829101562, 2.10709814453125, 1.0106634521484374, 1.0110371704101562, 1.0109081420898438, 1.0120171508789062, 1.0121226196289062, 1.011262451171875, 1.0106112060546875, 1.0108251953125, 1.0108651733398437, 1.0108108520507812, 1.0108057861328126, 1.0110238647460938, 1.0108446655273438, 1.01113037109375, 1.0110238647460938, 1.0114150390625, 1.01075146484375, 1.0109603881835938, 1.0105835571289064, 1.010745361328125, 1.0105211181640625, 1.0108262329101563, 1.01054052734375, 1.0108231811523438, 1.0112245483398437, 1.0115286865234374, 1.0111610717773438, 1.0113402709960937, 1.0108292846679687, 1.0113024291992188, 1.0110156860351562, 1.011251220703125, 1.0109337768554687, 1.0112122802734376, 1.0112041015625, 1.0106275634765625, 1.0104483642578126, 1.0109224853515626, 1.0123694458007813, 1.0121512451171875, 1.0115277099609374, 1.011831787109375, 1.0113003540039063, 1.0119905395507813, 1.0121932983398438, 1.01144677734375, 1.0107801513671875, 1.011119140625, 1.0112194213867187, 1.0110924682617188, 1.0110105590820313, 1.0112767944335936, 1.0110873413085937, 1.0113557739257812, 1.0116627807617187, 1.0114488525390626, 1.01161474609375, 1.0121482543945313, 1.0110945434570313, 1.0111918334960937, 1.0114703369140625, 1.011420166015625, 2.107243408203125, 1.011304443359375, 1.0116792602539062, 1.0110904541015624, 1.0111897583007812, 1.0108170166015624, 1.0107166748046874, 1.010555908203125, 1.0108446655273438, 1.0105784301757812, 1.0112839965820313, 1.0110791625976563, 1.011198974609375, 1.0106951904296875, 1.0111498413085938, 1.0108477172851562, 1.0110279541015625, 1.0106736450195313, 1.01078125, 1.0110596313476563, 1.0108917846679688, 1.0110576782226564, 1.0108016357421874, 1.01058251953125, 1.0113638305664063, 1.01103515625, 1.0112214965820312, 1.0109030151367187, 1.0114221801757812, 1.01172021484375, 1.011646484375, 1.0109634399414062, 1.0110392456054687, 1.01072998046875, 1.0113239135742187, 1.0107545776367188, 1.0113812255859376, 1.0107801513671875, 1.0113116455078126, 1.011267578125, 1.0112327880859375, 1.0109522094726562, 1.0110904541015624, 1.0106736450195313, 1.0117355346679688, 1.0115389404296875, 1.0110474243164063, 1.0108436279296875, 1.01097265625, 1.0110955810546876, 1.011800048828125, 1.0115277099609374, 1.0110167236328125, 1.0107422485351563, 1.0111610717773438, 1.0109552612304689, 1.01119384765625, 1.0112645263671876, 1.0109450073242188, 1.0111027221679687, 1.0116802368164062, 1.0114508666992188, 1.0115563354492187, 2.106843017578125, 1.01136181640625, 1.0113515625, 1.0108784790039063, 1.0107105102539062, 1.0108948364257813, 1.0118154296875, 1.0109542236328124, 1.0106787719726562, 1.0108887329101564, 1.0112225341796874, 1.0112604370117189, 1.0115020751953125, 1.010966552734375, 1.0110965576171875, 1.0111488037109375, 1.0121779174804688, 1.01123583984375, 1.0113885498046875, 1.011718017578125, 1.0113597412109374, 1.0111610717773438, 1.0113341674804688, 1.0110709838867187, 1.01110986328125, 1.0114406127929687, 1.0113106079101561, 1.0103971557617188, 1.0105989379882812, 1.0106224365234375, 1.010735107421875, 1.0104473876953124, 1.010703369140625, 1.0104063720703125, 1.01077099609375, 1.0111979370117188, 1.01096240234375, 1.0108262329101563, 1.011472412109375, 1.011409912109375, 1.0115133666992187, 1.0109255981445313, 1.0116690063476563, 1.0109921264648438, 1.0113484497070313, 1.011146728515625, 1.0115604248046874, 1.0110126342773438, 1.0111477661132813, 1.0111273193359376, 1.01097265625, 1.0108907470703126, 1.0107822265625, 1.0106972045898437, 1.0111293334960938, 1.011146728515625, 1.0112236938476562, 1.0110370483398436, 1.0111190795898437, 1.0111826171875, 1.01117236328125, 1.0109869995117187, 1.0112276611328126, 2.108168212890625, 1.0113248901367187, 1.0116497192382812, 1.0111066284179688, 1.0111590576171876, 1.010862060546875, 1.0108671875, 1.0107658081054687, 1.0111181030273437, 1.010798583984375, 1.0113054809570312, 1.0110842895507812, 1.01054052734375, 1.0102753295898437, 1.0105231323242188, 1.0106941528320312, 1.0107473754882812, 1.0110699462890624, 1.01083544921875, 1.0106234741210938, 1.0105692138671876, 1.0113546142578125, 1.0107955322265625, 1.01106689453125, 1.011330078125, 1.0107811889648437, 1.010820068359375, 1.0105374755859375, 1.0116658935546874, 1.011794921875, 1.0121381225585937, 1.0115756225585937, 1.0113126220703126, 1.0110075073242188, 1.0110658569335937, 1.01071875, 1.0106624145507812, 1.0105538330078125, 1.0108600463867188, 1.0104791259765624, 1.0111047973632812, 1.010713623046875, 1.0108712768554688, 1.0103910522460937, 1.0111826171875, 1.0108528442382811, 1.0117273559570312, 1.0112481079101563, 1.0113095703125, 1.0112542724609375, 1.0117642211914062, 1.0118338623046874, 1.0117969970703125, 1.0115205078125, 1.0120662841796875, 1.0122158203125, 1.011726318359375, 1.0116546630859375, 1.011726318359375, 1.0113873901367187, 1.0115338134765626, 1.0122117309570313, 1.0120550537109374, 2.11171728515625, 1.0110648193359375, 1.0112655639648438, 1.011040283203125, 1.0112337646484375, 1.011103759765625, 1.0117447509765625, 1.011694580078125, 1.0115451049804687, 1.0108006591796874, 1.0108999633789062, 1.0106654663085937, 1.0113208618164062, 1.0110341186523437, 1.0105753784179687, 1.0110699462890624, 1.0127493286132812, 1.0114774780273437, 1.0117990112304687, 1.0114754638671875, 1.01243701171875, 1.0126602172851562, 1.0127821044921874, 1.012316162109375, 1.0116639404296874, 1.0120989990234375, 1.0121307983398438, 1.0107197265625, 1.0107975463867187, 1.0107320556640624, 1.0111702880859375, 1.0104524536132813, 1.0106388549804688, 1.01058251953125, 1.0108671875, 1.0108385009765626, 1.0115369262695313, 1.0105947875976562, 1.010924560546875, 1.0109214477539064, 1.0113648681640626, 1.0106675415039061, 1.0112532348632812, 1.0106183471679688, 1.0113659057617188, 1.0111826171875, 1.0120530395507812, 1.0109584350585938, 1.0110103759765625, 1.0113085327148437, 1.0110975952148438, 1.0110146484375, 1.01137109375, 1.0108394775390626, 1.0119804077148438, 1.01148046875, 1.0113474731445313, 1.0114826049804688, 1.0118276977539062, 1.0114006958007813, 1.011209228515625, 1.0110105590820313, 1.0114006958007813, 2.110498779296875, 1.0119219360351563, 1.01136181640625, 1.0108313598632812, 1.0110064697265626, 1.0115112915039062, 1.0112849731445313, 1.011282958984375, 1.0111334228515625, 1.01097265625, 1.011146728515625, 1.0113351440429688, 1.0111365356445312, 1.010808837890625, 1.010713623046875, 1.0107698974609376, 1.0108856201171874, 1.0107218017578126, 1.01076171875, 1.0107422485351563, 1.0111477661132813, 1.011072998046875, 1.01098291015625, 1.0104760131835937, 1.0106972045898437, 1.01064501953125, 1.0107012939453126, 1.010724853515625, 1.0107740478515626, 1.0106009521484376, 1.010820068359375, 1.0108784790039063, 1.0107863159179689, 1.0108590698242188, 1.01104736328125, 1.0110914306640626, 1.0109132690429687, 1.0108549194335938, 1.0113065795898437, 1.0108507690429687, 1.0111047973632812, 1.0121011352539062, 1.0114979858398438, 1.0112440185546876, 1.0114396362304687, 1.0111139526367188, 1.0110658569335937, 1.01098291015625, 1.011177490234375, 1.0114918212890625, 1.0113802490234376, 1.0121666259765625, 1.0110453491210938, 1.0111324462890625, 1.0114744262695312, 1.0106736450195313, 1.0109685668945312, 1.0110310668945313, 1.0110238647460938, 1.0108549194335938, 1.0109911499023438, 1.0115542602539063, 1.010951171875, 2.10921875, 1.0104708862304688, 1.0106941528320312, 1.0109368286132812, 1.0107924194335938, 1.0110361328125, 1.0113505249023438, 1.0111631469726563, 1.010882568359375, 1.011051513671875, 1.0112184448242187, 1.0114017333984375, 1.0117243041992188, 1.0111027221679687, 1.0108477172851562, 1.0110914306640626, 1.0114437255859374, 1.01085595703125, 1.0109983520507813, 1.0107094116210937, 1.0110496215820313, 1.010912109375, 1.0108661499023437, 1.010777099609375, 1.0107576904296875, 1.0106285400390624, 1.0111324462890625, 1.011009521484375, 1.01085693359375, 1.0107545776367188, 1.0111273193359376, 1.010572265625, 1.0108016357421874, 1.0107443237304687, 1.0111344604492187, 1.010861083984375, 1.0115419921875, 1.0107218017578126, 1.0108344116210937, 1.0103900146484375, 1.0109389038085939, 1.0106685180664063, 1.0109644775390625, 1.0104258422851562, 1.0109030151367187, 1.0110167236328125, 1.0119915771484376, 1.011103759765625, 1.0113802490234376, 1.0111365356445312, 1.0113638305664063, 1.0108446655273438, 1.0109859619140624, 1.01060400390625, 1.0110167236328125, 1.0108129272460937, 1.0115399780273437, 1.0108231811523438, 1.0112481079101563, 1.0108765869140626, 1.0110595703125, 1.0108630981445312, 1.0109900512695313]",tokens/s,0.9739130451218064,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-30b,facebook/opt-30b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,5984.595968,19933.954048,0.0,19287.506944,18376.399872,s,10,24.56320629882812,2.456320629882813,0.001691664801696018,2.4561865234375,2.4583423828125,2.4586625244140623,2.4589186376953127,"[2.456571533203125, 2.458982666015625, 2.4545791015625, 2.455032470703125, 2.454942626953125, 2.453741943359375, 2.455801513671875, 2.458271240234375, 2.45821728515625, 2.45706591796875]",tokens/s,104.22092168489151,kWh,2.8975297907988232e-05,1.5879395596475657e-05,0.00013865674981416998,0.00018351144331863386,tokens/kWh,1395008.3731590684,MB,5988.143104,19933.954048,0.0,19287.506944,18871.04,s,10,1458.421484375,145.84214843750001,0.012211079453744932,145.84282812499998,145.852815625,145.8578921875,145.8619534375,"[145.82440625, 145.837078125, 145.840359375, 145.8516875, 145.845296875, 145.8391875, 145.86296875, 145.8513125, 145.82034375, 145.84884375]",tokens/s,0.43197388872119064,kWh,0.001721854172928466,0.000943728472916573,0.008154870135002046,0.010820452780847084,tokens/kWh,5822.307187691273,,s,629,1478.2121892089838,2.350098869966589,0.29170661243792756,2.3148583984375,2.3159804199218748,2.3165888183593752,4.76929474609375,"[2.31657470703125, 2.316729248046875, 2.31575048828125, 2.31400439453125, 2.31361328125, 2.314355712890625, 2.31430859375, 2.313990234375, 2.314439697265625, 2.31400439453125, 2.314337158203125, 2.31450830078125, 2.3143525390625, 2.313754638671875, 2.3145595703125, 2.314746826171875, 2.313860107421875, 2.314616943359375, 2.31397998046875, 2.314455078125, 2.314322021484375, 2.315072509765625, 2.314198974609375, 2.314560546875, 2.31431689453125, 2.314648681640625, 2.314249267578125, 2.3142431640625, 2.31486767578125, 2.314310546875, 2.31423681640625, 2.314349609375, 2.31522509765625, 2.314227783203125, 2.314047607421875, 2.314390380859375, 2.31465380859375, 2.31459130859375, 2.3147294921875, 2.31495263671875, 2.31514306640625, 2.314708984375, 2.314705810546875, 2.314702880859375, 2.31510009765625, 2.314967041015625, 2.314390625, 2.3144326171875, 2.31436279296875, 2.3152353515625, 2.31499169921875, 2.31524560546875, 2.314324951171875, 2.3151728515625, 2.31494970703125, 2.314418212890625, 2.315260009765625, 2.3155927734375, 2.314987548828125, 2.3148759765625, 2.314437744140625, 2.314975341796875, 4.77530908203125, 2.315552734375, 2.313754638671875, 2.31431787109375, 2.314470458984375, 2.31491064453125, 2.313483154296875, 2.3140751953125, 2.314690673828125, 2.313935791015625, 2.31429541015625, 2.313943115234375, 2.3145830078125, 2.314388427734375, 2.3144130859375, 2.314663818359375, 2.314819580078125, 2.31438134765625, 2.31452978515625, 2.313989013671875, 2.314265625, 2.31451025390625, 2.31416015625, 2.313996337890625, 2.314390625, 2.314997802734375, 2.31469970703125, 2.31465380859375, 2.314850341796875, 2.315021240234375, 2.31556396484375, 2.314933349609375, 2.315074462890625, 2.314526611328125, 2.31516162109375, 2.314416015625, 2.31450634765625, 2.3146474609375, 2.31455322265625, 2.314646484375, 2.315080810546875, 2.314736572265625, 2.315146240234375, 2.316416015625, 2.316883056640625, 2.317212646484375, 2.317498291015625, 2.316708984375, 2.314590087890625, 2.31486669921875, 2.314609619140625, 2.314794921875, 2.3152783203125, 2.314463134765625, 2.31488720703125, 2.3152783203125, 2.3158486328125, 2.315106201171875, 2.31514208984375, 2.315134033203125, 2.314939453125, 2.31448974609375, 2.31490966796875, 4.76900048828125, 2.314322998046875, 2.31459423828125, 2.314704833984375, 2.3145400390625, 2.314397705078125, 2.315212890625, 2.315070556640625, 2.3151708984375, 2.314818603515625, 2.31474072265625, 2.314556396484375, 2.314756103515625, 2.314300537109375, 2.314501220703125, 2.314758056640625, 2.31469580078125, 2.314715087890625, 2.314291259765625, 2.313954345703125, 2.3144990234375, 2.3141162109375, 2.314789794921875, 2.3138662109375, 2.314201171875, 2.31396044921875, 2.314891357421875, 2.313994140625, 2.3142041015625, 2.31410595703125, 2.31444384765625, 2.314310546875, 2.3143935546875, 2.3148564453125, 2.314733642578125, 2.314330078125, 2.31438134765625, 2.3142685546875, 2.3148154296875, 2.315052978515625, 2.31476318359375, 2.31545751953125, 2.315313232421875, 2.315279296875, 2.315707275390625, 2.314758056640625, 2.3152158203125, 2.31581689453125, 2.315241455078125, 2.3146220703125, 2.315763671875, 2.3165615234375, 2.31669873046875, 2.3161025390625, 2.31678759765625, 2.31653173828125, 2.316707763671875, 2.3152138671875, 2.31585693359375, 2.314685546875, 2.315296875, 2.3153837890625, 2.315890625, 4.7694091796875, 2.314258544921875, 2.3142666015625, 2.31421435546875, 2.314051513671875, 2.314406982421875, 2.314616943359375, 2.314188720703125, 2.31503662109375, 2.31619287109375, 2.31446533203125, 2.313934814453125, 2.315040771484375, 2.314817626953125, 2.315828125, 2.3156357421875, 2.316775390625, 2.314610595703125, 2.31668017578125, 2.31729052734375, 2.31545556640625, 2.315461669921875, 2.315554931640625, 2.31465576171875, 2.314517578125, 2.3154697265625, 2.31606689453125, 2.314260498046875, 2.314863525390625, 2.314272705078125, 2.314649658203125, 2.3157421875, 2.315529296875, 2.3159091796875, 2.314768310546875, 2.3160771484375, 2.315504638671875, 2.314666015625, 2.314771484375, 2.315051025390625, 2.3154248046875, 2.314206298828125, 2.314478515625, 2.31497021484375, 2.315460693359375, 2.314807373046875, 2.31501708984375, 2.315440185546875, 2.315284423828125, 2.31450634765625, 2.314502197265625, 2.314387451171875, 2.314957763671875, 2.3149384765625, 2.314912841796875, 2.314670166015625, 2.3156572265625, 2.316516357421875, 2.3157412109375, 2.315861083984375, 2.315828125, 2.315040771484375, 2.3151728515625, 4.7702314453125, 2.31434033203125, 2.314682373046875, 2.313740234375, 2.314062744140625, 2.31486767578125, 2.314884033203125, 2.314498046875, 2.314965087890625, 2.314472412109375, 2.314133544921875, 2.31417041015625, 2.313879638671875, 2.313923583984375, 2.314146728515625, 2.314799072265625, 2.3140986328125, 2.314390625, 2.315798583984375, 2.31642822265625, 2.315537353515625, 2.316564453125, 2.314660888671875, 2.313965576171875, 2.314789794921875, 2.31402294921875, 2.314662841796875, 2.314347412109375, 2.315102294921875, 2.3144150390625, 2.315137939453125, 2.31604833984375, 2.315811767578125, 2.3149404296875, 2.31493115234375, 2.315124755859375, 2.3152158203125, 2.314859619140625, 2.315197509765625, 2.314984375, 2.315462646484375, 2.31446533203125, 2.3143095703125, 2.31438037109375, 2.31520556640625, 2.31512158203125, 2.314987548828125, 2.314483642578125, 2.31514208984375, 2.314712158203125, 2.31526513671875, 2.314966064453125, 2.314924072265625, 2.314987548828125, 2.31532958984375, 2.31476318359375, 2.3151298828125, 2.315618408203125, 2.316012451171875, 2.317046875, 2.316642333984375, 2.317076416015625, 2.317365234375, 4.7678056640625, 2.314080322265625, 2.31505908203125, 2.315978759765625, 2.315252685546875, 2.31602392578125, 2.3155966796875, 2.31449609375, 2.31440087890625, 2.31406396484375, 2.31448583984375, 2.314968017578125, 2.314263671875, 2.31400146484375, 2.314314697265625, 2.314521484375, 2.3153408203125, 2.31357861328125, 2.3147294921875, 2.31446728515625, 2.315052978515625, 2.314033203125, 2.314588134765625, 2.314469482421875, 2.31505517578125, 2.3144765625, 2.3147109375, 2.31465478515625, 2.315134033203125, 2.31459521484375, 2.31478466796875, 2.314600341796875, 2.314883056640625, 2.315136962890625, 2.315252685546875, 2.31488818359375, 2.315322265625, 2.31524755859375, 2.31511767578125, 2.3150263671875, 2.3153603515625, 2.315388916015625, 2.31535009765625, 2.31446826171875, 2.31495068359375, 2.31469775390625, 2.315263916015625, 2.314312744140625, 2.3146865234375, 2.314904541015625, 2.31448974609375, 2.314483642578125, 2.314174560546875, 2.314915771484375, 2.314462158203125, 2.3165869140625, 2.317093994140625, 2.31678759765625, 2.315781005859375, 2.315177978515625, 2.315336669921875, 2.31464453125, 2.31521484375, 4.770552734375, 2.314354736328125, 2.31482568359375, 2.313882568359375, 2.31376171875, 2.313788330078125, 2.315693115234375, 2.31585986328125, 2.31478271484375, 2.315926513671875, 2.3154453125, 2.315216796875, 2.314701904296875, 2.315554931640625, 2.31552197265625, 2.315801513671875, 2.316396484375, 2.314292236328125, 2.314648681640625, 2.31490771484375, 2.315061279296875, 2.3164384765625, 2.316370849609375, 2.316866455078125, 2.31659521484375, 2.31632080078125, 2.315591796875, 2.31587646484375, 2.3159716796875, 2.315903076171875, 2.315663330078125, 2.31596240234375, 2.316310546875, 2.314661865234375, 2.315462646484375, 2.31440380859375, 2.31526611328125, 2.315337646484375, 2.315576416015625, 2.31480419921875, 2.314577880859375, 2.316158935546875, 2.316590087890625, 2.31474072265625, 2.314672119140625, 2.315419677734375, 2.315107421875, 2.315031494140625, 2.314947509765625, 2.31512158203125, 2.31507861328125, 2.315275146484375, 2.31488916015625, 2.314337158203125, 2.315274169921875, 2.315357177734375, 2.314990478515625, 2.314895263671875, 2.315843505859375, 2.315281494140625, 2.315041748046875, 2.314649658203125, 2.3151708984375, 4.7742177734375, 2.314451904296875, 2.314017822265625, 2.316307373046875, 2.31636376953125, 2.315851806640625, 2.3162041015625, 2.315658203125, 2.315419677734375, 2.315569091796875, 2.31410693359375, 2.3144345703125, 2.3141181640625, 2.3143701171875, 2.314511474609375, 2.314544189453125, 2.31514208984375, 2.314577880859375, 2.314501220703125, 2.314337158203125, 2.315443115234375, 2.314441650390625, 2.314269775390625, 2.3141171875, 2.31440380859375, 2.315146240234375, 2.31492919921875, 2.314220458984375, 2.31446533203125, 2.3145595703125, 2.314896484375, 2.3150439453125, 2.314417236328125, 2.31532861328125, 2.315260009765625, 2.315807861328125, 2.31515234375, 2.315829345703125, 2.315790283203125, 2.3159296875, 2.31573193359375, 2.315969482421875, 2.316851318359375, 2.3155537109375, 2.314724365234375, 2.3145625, 2.315716552734375, 2.31531005859375, 2.315375732421875, 2.315345947265625, 2.31587744140625, 2.31564599609375, 2.315336669921875, 2.3145, 2.31510107421875, 2.31535400390625, 2.314984375, 2.31453076171875, 2.3153857421875, 2.314976318359375, 2.315102294921875, 2.31464453125, 2.31491064453125, 4.77313525390625, 2.31535205078125, 2.314526611328125, 2.3150869140625, 2.315890625, 2.314620849609375, 2.3140322265625, 2.313732177734375, 2.31478466796875, 2.31486669921875, 2.3156357421875, 2.31585986328125, 2.314743896484375, 2.314322021484375, 2.3139072265625, 2.31398291015625, 2.314142822265625, 2.31423388671875, 2.31518408203125, 2.31429833984375, 2.314175537109375, 2.3156572265625, 2.3145810546875, 2.313807861328125, 2.314330078125, 2.31499169921875, 2.314958740234375, 2.3144365234375, 2.3145185546875, 2.31440185546875, 2.315052978515625, 2.314460205078125, 2.3145419921875, 2.3140576171875, 2.314851318359375, 2.314546142578125, 2.314408935546875, 2.31389892578125, 2.31480419921875, 2.31436279296875, 2.3144580078125, 2.314177490234375, 2.314638427734375, 2.3145419921875, 2.31480322265625, 2.314745849609375, 2.314291259765625, 2.31537255859375, 2.31572265625, 2.31389599609375, 2.313974853515625, 2.314586181640625, 2.314895263671875, 2.314090576171875, 2.314375244140625, 2.314859619140625, 2.3148583984375, 2.3144326171875, 2.314412109375, 2.314270751953125, 2.31518115234375, 2.314324951171875, 2.314473388671875, 4.7714609375, 2.31414794921875, 2.313611328125, 2.314270751953125, 2.314892333984375, 2.31461572265625, 2.315198486328125, 2.3143720703125, 2.314142822265625, 2.31568701171875, 2.315873291015625, 2.31440478515625, 2.314210205078125, 2.31602587890625, 2.316198974609375, 2.31553857421875, 2.315966552734375, 2.31545947265625, 2.31545849609375, 2.31590185546875, 2.31619287109375, 2.31564599609375, 2.316030029296875, 2.315187255859375, 2.314642333984375, 2.313966552734375, 2.3152138671875, 2.31450830078125, 2.315548583984375, 2.314388427734375, 2.315707275390625, 2.3161650390625, 2.31663623046875, 2.316718994140625, 2.315875244140625, 2.315216796875, 2.31624609375, 2.315987060546875, 2.315421630859375, 2.315747314453125, 2.314558349609375, 2.313786376953125, 2.313956298828125, 2.3144541015625, 2.3148984375, 2.314367919921875, 2.31458203125, 2.315187255859375, 2.31473046875, 2.314349609375, 2.314651611328125, 2.31442626953125, 2.315454345703125, 2.31490771484375, 2.314850341796875, 2.315431884765625, 2.31543798828125, 2.314609619140625, 2.31511767578125, 2.314015625, 2.31457177734375, 2.314642333984375, 2.31451953125]",tokens/s,0.4255140125292758,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-1.3B,,cuda,0,42,,,,,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,,MB,1562.181632,1957.167104,0.0,1310.72,1163.955712,s,10,1.3245970916748049,0.13245970916748048,0.001266998397567483,0.13200260162353517,0.1341699493408203,0.13443305816650392,0.1346435452270508,"[0.1346961669921875, 0.1341114807128906, 0.1312022705078125, 0.13115029907226564, 0.13112367248535156, 0.1315875244140625, 0.13163551330566406, 0.13236968994140624, 0.13365353393554688, 0.13306694030761718]",tokens/s,1932.6631593031557,kWh,1.5487117942793545e-06,8.486256414471426e-07,6.508118482104666e-06,8.905455917831163e-06,tokens/kWh,28746422.683134936,MB,1562.181632,1959.264256,0.0,1312.817152,1232.77568,s,10,76.92137255859376,7.692137255859374,0.017487914910414803,7.693824951171875,7.702772509765625,7.711682641601563,7.718810747070313,"[7.69504345703125, 7.69082470703125, 7.6926064453125, 7.6866689453125, 7.7205927734375, 7.69639892578125, 7.6467568359375, 7.70079248046875, 7.69208056640625, 7.699607421875]",tokens/s,8.190181467707257,kWh,9.084797721340027e-05,4.979003748630931e-05,0.0003722497123017,0.0005128877270014096,tokens/kWh,122833.9004489902,,s,629,77.98684770965575,0.12398544945891218,0.01575529620495015,0.121744384765625,0.12330782623291016,0.12368689422607422,0.2534359008789062,"[0.12507750701904297, 0.12424604797363281, 0.1234441909790039, 0.12341862487792969, 0.12180786895751954, 0.1224120330810547, 0.12155903625488282, 0.12088832092285157, 0.12079001617431641, 0.12080332946777343, 0.1209518051147461, 0.12134912109375, 0.12099378967285156, 0.12220416259765625, 0.12147408294677735, 0.12154876708984375, 0.12117708587646485, 0.12111360168457032, 0.12086067199707032, 0.12105318450927735, 0.12099174499511718, 0.12216934204101562, 0.12297727966308594, 0.12166143798828125, 0.12153446197509765, 0.12155801391601563, 0.12161023712158203, 0.12131839752197265, 0.12115455627441406, 0.12147609710693359, 0.12320972442626953, 0.12328857421875, 0.12294041442871094, 0.1213460464477539, 0.12115360260009765, 0.12151289367675781, 0.12153343963623046, 0.12244172668457032, 0.12324454498291015, 0.12321791839599609, 0.1221396484375, 0.123504638671875, 0.12190930938720704, 0.12093536376953125, 0.12353228759765625, 0.1231247329711914, 0.12405248260498047, 0.12286566162109375, 0.12216831970214843, 0.1221580810546875, 0.12220928192138672, 0.12256153869628907, 0.12320460510253907, 0.12468940734863282, 0.12308889770507812, 0.1220997085571289, 0.12128972625732422, 0.12146380615234376, 0.12163587188720704, 0.12112380981445313, 0.12155289459228516, 0.12149759674072266, 0.2558975982666016, 0.12277452850341797, 0.12251545715332031, 0.12256665802001954, 0.12305101013183593, 0.12229837036132812, 0.1212968978881836, 0.1239582748413086, 0.12347596740722656, 0.12176076507568359, 0.12155289459228516, 0.12141977691650391, 0.1213685760498047, 0.1208463363647461, 0.12073062133789063, 0.1210040283203125, 0.12351385498046875, 0.12159487915039062, 0.12144640350341797, 0.12170649719238281, 0.12249292755126953, 0.12232396697998046, 0.12123442840576172, 0.12141567993164062, 0.12308889770507812, 0.12184268951416016, 0.1215098876953125, 0.12254208374023437, 0.12142387390136719, 0.12173619079589844, 0.12151910400390625, 0.12153446197509765, 0.12306739044189453, 0.12236492919921875, 0.12201881408691406, 0.12173824310302735, 0.12155699157714844, 0.12209356689453126, 0.12139315032958985, 0.12139520263671875, 0.12346572875976562, 0.12301107025146485, 0.12293836975097656, 0.12192870330810547, 0.12156313323974609, 0.12141056060791015, 0.12382415771484374, 0.12120060729980468, 0.12172902679443359, 0.12317593383789062, 0.12274995422363282, 0.12246323394775391, 0.12149657440185548, 0.12139110565185547, 0.12291788482666016, 0.12123340606689453, 0.12196150207519531, 0.12387631988525391, 0.12276838684082031, 0.12148223876953125, 0.12265267181396484, 0.1214044189453125, 0.12125389099121094, 0.25443635559082034, 0.12313497924804688, 0.12153446197509765, 0.12169830322265625, 0.12147097778320312, 0.12127334594726563, 0.12249088287353516, 0.1221396484375, 0.12196044921875, 0.12192460632324219, 0.12144742584228516, 0.1228226547241211, 0.12275917053222657, 0.12112697601318359, 0.12176787567138672, 0.1217228775024414, 0.12139424133300782, 0.12269356536865235, 0.1223720932006836, 0.12153343963623046, 0.1215478057861328, 0.12162351989746094, 0.12261682891845703, 0.12271308898925781, 0.1215098876953125, 0.12233932495117188, 0.12170342254638672, 0.121238525390625, 0.12113817596435547, 0.12111769866943359, 0.12136653137207032, 0.12165529632568359, 0.1218897933959961, 0.121744384765625, 0.12428594970703125, 0.12338790130615235, 0.12270489501953125, 0.1215129623413086, 0.12190310668945313, 0.12194918060302734, 0.12152114868164063, 0.12212838745117187, 0.12310630035400391, 0.12293427276611328, 0.12179046630859375, 0.12117298889160157, 0.12185906982421875, 0.12301107025146485, 0.1216522216796875, 0.120953857421875, 0.12247654724121093, 0.12268236541748047, 0.12308684539794922, 0.12277657318115234, 0.12169420623779297, 0.12158668518066407, 0.12113510131835938, 0.1211361312866211, 0.1227540512084961, 0.12328550720214844, 0.12320972442626953, 0.12324147033691406, 0.12298242950439453, 0.25302217102050784, 0.1215467529296875, 0.12238642883300781, 0.12331622314453125, 0.12284416198730469, 0.12259225463867188, 0.1225902099609375, 0.12293222045898437, 0.12212531280517579, 0.12088832092285157, 0.12144127655029296, 0.12321894073486328, 0.12279398345947265, 0.12161740875244141, 0.12199628448486328, 0.12130611419677734, 0.12155699157714844, 0.12218367767333985, 0.12152627563476562, 0.12282061004638672, 0.12332236480712891, 0.12172697448730468, 0.1214730224609375, 0.12131743621826171, 0.12206380462646485, 0.12272434997558594, 0.12205158233642578, 0.12277043151855468, 0.12318822479248047, 0.12324352264404297, 0.12150886535644531, 0.12119551849365234, 0.12129280090332031, 0.12111567687988281, 0.1214658203125, 0.12191129302978515, 0.12111974334716796, 0.1216358413696289, 0.12116377258300781, 0.12117708587646485, 0.12161539459228515, 0.1231072998046875, 0.12339097595214844, 0.12407295989990234, 0.12407295989990234, 0.12135116577148437, 0.12155903625488282, 0.12125183868408203, 0.12111154937744141, 0.1215272979736328, 0.12139315032958985, 0.12157440185546875, 0.12285030364990235, 0.12157644653320313, 0.1217976303100586, 0.12137881469726562, 0.12175360107421875, 0.12123750305175782, 0.12148838043212891, 0.12119859313964844, 0.12210585784912109, 0.12211199951171875, 0.12212940979003906, 0.2530303955078125, 0.12125593566894531, 0.12254003143310546, 0.1212252197265625, 0.12286156463623046, 0.12248985290527344, 0.12266598510742187, 0.12236185455322265, 0.12159385681152343, 0.1212590103149414, 0.12165631866455077, 0.12115660858154297, 0.12235878753662109, 0.12386713409423829, 0.1217802276611328, 0.12304691314697265, 0.12239974212646484, 0.12245503997802734, 0.12413849639892578, 0.12299366760253906, 0.12299263763427734, 0.12301107025146485, 0.12142489624023438, 0.12230963134765625, 0.12332236480712891, 0.12175052642822265, 0.12143718719482421, 0.12149247741699219, 0.1212938232421875, 0.12226150512695312, 0.12151193237304687, 0.12162457275390624, 0.12326092529296875, 0.12302438354492187, 0.12452352142333985, 0.12339405059814453, 0.12212735748291016, 0.12276131439208984, 0.12299766540527343, 0.1216911392211914, 0.12096409606933593, 0.12123033905029297, 0.12140953826904297, 0.12123238372802735, 0.12380467224121093, 0.12290969848632813, 0.12300806427001953, 0.12272940826416015, 0.12320665740966796, 0.12299878692626953, 0.12275917053222657, 0.12293119812011719, 0.12306432342529297, 0.12317798614501953, 0.12322406768798828, 0.12217958068847656, 0.1228267822265625, 0.12593353271484375, 0.1234872283935547, 0.12296601867675781, 0.12340940856933594, 0.12347187042236328, 0.12346470642089843, 0.25439436340332033, 0.12301414489746093, 0.1228062744140625, 0.1231431655883789, 0.12312268829345703, 0.12464742279052735, 0.12372991943359375, 0.1234708480834961, 0.12335411071777344, 0.12233523559570313, 0.12217958068847656, 0.12264959716796875, 0.12329881286621094, 0.12281549072265625, 0.12341248321533203, 0.123109375, 0.12312268829345703, 0.12331110382080078, 0.12178943634033203, 0.12153343963623046, 0.12128153228759765, 0.12127027130126954, 0.12139008331298828, 0.12159487915039062, 0.12141875457763672, 0.12109516906738281, 0.12095590209960938, 0.12107469177246094, 0.12224205017089844, 0.1224263687133789, 0.12177005004882813, 0.12150163269042968, 0.12146585845947265, 0.12140953826904297, 0.12113005065917969, 0.12179347229003906, 0.12241817474365234, 0.12146585845947265, 0.1236316146850586, 0.12164096069335938, 0.12152934265136718, 0.12139622497558594, 0.12146688079833984, 0.12121600341796875, 0.12141567993164062, 0.12133478546142579, 0.12168498992919922, 0.1214167709350586, 0.12150675201416015, 0.12333977508544922, 0.12334899139404297, 0.12139826965332032, 0.12266291046142579, 0.12162457275390624, 0.1213675537109375, 0.12148838043212891, 0.123219970703125, 0.12320358276367188, 0.12155494689941407, 0.12110848236083985, 0.12138294219970704, 0.12153238677978516, 0.12358963012695312, 0.25359359741210935, 0.12145458984375, 0.12112588500976562, 0.12119039916992187, 0.12124671936035156, 0.12149964904785156, 0.12119961547851563, 0.12119245147705078, 0.12146688079833984, 0.12120269012451172, 0.12097740936279297, 0.1208616943359375, 0.12124671936035156, 0.12146176147460938, 0.12102349090576171, 0.1214505615234375, 0.12145657348632813, 0.12126924896240235, 0.12115869140625, 0.12115657806396485, 0.12137574768066406, 0.12156928253173828, 0.12129894256591797, 0.1242234878540039, 0.12195740509033202, 0.12128765106201173, 0.12127436828613282, 0.12119142150878906, 0.12085968017578125, 0.12119548797607421, 0.12113203430175781, 0.12153139495849609, 0.12129587554931641, 0.12157234954833984, 0.12127641296386718, 0.12110438537597656, 0.12128870391845703, 0.12145970916748047, 0.12127846527099609, 0.12160921478271484, 0.12154879760742188, 0.1213306884765625, 0.12123238372802735, 0.12116070556640625, 0.12110643005371094, 0.12110031890869141, 0.12115042877197266, 0.12125593566894531, 0.12130406188964844, 0.12108185577392579, 0.12092108917236329, 0.12128562927246093, 0.12140748596191406, 0.12148838043212891, 0.12150784301757812, 0.12131737518310547, 0.12162969970703125, 0.12153958129882812, 0.122281982421875, 0.12196249389648438, 0.12154879760742188, 0.1211883544921875, 0.12114022064208985, 0.25422848510742185, 0.12144435119628906, 0.12121600341796875, 0.12331622314453125, 0.12173926544189453, 0.12114329528808594, 0.1231800308227539, 0.1233438720703125, 0.1232721939086914, 0.12304691314697265, 0.12367155456542969, 0.12309913635253907, 0.12304998779296875, 0.12310630035400391, 0.12295680236816406, 0.12294348907470704, 0.12386918640136718, 0.12152320098876954, 0.12160205078125, 0.121059326171875, 0.12094258880615234, 0.12096717071533203, 0.12145868682861329, 0.12323331451416016, 0.12289020538330078, 0.12314832305908203, 0.1236971206665039, 0.12315955352783203, 0.12311039733886718, 0.12137471771240234, 0.12235366058349609, 0.1222451171875, 0.1222973403930664, 0.12167884826660157, 0.12279296112060546, 0.12148735809326172, 0.12146995544433593, 0.12277760314941406, 0.12156723022460937, 0.1216358413696289, 0.12169728088378906, 0.12312166595458984, 0.12145664215087891, 0.12275507354736329, 0.1232353286743164, 0.12311347198486328, 0.123072509765625, 0.1227171859741211, 0.12111154937744141, 0.12276019287109376, 0.12255232238769531, 0.12235059356689452, 0.12264141082763672, 0.1211514892578125, 0.12123545837402344, 0.12166246032714843, 0.12119347381591797, 0.12130201721191407, 0.12149964904785156, 0.12166246032714843, 0.12141875457763672, 0.12106034851074218, 0.12132147216796875, 0.25745306396484374, 0.12312064361572266, 0.12307865905761718, 0.12312166595458984, 0.12121913909912109, 0.12116063690185547, 0.1212774429321289, 0.12236185455322265, 0.1215498275756836, 0.12132454681396485, 0.12093132781982421, 0.12260454559326171, 0.12201676940917969, 0.12116480255126953, 0.12119449615478516, 0.121206787109375, 0.12122112274169922, 0.12207103729248046, 0.12143001556396485, 0.12134400177001953, 0.12148531341552735, 0.12378009796142578, 0.12321279907226562, 0.12303257751464844, 0.12391117095947266, 0.12308889770507812, 0.12307762908935548, 0.12230758666992188, 0.12212636566162109, 0.12145769500732422, 0.12122207641601562, 0.12131123352050781, 0.12108799743652343, 0.1211156463623047, 0.12251033782958984, 0.12306432342529297, 0.12318310546875, 0.12159795379638672, 0.12140850830078125, 0.12139520263671875, 0.12131839752197265, 0.12117196655273438, 0.12120371246337891, 0.12156723022460937, 0.12148941040039063, 0.12145664215087891, 0.12146892547607421, 0.1215447006225586, 0.12250316619873047, 0.12295782470703125, 0.12217036437988281, 0.12270387268066406, 0.12305919647216797, 0.12331520080566406, 0.12305203247070312, 0.12250521850585938, 0.12303667449951172, 0.12276838684082031, 0.12293427276611328, 0.12293427276611328, 0.12159283447265624, 0.12137062072753907, 0.12141670227050781, 0.2548654022216797, 0.1212416000366211, 0.12151193237304687, 0.12308480072021484, 0.12140338897705077, 0.12150377655029297, 0.12124156951904297, 0.121275390625, 0.12150374603271484, 0.12148838043212891, 0.12146482849121094, 0.12227788543701172, 0.12289740753173828, 0.12292819213867187, 0.12136339569091797, 0.1211822052001953, 0.12195027160644531, 0.12174432373046876, 0.12153241729736328, 0.12099174499511718, 0.12128463745117188, 0.12138390350341798, 0.121385986328125, 0.12258918762207031, 0.1230161895751953, 0.12225638580322265, 0.12275917053222657, 0.12151602935791016, 0.12140850830078125, 0.12134809875488281, 0.12176076507568359, 0.1217791976928711, 0.12272946929931641, 0.12169420623779297, 0.12219187164306641, 0.12145561981201172, 0.12164096069335938, 0.12145254516601563, 0.12175360107421875, 0.12436787414550782, 0.1233602523803711, 0.12332339477539063, 0.1232701416015625, 0.12218675231933594, 0.1233039321899414, 0.12280217742919922, 0.12283596801757812, 0.1221048355102539, 0.12304793548583984, 0.1227509765625, 0.12293023681640625, 0.12202182769775391, 0.12274483489990234, 0.1233070068359375, 0.12290560150146485, 0.12302130889892578, 0.1229496307373047, 0.12324864196777344, 0.12319129943847656, 0.12199935913085938, 0.12175363159179688, 0.12325577545166015, 0.12313600158691407]",tokens/s,8.06546255519598,,,main,False,False,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,databricks/dbrx-base,databricks/dbrx-base,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/databricks/dbrx-base/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aa508-06e716dc7bd8955e0cad2f3c;2b0f0b48-bf56-40d7-b02f-9d9f146c3922) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/databricks/dbrx-base/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like databricks/dbrx-base is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,huggyllama/llama-7b,huggyllama/llama-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,1938.939904,5480.382464,0.0,4833.93536,4503.41376,s,10,5.748573669433593,0.5748573669433594,0.0017742246285147622,0.5745578002929688,0.5774038818359375,0.5778909057617188,0.5782805249023437,"[0.5783779296875, 0.577295654296875, 0.57393994140625, 0.573126220703125, 0.5733052978515625, 0.5724807739257812, 0.5741940307617187, 0.5749215698242187, 0.5756366577148437, 0.5752955932617188]",tokens/s,445.32785821499897,kWh,6.7656206220020495e-06,3.7063076173277895e-06,3.206590528233145e-05,4.2537833521661295e-05,tokens/kWh,6018172.0319545325,MB,1940.045824,5480.382464,0.0,4833.93536,4688.700416,s,10,334.972359375,33.49723593750001,0.0033039843001841955,33.498505859375,33.500188281250004,33.500191796875,33.500194609375,"[33.5001953125, 33.49962890625, 33.4886484375, 33.4951328125, 33.496859375, 33.49850390625, 33.49586328125, 33.5001875, 33.49883203125, 33.4985078125]",tokens/s,1.8807521945257513,kWh,0.0003954924626188515,0.00021676498843461862,0.0018363540061190774,0.0024486114571725475,tokens/kWh,25728.867605948046,,s,629,339.6100241088865,0.5399205470729518,0.06828609533356728,0.531684326171875,0.532189599609375,0.5323219116210938,1.1051400146484376,"[0.5314457397460938, 0.5318953247070313, 0.5313341674804688, 0.5320222778320313, 0.53230078125, 0.5322966918945312, 0.532052978515625, 0.5319219360351563, 0.5317222290039062, 0.5324605712890625, 0.5316915283203125, 0.5319751586914062, 0.5320540161132813, 0.532116455078125, 0.5318584594726562, 0.5316935424804687, 0.5311477661132813, 0.5319874267578125, 0.5314037475585938, 0.5320007934570312, 0.5313106079101563, 0.5315860595703125, 0.5312788696289062, 0.5318707275390625, 0.5317847290039063, 0.5322864379882812, 0.5312890625, 0.5320038452148438, 0.5311918334960938, 0.5317908325195313, 0.53142529296875, 0.5319515991210938, 0.5317119750976562, 0.5320806274414063, 0.5313812255859375, 0.53172119140625, 0.5311549682617187, 0.532105224609375, 0.5318379516601562, 0.5316167602539063, 0.5311580200195313, 0.5316771850585937, 0.5315532836914062, 0.532158447265625, 0.531968017578125, 0.5320550537109375, 0.5314191284179688, 0.5322158203125, 0.5313853149414063, 0.5317304077148437, 0.5311661987304688, 0.5319700317382813, 0.5315399780273438, 0.5316935424804687, 0.531431396484375, 0.5315706787109375, 0.53121435546875, 0.5316372680664062, 0.5313402709960937, 0.531820556640625, 0.5316812744140625, 0.532168701171875, 1.109749755859375, 0.531346435546875, 0.5321318359375, 0.5310955810546875, 0.5315061645507813, 0.531493896484375, 0.5316331787109375, 0.5316638793945313, 0.5315604248046875, 0.531146728515625, 0.5316085815429688, 0.5309306640625, 0.5317273559570312, 0.5312214965820312, 0.5316608276367187, 0.5313013916015625, 0.5319352416992188, 0.5315829467773437, 0.5317447509765625, 0.5317273559570312, 0.53187890625, 0.5315963134765626, 0.53159423828125, 0.5313740844726562, 0.532463623046875, 0.5318870849609375, 0.5320325317382812, 0.5313720092773437, 0.531726318359375, 0.5309389038085938, 0.5318615112304688, 0.5310996704101563, 0.5315245971679687, 0.5312379150390625, 0.5315072021484375, 0.5316474609375, 0.5321410522460938, 0.5322495727539063, 0.5316792602539062, 0.5317243041992188, 0.5318717651367187, 0.531431396484375, 0.5317796020507812, 0.5316239624023438, 0.5317867431640625, 0.5317509155273438, 0.533918701171875, 0.5316608276367187, 0.5317805786132812, 0.5316536254882812, 0.5319649047851562, 0.531599365234375, 0.531768310546875, 0.5318400268554687, 0.5323099975585938, 0.5319802856445313, 0.5322926025390625, 0.5316300659179688, 0.532094970703125, 0.5318133544921875, 0.531968017578125, 0.5319454956054688, 0.5321953125, 1.106044921875, 0.5311734008789063, 0.5317723999023437, 0.5311109008789062, 0.5316505737304688, 0.5316710205078125, 0.5314662475585937, 0.531103759765625, 0.5316044921875, 0.531599365234375, 0.5319229736328125, 0.5315625, 0.5319280395507813, 0.5312767944335938, 0.5315369262695312, 0.5314559936523438, 0.531651611328125, 0.5313187866210938, 0.5317222290039062, 0.5312973022460937, 0.5317273559570312, 0.5311549682617187, 0.531431396484375, 0.531472412109375, 0.532068359375, 0.5310628051757813, 0.5315819702148438, 0.5311876831054687, 0.5319567260742187, 0.5312583618164063, 0.5321011352539062, 0.5313074951171874, 0.5321062622070313, 0.531146728515625, 0.5314539794921875, 0.5311528930664062, 0.5316218872070313, 0.531188720703125, 0.531778564453125, 0.53129931640625, 0.53172021484375, 0.53125634765625, 0.5316557006835938, 0.5312962646484375, 0.5317734375, 0.531356689453125, 0.5318450927734375, 0.5314447631835938, 0.5319024658203125, 0.531715087890625, 0.5318553466796875, 0.5314600830078124, 0.5316638793945313, 0.5312317504882812, 0.5315184936523437, 0.5314949340820313, 0.5321676635742187, 0.5318441162109375, 0.5320765380859375, 0.5313792114257813, 0.5318953247070313, 0.5314150390625, 0.53174169921875, 1.1051417236328125, 0.531535888671875, 0.5323489379882812, 0.5317929077148438, 0.53146728515625, 0.5314027709960938, 0.5317069091796875, 0.5312337646484375, 0.5317662963867188, 0.531146728515625, 0.5315819702148438, 0.531715087890625, 0.5320007934570312, 0.531135498046875, 0.5316823120117188, 0.5312440185546875, 0.531652587890625, 0.5312808837890625, 0.53197314453125, 0.53145703125, 0.5322874755859375, 0.5318850708007813, 0.5318379516601562, 0.5312071533203125, 0.531473388671875, 0.531135498046875, 0.531435546875, 0.5312982788085937, 0.5317243041992188, 0.5311580200195313, 0.5318113403320313, 0.5314652099609375, 0.5317734375, 0.53153076171875, 0.5319024658203125, 0.5314232177734375, 0.5322066040039063, 0.5310986328125, 0.5315245971679687, 0.53191064453125, 0.5318922119140626, 0.53157275390625, 0.5320130615234375, 0.5313024291992188, 0.5326827392578125, 0.5315870971679687, 0.5319966430664063, 0.5312655639648437, 0.5317652587890624, 0.5317560424804687, 0.532094970703125, 0.5314744262695312, 0.5314744262695312, 0.5314232177734375, 0.5318430786132813, 0.531373046875, 0.5317376098632812, 0.5314959106445313, 0.53226904296875, 0.5319168090820312, 0.5326305541992188, 0.5315430297851562, 0.5318215942382812, 1.1051356201171876, 0.5312973022460937, 0.5315819702148438, 0.5314774780273438, 0.531794921875, 0.5313423461914063, 0.5317007446289063, 0.5312061157226563, 0.532337646484375, 0.5311897583007813, 0.5318450927734375, 0.5311968994140625, 0.5319700317382813, 0.5320519409179687, 0.5319915771484375, 0.531267578125, 0.5318911743164062, 0.531177490234375, 0.5319485473632812, 0.5313751220703125, 0.5316874389648437, 0.5312163696289063, 0.5318246459960938, 0.5318041381835937, 0.5321021728515625, 0.5312501831054688, 0.5314898071289063, 0.5313935546875, 0.5318809814453125, 0.5312296752929687, 0.5316013793945312, 0.5316351928710937, 0.5323223266601562, 0.5317406616210938, 0.5318369140625, 0.5311907958984375, 0.5315348510742187, 0.5314457397460938, 0.531673095703125, 0.53165771484375, 0.5315717163085938, 0.531652587890625, 0.5317069091796875, 0.5313167114257813, 0.53161474609375, 0.5311897583007813, 0.5315440673828125, 0.5312553100585937, 0.5322280883789062, 0.5312880859375, 0.5325035400390625, 0.5322506103515625, 0.5319270629882813, 0.5315594482421875, 0.53212158203125, 0.5314180908203125, 0.5320048828125, 0.5329141845703125, 0.5321277465820312, 0.5319035034179688, 0.5321103515625, 0.5317222290039062, 0.5319536743164063, 1.104712646484375, 0.531314697265625, 0.531794921875, 0.531409912109375, 0.5321390380859375, 0.5317406616210938, 0.5324042358398438, 0.5315665893554687, 0.531862548828125, 0.5314232177734375, 0.5315747680664062, 0.5311682739257813, 0.5318154296875, 0.53142529296875, 0.5318092651367188, 0.5314898071289063, 0.5318870849609375, 0.531378173828125, 0.53172119140625, 0.5311692504882812, 0.531430419921875, 0.5313966064453125, 0.531857421875, 0.5313710327148438, 0.5323397216796875, 0.5316075439453125, 0.5323120727539062, 0.5317294311523437, 0.5321543579101562, 0.5312010498046875, 0.5318461303710937, 0.5319188232421875, 0.532210693359375, 0.5314221801757812, 0.5321881713867187, 0.5314324340820312, 0.5315921630859375, 0.5312501831054688, 0.5330370483398438, 0.531240966796875, 0.5315143432617188, 0.5314150390625, 0.5320355834960937, 0.531736572265625, 0.532005859375, 0.5315584106445312, 0.5322998046875, 0.5318276977539063, 0.5317406616210938, 0.53136279296875, 0.5315768432617187, 0.531726318359375, 0.5317335205078125, 0.5318584594726562, 0.5316566772460938, 0.531314697265625, 0.5316884765625, 0.5317488403320313, 0.5318338623046875, 0.5312901000976562, 0.53187890625, 0.5315584106445312, 0.5323212890625, 1.1061114501953124, 0.5317314453125, 0.5319290771484375, 0.531072998046875, 0.5316290283203124, 0.5314488525390625, 0.5314508666992187, 0.5313126220703125, 0.5319659423828125, 0.5314447631835938, 0.5315645141601563, 0.5322833862304688, 0.5317775268554688, 0.5313228759765625, 0.5322045288085937, 0.53157373046875, 0.5320222778320313, 0.53178369140625, 0.5322998046875, 0.5316710205078125, 0.5317069091796875, 0.531219482421875, 0.5317181396484375, 0.5315245971679687, 0.531736572265625, 0.5315369262695312, 0.53231103515625, 0.5319618530273438, 0.5321339111328125, 0.5312696533203125, 0.5317406616210938, 0.53157275390625, 0.5317908325195313, 0.5312593994140625, 0.5319639282226563, 0.5313003540039063, 0.5317816162109374, 0.5313792114257813, 0.531936279296875, 0.5312071533203125, 0.5314293823242188, 0.5310996704101563, 0.531589111328125, 0.5313955688476563, 0.53193115234375, 0.5313556518554687, 0.5319588012695312, 0.531583984375, 0.532021240234375, 0.5315205078125, 0.5315798950195313, 0.5314150390625, 0.5316608276367187, 0.53139453125, 0.5319014282226563, 0.5313863525390625, 0.5320611572265626, 0.5318748168945312, 0.5321339111328125, 0.5318461303710937, 0.5317355346679687, 0.5316792602539062, 0.5318819580078125, 1.1075174560546874, 0.531072021484375, 0.5316382446289063, 0.5312890625, 0.53142529296875, 0.531535888671875, 0.5324411010742187, 0.5314232177734375, 0.5321287841796875, 0.5315972900390625, 0.5322034912109375, 0.5314857177734374, 0.5315389404296875, 0.531357666015625, 0.5317713623046875, 0.5311488037109375, 0.5317294311523437, 0.5311682739257813, 0.531652587890625, 0.5319782104492188, 0.5317386474609375, 0.5313341674804688, 0.5316669311523438, 0.5313074951171874, 0.5316454467773437, 0.5311918334960938, 0.5317662963867188, 0.5315706787109375, 0.5321380004882813, 0.5316188354492187, 0.5322280883789062, 0.5312737426757812, 0.5317304077148437, 0.5311057739257813, 0.5317775268554688, 0.5316751098632813, 0.5317723999023437, 0.5313269653320313, 0.532220947265625, 0.531409912109375, 0.531726318359375, 0.5313054809570312, 0.5317427368164063, 0.5314866943359375, 0.5319229736328125, 0.5314406127929687, 0.532294677734375, 0.5323673706054688, 0.5322874755859375, 0.5322711181640625, 0.5320068969726562, 0.5316044921875, 0.5322977294921875, 0.531583984375, 0.5319116821289063, 0.5315798950195313, 0.532516845703125, 0.5323622436523437, 0.5323622436523437, 0.5315491943359375, 0.5319935913085938, 0.5320017700195312, 0.531962890625, 1.1080765380859374, 0.5319547119140625, 0.5317703857421875, 0.53150927734375, 0.5314652099609375, 0.5311805419921874, 0.53163623046875, 0.5313218383789062, 0.531493896484375, 0.531631103515625, 0.5318482055664062, 0.5316925659179688, 0.5316792602539062, 0.5311713256835937, 0.53180517578125, 0.5314037475585938, 0.531620849609375, 0.5318748168945312, 0.5327401123046875, 0.5320171508789062, 0.5328353271484375, 0.5319864501953125, 0.5319833374023437, 0.5312788696289062, 0.53235302734375, 0.5315451049804687, 0.5321615600585937, 0.5315635375976563, 0.5320089721679687, 0.5313218383789062, 0.5321666259765625, 0.5312501831054688, 0.5315911865234375, 0.5312686157226563, 0.5321390380859375, 0.531409912109375, 0.531768310546875, 0.5311641845703124, 0.5316935424804687, 0.5315829467773437, 0.5322066040039063, 0.5317447509765625, 0.5320120239257813, 0.5312399291992187, 0.5316751098632813, 0.5313863525390625, 0.5317621459960937, 0.531684326171875, 0.5316658935546875, 0.5314549560546875, 0.5319035034179688, 0.531262451171875, 0.5315963134765626, 0.5311375122070312, 0.5317283935546875, 0.5315194702148438, 0.5318932495117188, 0.532084716796875, 0.5324154663085937, 0.5319188232421875, 0.5318717651367187, 0.5313474731445312, 0.5320181884765625, 1.10763720703125, 0.5311539306640625, 0.5317109985351562, 0.5316484985351563, 0.5316484985351563, 0.5312767944335938, 0.5317509155273438, 0.5310279541015624, 0.5314611206054688, 0.5312849731445313, 0.53146728515625, 0.5310648193359375, 0.5322076416015625, 0.53161572265625, 0.5324257202148438, 0.5314857177734374, 0.5316536254882812, 0.5312450561523437, 0.5317069091796875, 0.5315901489257813, 0.5321072387695313, 0.5315020751953125, 0.531820556640625, 0.5321041870117188, 0.5318809814453125, 0.5312317504882812, 0.532041748046875, 0.5315451049804687, 0.5318154296875, 0.5316792602539062, 0.5321062622070313, 0.5322495727539063, 0.5319905395507812, 0.531694580078125, 0.5317621459960937, 0.5315451049804687, 0.5317488403320313, 0.5312747802734376, 0.5321328735351563, 0.531746826171875, 0.5318154296875, 0.5317119750976562, 0.531684326171875, 0.5313290405273438, 0.5314979858398438, 0.5311682739257813, 0.5318164672851563, 0.5316884765625, 0.53195263671875, 0.5318246459960938, 0.532116455078125, 0.5315665893554687, 0.5320376586914063, 0.5314426879882812, 0.5319772338867188, 0.5315123291015625, 0.53215234375, 0.53184716796875, 0.5321748657226563, 0.5318123779296875, 0.5321236572265625, 0.5316618041992187, 0.5318829956054687]",tokens/s,1.8521243642629592,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neox-20b,,cuda,0,42,,,,,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,,MB,1888.280576,15194.390528,0.0,14547.943424,13898.252288,s,10,16.972713500976564,1.6972713500976564,0.0011802396473078822,1.6969771728515626,1.6988580200195311,1.699245782470703,1.6995559924316406,"[1.6969825439453126, 1.6969718017578126, 1.6960919189453125, 1.6958106689453125, 1.6964619140625, 1.69635986328125, 1.6974122314453124, 1.6982171630859375, 1.699633544921875, 1.6987718505859375]",tokens/s,150.83033127570937,kWh,2.0026979611979592e-05,1.097496669508473e-05,9.603446571639917e-05,0.0001270364120234635,tokens/kWh,2015170.2643547354,MB,1888.280576,15194.390528,0.0,14547.943424,14315.97312,s,10,986.8612499999999,98.68612499999999,0.014850841233799409,98.68501953125,98.70599609374999,98.707103515625,98.707989453125,"[98.6797109375, 98.687671875, 98.6747890625, 98.6555546875, 98.6919140625, 98.6783203125, 98.6969609375, 98.7082109375, 98.70575, 98.6823671875]",tokens/s,0.6383876152802637,kWh,0.0011649224273860455,0.0006384806384544572,0.00558826860950019,0.007391671675340693,tokens/kWh,8523.105836826315,,s,629,1000.5712542724599,1.5907333136287136,0.2015788363996436,1.5663739013671876,1.5674868164062499,1.5676927978515625,3.262457841796875,"[1.5671644287109374, 1.5661424560546875, 1.56664013671875, 1.5672913818359375, 1.566525390625, 1.5666636962890625, 1.56687158203125, 1.566798828125, 1.565517822265625, 1.56556494140625, 1.56592333984375, 1.56533251953125, 1.5655465087890625, 1.5658004150390625, 1.565998046875, 1.5670538330078125, 1.5664508056640625, 1.56584033203125, 1.5667855224609375, 1.5674542236328124, 1.567055908203125, 1.5672022705078126, 1.5673907470703126, 1.56657666015625, 1.5663001708984374, 1.566665771484375, 1.5659376220703125, 1.5659520263671876, 1.5661055908203125, 1.566899169921875, 1.5660892333984375, 1.566604248046875, 1.5656611328125, 1.5667138671875, 1.565654052734375, 1.566857177734375, 1.5661802978515624, 1.5664854736328124, 1.5667210693359375, 1.5670538330078125, 1.56666162109375, 1.566857177734375, 1.5659581298828125, 1.565955078125, 1.5662294921875, 1.5661065673828125, 1.566191650390625, 1.5659642333984376, 1.5664527587890624, 1.5659898681640625, 1.5656510009765625, 1.5664803466796875, 1.566614501953125, 1.566317626953125, 1.5665059814453124, 1.565919189453125, 1.5657093505859374, 1.5657308349609376, 1.5665244140625, 1.565981689453125, 1.5660799560546874, 1.56632373046875, 3.263318115234375, 1.5650416259765625, 1.5667425537109374, 1.5671868896484376, 1.5674766845703125, 1.566899169921875, 1.5670477294921874, 1.567247314453125, 1.5669462890625, 1.566482421875, 1.56721044921875, 1.567352783203125, 1.56718603515625, 1.5668541259765625, 1.565632568359375, 1.5655035400390624, 1.565454345703125, 1.566171142578125, 1.5655423583984376, 1.5656632080078126, 1.565739013671875, 1.5655546875, 1.56594384765625, 1.565739990234375, 1.5656090087890624, 1.5654093017578126, 1.5656141357421876, 1.5657379150390625, 1.565550537109375, 1.5656468505859376, 1.566182373046875, 1.56573388671875, 1.5660994873046874, 1.566614501953125, 1.5671705322265626, 1.5668858642578125, 1.5673907470703126, 1.5655372314453124, 1.5656827392578125, 1.567224853515625, 1.5674173583984374, 1.567141845703125, 1.567562744140625, 1.56704150390625, 1.566593994140625, 1.5660902099609375, 1.56824267578125, 1.56689404296875, 1.5673763427734375, 1.5663564453125, 1.566992431640625, 1.567836181640625, 1.5679334716796876, 1.5667803955078126, 1.5659478759765626, 1.566908447265625, 1.566277587890625, 1.5658690185546875, 1.5664803466796875, 1.5664271240234375, 1.5662049560546876, 1.5662171630859374, 1.5660595703125, 3.26247021484375, 1.5656990966796875, 1.5657799072265626, 1.566017578125, 1.5667598876953126, 1.565744140625, 1.566393310546875, 1.56580859375, 1.56586083984375, 1.5654072265625, 1.56617529296875, 1.5659202880859375, 1.5659725341796875, 1.5665029296875, 1.5664486083984375, 1.5655260009765626, 1.5660780029296875, 1.5663062744140626, 1.5667117919921876, 1.566983154296875, 1.5657728271484375, 1.565895751953125, 1.566373779296875, 1.5677890625, 1.566457763671875, 1.5663277587890625, 1.566962646484375, 1.566697509765625, 1.5665489501953125, 1.566899169921875, 1.56623046875, 1.566688232421875, 1.5666124267578125, 1.5657093505859374, 1.566341064453125, 1.5660462646484374, 1.5663482666015625, 1.566123046875, 1.5662838134765624, 1.5664619140625, 1.56621826171875, 1.566017578125, 1.5659622802734374, 1.565760498046875, 1.5661669921875, 1.566396484375, 1.5663912353515625, 1.5682620849609374, 1.5664476318359375, 1.5661475830078124, 1.56608203125, 1.56626123046875, 1.5659765625, 1.5659632568359374, 1.565811767578125, 1.56657763671875, 1.5658956298828124, 1.5657799072265626, 1.5660042724609375, 1.566275634765625, 1.5663533935546874, 1.56682958984375, 1.566255126953125, 3.26236865234375, 1.5657420654296874, 1.5653990478515625, 1.5662427978515625, 1.565697021484375, 1.5651962890625, 1.56592333984375, 1.5652525634765626, 1.5654676513671875, 1.5651666259765624, 1.56598583984375, 1.5653150634765625, 1.5663011474609374, 1.5655894775390624, 1.5655516357421875, 1.565338623046875, 1.5655577392578126, 1.5656806640625, 1.56554345703125, 1.56598779296875, 1.5655843505859375, 1.5654718017578124, 1.5656263427734376, 1.5660400390625, 1.5658824462890626, 1.5657850341796875, 1.5657625732421876, 1.5657196044921875, 1.56554443359375, 1.566350341796875, 1.5656375732421874, 1.5655997314453125, 1.566192626953125, 1.565496337890625, 1.5660390625, 1.566271484375, 1.5658936767578124, 1.565685791015625, 1.5660216064453125, 1.5656141357421876, 1.565828125, 1.566434326171875, 1.5664906005859376, 1.5654307861328125, 1.56860107421875, 1.56617529296875, 1.566096435546875, 1.5658311767578126, 1.5665806884765625, 1.5662591552734375, 1.5666104736328126, 1.56617724609375, 1.566076904296875, 1.5665244140625, 1.566587890625, 1.56634521484375, 1.5658741455078125, 1.5661434326171875, 1.566371826171875, 1.5663533935546874, 1.566328857421875, 1.5666011962890625, 1.5661240234375, 3.263153076171875, 1.56556494140625, 1.565706298828125, 1.5666165771484375, 1.566034912109375, 1.566002197265625, 1.5660831298828124, 1.565319091796875, 1.5665797119140625, 1.5655751953125, 1.5664332275390624, 1.5658896484375, 1.5659744873046875, 1.5658916015625, 1.566130126953125, 1.5668695068359375, 1.5666226806640624, 1.5660933837890625, 1.566509033203125, 1.5661363525390626, 1.5657738037109374, 1.5658199462890625, 1.5663206787109376, 1.5657840576171875, 1.566213134765625, 1.5667742919921874, 1.5661240234375, 1.5659315185546876, 1.565498291015625, 1.5660472412109374, 1.5656407470703124, 1.56573486328125, 1.566254150390625, 1.56568359375, 1.5663575439453126, 1.566123046875, 1.5667579345703124, 1.5663963623046875, 1.5667947998046876, 1.5661865234375, 1.5662919921875, 1.5661158447265624, 1.5670897216796875, 1.56710302734375, 1.567573974609375, 1.5670128173828124, 1.5670006103515626, 1.5675535888671874, 1.568489501953125, 1.5677030029296875, 1.5673804931640625, 1.5675074462890626, 1.5675924072265626, 1.56819970703125, 1.56786181640625, 1.5679241943359374, 1.5680113525390624, 1.5678955078125, 1.56760888671875, 1.5664681396484375, 1.5662049560546876, 1.5662213134765626, 1.5661905517578125, 3.262426025390625, 1.5655628662109375, 1.56493212890625, 1.5661669921875, 1.5657431640625, 1.5655537109375, 1.5660482177734374, 1.5657359619140625, 1.5666063232421874, 1.5661004638671876, 1.5656785888671876, 1.5670230712890625, 1.5668284912109376, 1.5666944580078126, 1.566755859375, 1.566581787109375, 1.5667763671875, 1.5658946533203124, 1.5655372314453124, 1.5659530029296875, 1.565708251953125, 1.5669320068359376, 1.56744091796875, 1.566341064453125, 1.5660360107421876, 1.5659837646484376, 1.5658629150390626, 1.5660892333984375, 1.565887451171875, 1.56593359375, 1.5656898193359374, 1.5661455078125, 1.566086181640625, 1.5658076171875, 1.5659263916015624, 1.5658218994140625, 1.56609033203125, 1.56613525390625, 1.5673641357421875, 1.5665848388671875, 1.5666175537109375, 1.5664681396484375, 1.5664619140625, 1.566224365234375, 1.5664588623046876, 1.5661363525390626, 1.5667579345703124, 1.5667978515625, 1.5667178955078125, 1.5658916015625, 1.5658680419921875, 1.5657728271484375, 1.5670528564453126, 1.567161376953125, 1.5674490966796875, 1.567394775390625, 1.567477783203125, 1.56653369140625, 1.5658792724609376, 1.56665966796875, 1.5675699462890624, 1.56628076171875, 1.5665899658203124, 3.26359765625, 1.5656744384765624, 1.565328369140625, 1.566392333984375, 1.565917236328125, 1.566066650390625, 1.567247314453125, 1.56674560546875, 1.567057861328125, 1.566623779296875, 1.5667916259765624, 1.5669647216796876, 1.56680810546875, 1.5664117431640625, 1.566813232421875, 1.5673231201171876, 1.566376953125, 1.5661844482421876, 1.567056884765625, 1.567025146484375, 1.566773193359375, 1.5668214111328125, 1.5673548583984376, 1.568668701171875, 1.567552490234375, 1.566614501953125, 1.566899169921875, 1.567447021484375, 1.5675135498046875, 1.5669381103515625, 1.5667794189453126, 1.567139892578125, 1.5667916259765624, 1.566656494140625, 1.566287841796875, 1.5665806884765625, 1.5664046630859374, 1.5664854736328124, 1.56649169921875, 1.56577587890625, 1.5662652587890624, 1.5666688232421875, 1.5665521240234375, 1.56607080078125, 1.5668214111328125, 1.5663809814453125, 1.5665684814453125, 1.5666226806640624, 1.566630859375, 1.5663360595703124, 1.566734375, 1.5659100341796874, 1.566339111328125, 1.5664271240234375, 1.5667056884765624, 1.5661854248046876, 1.5660308837890624, 1.56668310546875, 1.5665264892578126, 1.5664742431640626, 1.5667752685546874, 1.5658568115234375, 1.566477294921875, 3.26474853515625, 1.5660062255859375, 1.5659560546875, 1.5659263916015624, 1.5657471923828126, 1.5661905517578125, 1.5662623291015625, 1.5657779541015624, 1.5659345703125, 1.5662991943359375, 1.5659427490234374, 1.565487060546875, 1.566613525390625, 1.56559765625, 1.56596533203125, 1.566318603515625, 1.566983154296875, 1.5657728271484375, 1.5661322021484374, 1.566572509765625, 1.5666114501953126, 1.5659100341796874, 1.56704150390625, 1.56611376953125, 1.5677716064453124, 1.5673876953125, 1.5676068115234374, 1.5671654052734374, 1.5670262451171875, 1.5657738037109374, 1.5667547607421874, 1.5669217529296875, 1.5676558837890624, 1.56761083984375, 1.567972412109375, 1.567614990234375, 1.5675115966796875, 1.5676702880859374, 1.56752587890625, 1.567363037109375, 1.567826904296875, 1.567635498046875, 1.5676488037109375, 1.5662724609375, 1.5680296630859376, 1.567309814453125, 1.567677490234375, 1.567458251953125, 1.5668489990234375, 1.565865966796875, 1.566993408203125, 1.5673487548828124, 1.567581298828125, 1.566005126953125, 1.5679825439453126, 1.567009765625, 1.566161865234375, 1.5676138916015625, 1.56758935546875, 1.5667568359375, 1.5663258056640625, 1.566224365234375, 1.5670302734375, 3.26736279296875, 1.5667579345703124, 1.5664261474609376, 1.5665531005859374, 1.5668193359375, 1.566329833984375, 1.5663739013671876, 1.566214111328125, 1.5671173095703126, 1.565612060546875, 1.5653099365234375, 1.565854736328125, 1.5658486328125, 1.5660902099609375, 1.566256103515625, 1.5667547607421874, 1.5674736328125, 1.56900244140625, 1.5678065185546874, 1.567677490234375, 1.5673057861328126, 1.5676702880859374, 1.5678648681640626, 1.566625732421875, 1.566234619140625, 1.5665244140625, 1.565811767578125, 1.566467041015625, 1.5663380126953126, 1.5663995361328125, 1.56670458984375, 1.566562255859375, 1.56580859375, 1.5658076171875, 1.567499267578125, 1.566665771484375, 1.5675023193359374, 1.5676190185546874, 1.5666585693359374, 1.5654676513671875, 1.5674849853515624, 1.5671900634765625, 1.567220703125, 1.5676558837890624, 1.5666514892578125, 1.5663369140625, 1.5666441650390626, 1.566841796875, 1.566866455078125, 1.56685107421875, 1.567130615234375, 1.5664896240234376, 1.5665152587890625, 1.5667691650390625, 1.56735693359375, 1.5668797607421876, 1.5669544677734375, 1.565843505859375, 1.5657564697265625, 1.5662509765625, 1.5679661865234376, 1.567494140625, 1.5670753173828125, 3.26488671875, 1.56588232421875, 1.5666165771484375, 1.56659912109375, 1.5660400390625, 1.565919189453125, 1.5653775634765625, 1.5651685791015626, 1.5665255126953126, 1.5655526123046875, 1.56554443359375, 1.565875244140625, 1.565812744140625, 1.56569091796875, 1.56594384765625, 1.566340087890625, 1.565750244140625, 1.5661793212890625, 1.5670374755859375, 1.5664691162109374, 1.5666708984375, 1.5669124755859376, 1.5670743408203125, 1.565885498046875, 1.567392822265625, 1.56645068359375, 1.567561767578125, 1.5667598876953126, 1.566843994140625, 1.566246826171875, 1.56635546875, 1.5664219970703126, 1.565961181640625, 1.5663524169921874, 1.5675914306640626, 1.56609326171875, 1.5661405029296875, 1.5667711181640624, 1.566587890625, 1.5662694091796876, 1.566982177734375, 1.566482421875, 1.566496826171875, 1.5665531005859374, 1.5663370361328126, 1.5661158447265624, 1.566561279296875, 1.566202880859375, 1.566286865234375, 1.5659674072265626, 1.5661629638671875, 1.5660155029296876, 1.566634033203125, 1.5668284912109376, 1.5664476318359375, 1.5665438232421875, 1.5667864990234375, 1.5668736572265625, 1.56661865234375, 1.5671204833984376, 1.566720947265625, 1.5667332763671875, 1.566159912109375]",tokens/s,0.6286408862079101,,,main,False,False,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-4B,Qwen/Qwen1.5-4B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,3203.223552,5128.060928,0.0,4481.613824,4276.256768,s,10,3.1983958740234373,0.31983958740234375,0.001538131138588488,0.31976911926269536,0.32137343139648433,0.3217927795410156,0.3221282580566406,"[0.31961990356445313, 0.3212218933105469, 0.3187176513671875, 0.3199183349609375, 0.3183829040527344, 0.3169226989746094, 0.31905938720703125, 0.32106072998046875, 0.32128024291992185, 0.32221212768554686]",tokens/s,800.4012326277909,kWh,3.750147018581629e-06,2.054515901244258e-06,1.6957018773937002e-05,2.276168169376289e-05,tokens/kWh,11246972.14574214,MB,3203.223552,5128.060928,0.0,4481.613824,4465.662976,s,10,186.97800781249998,18.69780078125,0.017296888320766722,18.693263671875002,18.720369140625,18.729755859375,18.737265234375002,"[18.718283203125, 18.689103515625, 18.695333984375, 18.678984375, 18.739142578125, 18.694673828125, 18.691853515625, 18.6855234375, 18.701947265625, 18.683162109375]",tokens/s,3.369380214125283,kWh,0.00022053018352016812,0.00012086672241809653,0.0009756264419588673,0.001317023347897132,tokens/kWh,47835.14286257035,,s,629,189.55983416748046,0.3013669859578386,0.038019298519587436,0.29649102783203124,0.2979925964355469,0.29949728393554687,0.6154752392578126,"[0.29722113037109377, 0.2962104187011719, 0.2994401550292969, 0.3005357666015625, 0.30041702270507814, 0.2999255065917969, 0.30030029296875, 0.3005870056152344, 0.30005453491210937, 0.3003166809082031, 0.2982318115234375, 0.29764404296875, 0.29678591918945313, 0.2977320861816406, 0.2972508239746094, 0.2988697509765625, 0.29594931030273436, 0.29598825073242185, 0.29595953369140626, 0.2961408081054687, 0.29663333129882813, 0.2959800415039063, 0.2984028015136719, 0.29736856079101565, 0.29788876342773435, 0.2962769775390625, 0.2963660888671875, 0.2973767700195312, 0.2967070617675781, 0.2959728698730469, 0.2965258178710938, 0.29611212158203126, 0.2976409606933594, 0.2969599914550781, 0.2967398376464844, 0.2962391052246094, 0.29641522216796873, 0.2959902648925781, 0.29629336547851565, 0.29671832275390625, 0.29604147338867187, 0.2959872131347656, 0.2960855102539062, 0.2966855773925781, 0.29605990600585935, 0.29604147338867187, 0.29675726318359374, 0.29607012939453126, 0.2964500427246094, 0.29650942993164064, 0.2962022399902344, 0.2960516967773438, 0.29594830322265625, 0.296783935546875, 0.296870849609375, 0.29710232543945314, 0.29665484619140625, 0.29788363647460936, 0.29600152587890627, 0.29640191650390624, 0.29616229248046877, 0.2960711669921875, 0.6185267333984374, 0.29601278686523436, 0.29589199829101565, 0.29706646728515623, 0.29594418334960937, 0.29600460815429686, 0.29592166137695314, 0.2975078430175781, 0.2961551208496094, 0.29602200317382815, 0.2963240966796875, 0.2960445556640625, 0.2958970947265625, 0.2962135009765625, 0.29608447265625, 0.29607730102539065, 0.29639883422851565, 0.29659750366210935, 0.29723443603515626, 0.29676031494140626, 0.2970972290039062, 0.296911865234375, 0.29686373901367186, 0.29711358642578123, 0.29724774169921875, 0.29758157348632813, 0.29762457275390625, 0.2963056640625, 0.29703271484375, 0.29604864501953126, 0.2962616271972656, 0.296153076171875, 0.2962769775390625, 0.2961919860839844, 0.2968739929199219, 0.29633843994140624, 0.2965504150390625, 0.29617355346679686, 0.29583154296875, 0.2962042846679688, 0.2970060729980469, 0.2965278625488281, 0.29644491577148435, 0.29606298828125, 0.29650225830078125, 0.2965186462402344, 0.3018721313476562, 0.29692620849609375, 0.2963681640625, 0.29674798583984374, 0.296806396484375, 0.29622067260742185, 0.2960076904296875, 0.2968924255371094, 0.2960343017578125, 0.2965831604003906, 0.2970480651855469, 0.29702047729492187, 0.296217529296875, 0.29626266479492186, 0.29842022705078125, 0.29761843872070315, 0.29686373901367186, 0.6160445556640625, 0.2960199890136719, 0.2965903015136719, 0.2964684753417969, 0.29644390869140624, 0.2965688171386719, 0.2976983032226562, 0.29753759765625, 0.29599737548828126, 0.29653094482421877, 0.2970838928222656, 0.2961479797363281, 0.29596878051757813, 0.29587353515625, 0.2960445556640625, 0.2966005859375, 0.2975467529296875, 0.2962995300292969, 0.2958847961425781, 0.2957864990234375, 0.2960732116699219, 0.295973876953125, 0.2962944030761719, 0.29851339721679687, 0.295984130859375, 0.29607730102539065, 0.2961203308105469, 0.2960425109863281, 0.29593295288085936, 0.29733270263671874, 0.29703067016601564, 0.29734808349609376, 0.29631283569335937, 0.296816650390625, 0.29681048583984376, 0.296489990234375, 0.29625653076171876, 0.29661590576171876, 0.29632000732421876, 0.2966937561035156, 0.2978652038574219, 0.2965299072265625, 0.2963240966796875, 0.29751806640625, 0.29830859375, 0.297481201171875, 0.29719039916992185, 0.29953536987304685, 0.29737368774414064, 0.29692108154296876, 0.296585205078125, 0.296052734375, 0.29637939453125, 0.29625753784179687, 0.29655654907226564, 0.29799423217773435, 0.2968268737792969, 0.29685043334960937, 0.29692108154296876, 0.29714944458007814, 0.29722930908203127, 0.2967623596191406, 0.297270263671875, 0.6168934326171875, 0.29688626098632814, 0.296279052734375, 0.29670297241210936, 0.2962923583984375, 0.29618795776367185, 0.2966793518066406, 0.2960650329589844, 0.29615924072265626, 0.29594418334960937, 0.2973972473144531, 0.29612338256835935, 0.2971064453125, 0.2982297668457031, 0.29685043334960937, 0.2960855102539062, 0.296089599609375, 0.29661489868164065, 0.2965350341796875, 0.29597695922851563, 0.2961705017089844, 0.29597695922851563, 0.29592276000976564, 0.29643975830078123, 0.29586328125, 0.29596466064453125, 0.2959923095703125, 0.2958981018066406, 0.29643777465820315, 0.2959923095703125, 0.29621759033203127, 0.2963650512695313, 0.29609677124023437, 0.29788980102539064, 0.2977914733886719, 0.29713613891601565, 0.2964951171875, 0.29727642822265626, 0.2966794128417969, 0.2964142150878906, 0.2965821533203125, 0.2960855102539062, 0.2957701416015625, 0.2966097717285156, 0.29938388061523435, 0.29615609741210935, 0.29603640747070314, 0.2962728271484375, 0.29596978759765624, 0.2960455627441406, 0.296300537109375, 0.29595040893554686, 0.29649197387695314, 0.29635379028320313, 0.296342529296875, 0.29667330932617186, 0.2967008972167969, 0.29659442138671877, 0.29601690673828124, 0.29651455688476563, 0.29645925903320314, 0.2973665466308594, 0.29616226196289064, 0.615267333984375, 0.29617767333984374, 0.29771673583984376, 0.29707980346679685, 0.298039306640625, 0.297238525390625, 0.2971115417480469, 0.29701837158203126, 0.29682891845703124, 0.2972641296386719, 0.2971668395996094, 0.29752627563476564, 0.2972252197265625, 0.29701632690429686, 0.29692620849609375, 0.29619302368164063, 0.2964234619140625, 0.2987468566894531, 0.296131591796875, 0.2958981018066406, 0.29623501586914064, 0.29586944580078123, 0.295846923828125, 0.29605377197265625, 0.29625753784179687, 0.29595443725585935, 0.29606195068359376, 0.29590631103515624, 0.29597491455078123, 0.29599948120117187, 0.296310791015625, 0.2964664306640625, 0.29622885131835935, 0.2963630065917969, 0.29594317626953126, 0.2961397705078125, 0.29812017822265624, 0.3001466979980469, 0.3003924560546875, 0.3000289306640625, 0.29983538818359373, 0.29985791015625, 0.29986712646484376, 0.29981594848632814, 0.30056344604492186, 0.29986407470703125, 0.2999111633300781, 0.29782833862304686, 0.2960773315429687, 0.2960137939453125, 0.29763687133789063, 0.2979921875, 0.2981877746582031, 0.2967510986328125, 0.29627288818359376, 0.29777716064453125, 0.2970634155273438, 0.30057470703125, 0.2982738037109375, 0.2991800231933594, 0.30042315673828124, 0.2963568725585938, 0.2961418151855469, 0.6139617309570312, 0.29619406127929687, 0.29618585205078124, 0.29586431884765624, 0.296079345703125, 0.29643060302734375, 0.2977740783691406, 0.2977832946777344, 0.2964633483886719, 0.29604147338867187, 0.2960609130859375, 0.29627084350585936, 0.29602822875976564, 0.2961315307617188, 0.29630160522460935, 0.29599432373046874, 0.2961469421386719, 0.296531982421875, 0.2960977783203125, 0.2960373840332031, 0.2965381164550781, 0.2960506896972656, 0.29609063720703127, 0.2962063293457031, 0.2960169677734375, 0.29643975830078123, 0.2966312866210937, 0.2963077087402344, 0.2960865173339844, 0.2961162109375, 0.2960281677246094, 0.2961766357421875, 0.296352783203125, 0.2976133117675781, 0.29756414794921876, 0.29727435302734373, 0.29644595336914065, 0.2974146423339844, 0.30141543579101565, 0.2976788330078125, 0.2980526123046875, 0.2980894775390625, 0.29788058471679685, 0.2963497009277344, 0.29696920776367186, 0.29795635986328123, 0.29711566162109376, 0.29751296997070314, 0.29756314086914065, 0.2974320678710938, 0.29744332885742186, 0.29654324340820315, 0.29747711181640624, 0.2984744873046875, 0.29649203491210935, 0.2962442321777344, 0.2966384582519531, 0.2963189697265625, 0.29589913940429685, 0.2962176208496094, 0.2958837585449219, 0.29590936279296876, 0.29634048461914064, 0.6155560913085938, 0.29629541015625, 0.2960916442871094, 0.2964561767578125, 0.29645208740234374, 0.2960373840332031, 0.2963875732421875, 0.2963620300292969, 0.2962861633300781, 0.29651251220703123, 0.29747915649414064, 0.2981519470214844, 0.29662823486328127, 0.2976133117675781, 0.29637530517578126, 0.29608038330078124, 0.29605682373046877, 0.2960496520996094, 0.29625140380859377, 0.29587966918945313, 0.2968821716308594, 0.29635174560546873, 0.29616537475585936, 0.3003904113769531, 0.2970214538574219, 0.2966312866210937, 0.29685556030273436, 0.29636813354492186, 0.29714739990234373, 0.29626266479492186, 0.29595135498046876, 0.2962196350097656, 0.29621759033203127, 0.29663333129882813, 0.2961408081054687, 0.2965729370117188, 0.2961131591796875, 0.2963865661621094, 0.296479736328125, 0.296900634765625, 0.2964776611328125, 0.296848388671875, 0.29701937866210937, 0.29931622314453127, 0.29786627197265625, 0.2962636413574219, 0.29623806762695315, 0.2966210632324219, 0.2961336364746094, 0.2963711853027344, 0.29625140380859377, 0.2967357482910156, 0.29616845703125, 0.2963599853515625, 0.29687493896484374, 0.296300537109375, 0.2967142333984375, 0.29815499877929685, 0.29702349853515625, 0.29747610473632813, 0.296627197265625, 0.29662515258789063, 0.2977822570800781, 0.6183147583007812, 0.29672344970703124, 0.296764404296875, 0.29664154052734376, 0.2967080993652344, 0.29716583251953127, 0.29694155883789064, 0.2967244873046875, 0.2972221374511719, 0.2974392395019531, 0.2972866516113281, 0.2969518127441406, 0.2982625427246094, 0.296279052734375, 0.29735015869140624, 0.2961131591796875, 0.2980966491699219, 0.29623806762695315, 0.29628826904296873, 0.2956605529785156, 0.29607012939453126, 0.2962001953125, 0.29612442016601564, 0.29620736694335936, 0.29724774169921875, 0.2960086975097656, 0.297548828125, 0.2967602844238281, 0.29649102783203124, 0.2963619384765625, 0.2962708740234375, 0.29639984130859376, 0.2968350830078125, 0.2963292236328125, 0.296099853515625, 0.29622784423828125, 0.29627093505859375, 0.296197021484375, 0.2974996337890625, 0.2962104187011719, 0.2960639953613281, 0.29599948120117187, 0.2959902648925781, 0.2965739440917969, 0.2961817626953125, 0.2968320007324219, 0.29708084106445315, 0.2968320007324219, 0.29710540771484373, 0.2960558166503906, 0.2963138427734375, 0.29637326049804685, 0.29592166137695314, 0.2962083740234375, 0.29609573364257813, 0.2962114562988281, 0.2961203308105469, 0.29699481201171873, 0.29659442138671877, 0.29610906982421875, 0.29717401123046877, 0.29671218872070315, 0.2965350341796875, 0.61785498046875, 0.29675213623046875, 0.29679409790039063, 0.29726925659179687, 0.29639578247070314, 0.29744537353515627, 0.2977556457519531, 0.29687704467773435, 0.2974320678710938, 0.297416748046875, 0.29766244506835937, 0.29807000732421873, 0.2972119140625, 0.2973388671875, 0.2967439270019531, 0.2972999572753906, 0.2976030578613281, 0.2986280822753906, 0.29623602294921875, 0.29600152587890627, 0.2964695739746094, 0.2963936767578125, 0.2961868896484375, 0.29702349853515625, 0.29616024780273437, 0.29621148681640624, 0.2964879150390625, 0.29709210205078124, 0.2963056640625, 0.2973767700195312, 0.2969722900390625, 0.29783346557617185, 0.2965718994140625, 0.2977576904296875, 0.2961418151855469, 0.2967091064453125, 0.2973644714355469, 0.29671218872070315, 0.29666201782226564, 0.296838134765625, 0.2965801086425781, 0.2962391052246094, 0.296089599609375, 0.2963804016113281, 0.29701119995117187, 0.2961469421386719, 0.29617767333984374, 0.296384521484375, 0.296369140625, 0.29728460693359376, 0.2963015747070312, 0.2967838745117187, 0.29621554565429686, 0.2963097534179687, 0.29686373901367186, 0.2963947448730469, 0.2962995300292969, 0.29706649780273436, 0.2967142333984375, 0.29672857666015623, 0.2966722412109375, 0.2982359008789062, 0.2979543151855469, 0.6189783325195313, 0.2962237548828125, 0.2969938049316406, 0.29632614135742186, 0.2961541748046875, 0.2965380554199219, 0.2963343505859375, 0.29643777465820315, 0.29634866333007814, 0.2962872314453125, 0.2963630065917969, 0.29629644775390623, 0.29665484619140625, 0.296310791015625, 0.29620120239257813, 0.29598513793945314, 0.29617767333984374, 0.2959667053222656, 0.2962749328613281, 0.2967930908203125, 0.29673779296875, 0.29692825317382815, 0.29633331298828125, 0.29688934326171873, 0.2962821044921875, 0.29612954711914063, 0.2987386779785156, 0.2962083740234375, 0.296374267578125, 0.2962135009765625, 0.29653402709960935, 0.2968299560546875, 0.2963097534179687, 0.2981970520019531, 0.2961192321777344, 0.2962135009765625, 0.2961069946289063, 0.295920654296875, 0.2961459655761719, 0.29606500244140627, 0.29621453857421876, 0.29600564575195315, 0.29583566284179685, 0.2993377380371094, 0.29691802978515625, 0.2961561584472656, 0.2958428039550781, 0.2965329895019531, 0.29672549438476564, 0.29595547485351564, 0.2960772705078125, 0.29593496704101563, 0.2966548767089844, 0.2969466552734375, 0.2978447265625, 0.296627197265625, 0.29650433349609373, 0.2975867004394531, 0.2982461853027344, 0.29639166259765626, 0.29643161010742186, 0.29704702758789064, 0.29672344970703124]",tokens/s,3.3182134958203435,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-1.8B,Qwen/Qwen1.5-1.8B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,1868.201984,3349.676032,0.0,2703.228928,2578.238464,s,10,1.410495346069336,0.1410495346069336,0.0017689689177857657,0.14056432342529296,0.14309543151855467,0.14402269134521484,0.14476449920654297,"[0.144949951171875, 0.1412518768310547, 0.1397279357910156, 0.13956182861328126, 0.13982972717285155, 0.13892410278320313, 0.13987677001953125, 0.14288937377929686, 0.1413212127685547, 0.14216256713867187]",tokens/s,1814.9652227737006,kWh,1.6439394082552122e-06,9.00797525271931e-07,6.874183740083411e-06,9.418920673610555e-06,tokens/kWh,27179334.96533712,MB,1868.201984,3349.676032,0.0,2703.228928,2667.098624,s,10,83.44500195312501,8.3445001953125,0.03302115511867202,8.3336259765625,8.37593193359375,8.403280419921876,8.425159208984375,"[8.33203515625, 8.335076171875, 8.333888671875, 8.3144501953125, 8.32368359375, 8.3150146484375, 8.3698544921875, 8.43062890625, 8.3570068359375, 8.33336328125]",tokens/s,7.549882979856609,kWh,9.832901556366757e-05,5.38916025818177e-05,0.0003962753054459165,0.0005484959235914018,tokens/kWh,114859.55918777513,,s,629,84.57831109619147,0.13446472352335676,0.016756456539684012,0.13199974060058595,0.1339588592529297,0.13428590698242188,0.27203026611328124,"[0.131915771484375, 0.13208883666992188, 0.13340570068359375, 0.13261517333984374, 0.1320437774658203, 0.13189222717285157, 0.13191372680664062, 0.13180621337890625, 0.13191270446777345, 0.1318174743652344, 0.13188607788085938, 0.13165977478027344, 0.13220761108398438, 0.13279539489746095, 0.132642822265625, 0.1319086151123047, 0.13183590698242187, 0.13183692932128907, 0.13191372680664062, 0.1319710693359375, 0.13185536193847655, 0.13166387939453125, 0.13191888427734375, 0.13172015380859375, 0.1317058563232422, 0.13175704956054687, 0.1318830108642578, 0.13170687866210937, 0.1325506591796875, 0.13195468139648436, 0.13218412780761718, 0.13190444946289062, 0.1321625671386719, 0.13238067626953126, 0.13378150939941405, 0.13233255004882813, 0.13199667358398437, 0.13365965270996094, 0.13226495361328125, 0.13205401611328124, 0.13234483337402345, 0.13211033630371094, 0.13187481689453126, 0.13180210876464843, 0.13335859680175782, 0.13234176635742187, 0.13278311157226563, 0.13272679138183593, 0.1324042205810547, 0.13250457763671875, 0.13250662231445312, 0.13336883544921874, 0.13242880249023437, 0.13245030212402345, 0.13245234680175783, 0.13265408325195313, 0.1332162628173828, 0.13244825744628907, 0.1322782745361328, 0.13222093200683593, 0.13194239807128907, 0.1318461456298828, 0.2743060607910156, 0.13452493286132813, 0.13292338562011718, 0.13269094848632812, 0.13221580505371094, 0.13191372680664062, 0.13204888916015625, 0.1319393310546875, 0.13207244873046875, 0.13193624877929688, 0.1332346954345703, 0.13184307861328126, 0.13165977478027344, 0.13174169921875, 0.1319331817626953, 0.13257011413574218, 0.13245234680175783, 0.13182669067382813, 0.13220249938964843, 0.1319772186279297, 0.13190451049804688, 0.13176934814453126, 0.13176832580566405, 0.13202943420410157, 0.131842041015625, 0.13187992858886718, 0.13245132446289062, 0.1319014434814453, 0.13174374389648438, 0.13188096618652342, 0.1329971160888672, 0.1325875244140625, 0.1321922607421875, 0.13208473205566407, 0.1319751739501953, 0.13210829162597656, 0.132173828125, 0.1320202178955078, 0.13244108581542968, 0.1323520050048828, 0.13193624877929688, 0.1318707275390625, 0.13215335083007812, 0.1319833526611328, 0.13199871826171874, 0.13258546447753905, 0.1363056640625, 0.13410917663574218, 0.132347900390625, 0.13224557495117187, 0.13346809387207031, 0.1320499267578125, 0.1319823303222656, 0.13196595764160157, 0.13193113708496093, 0.13199974060058595, 0.13258444213867188, 0.13302787780761718, 0.13199562072753906, 0.13191885375976561, 0.13196493530273437, 0.1318461456298828, 0.13196800231933595, 0.27239935302734375, 0.1318348846435547, 0.13182975769042968, 0.13207449340820313, 0.13185638427734375, 0.1317724151611328, 0.13244415283203126, 0.13195263671875, 0.1320099792480469, 0.13171302795410156, 0.13188505554199217, 0.13182054138183594, 0.13208575439453124, 0.1338101806640625, 0.13198439025878905, 0.13201919555664063, 0.13190348815917968, 0.13199871826171874, 0.13165977478027344, 0.13169357299804688, 0.13205708312988282, 0.13223014831542967, 0.13255679321289063, 0.13336166381835937, 0.1323274230957031, 0.1321175079345703, 0.13262745666503906, 0.13243597412109376, 0.13281074523925782, 0.13219737243652344, 0.13270425415039064, 0.1319403839111328, 0.13225782775878905, 0.13328172302246094, 0.13259266662597657, 0.13259056091308594, 0.13216152954101562, 0.1320622100830078, 0.13242469787597655, 0.1341696014404297, 0.13195263671875, 0.13167718505859374, 0.13170278930664062, 0.13178675842285156, 0.13180108642578126, 0.13196902465820312, 0.13176319885253907, 0.1331517791748047, 0.1319188232421875, 0.13173248291015624, 0.13183795166015624, 0.13228440856933593, 0.1320273895263672, 0.13176934814453126, 0.13183897399902345, 0.13179600524902343, 0.13171708679199218, 0.13180108642578126, 0.13352243041992187, 0.1350635528564453, 0.13252915954589845, 0.13388493347167968, 0.1324451904296875, 0.2718658447265625, 0.1323663330078125, 0.1318041534423828, 0.1317375946044922, 0.13224755859375, 0.13172837829589842, 0.13183692932128907, 0.13180825805664062, 0.13196493530273437, 0.13201612854003905, 0.13178060913085937, 0.13186866760253907, 0.13176422119140624, 0.13189222717285157, 0.13185331726074218, 0.13174476623535156, 0.13189631652832032, 0.1319833526611328, 0.13183180236816405, 0.13173965454101563, 0.13178880310058594, 0.13191474914550783, 0.13178163146972657, 0.13190553283691406, 0.132674560546875, 0.13195161437988281, 0.13180313110351563, 0.13180108642578126, 0.13176524353027344, 0.13174578857421876, 0.1317232666015625, 0.13185740661621093, 0.1315359344482422, 0.13187271118164062, 0.13189427185058594, 0.13176217651367186, 0.13183999633789062, 0.131842041015625, 0.13198028564453124, 0.1318461456298828, 0.13185740661621093, 0.13189222717285157, 0.1318656005859375, 0.13177548217773438, 0.13433549499511718, 0.1318778839111328, 0.13178163146972657, 0.1320099792480469, 0.13189529418945312, 0.13173043823242186, 0.13199974060058595, 0.1317908477783203, 0.13167718505859374, 0.1318707580566406, 0.1319085693359375, 0.1320079345703125, 0.1317375946044922, 0.1329776611328125, 0.13234278869628907, 0.13224960327148438, 0.13266841125488282, 0.13238578796386719, 0.13234994506835937, 0.2727505798339844, 0.13219532775878906, 0.1322977294921875, 0.1321318359375, 0.13251072692871094, 0.13448908996582032, 0.13275852966308593, 0.1344153594970703, 0.13294796752929688, 0.1335828552246094, 0.13194137573242187, 0.1318604736328125, 0.13178469848632812, 0.13185023498535156, 0.13183282470703125, 0.13243084716796874, 0.13204582214355468, 0.13287628173828125, 0.13193522644042968, 0.13247999572753907, 0.13198028564453124, 0.13178880310058594, 0.1318225860595703, 0.13189529418945312, 0.1317969970703125, 0.1317898254394531, 0.13174783325195313, 0.13180825805664062, 0.1317611541748047, 0.13181951904296876, 0.13173043823242186, 0.1318901824951172, 0.1316822967529297, 0.1317908477783203, 0.13169664001464843, 0.1318656005859375, 0.1333289031982422, 0.1319536590576172, 0.13187890625, 0.13186457824707032, 0.13181951904296876, 0.1317969970703125, 0.1317580871582031, 0.13171200561523438, 0.13176934814453126, 0.13164851379394532, 0.13176524353027344, 0.13194239807128907, 0.13188812255859375, 0.13191270446777345, 0.1322117156982422, 0.13192909240722656, 0.13183999633789062, 0.1317693786621094, 0.13185020446777343, 0.13187583923339843, 0.1317611541748047, 0.13403237915039062, 0.13205503845214844, 0.13186151123046874, 0.13183795166015624, 0.13185536193847655, 0.13172531127929688, 0.2711296081542969, 0.13194137573242187, 0.131852294921875, 0.13176217651367186, 0.13171609497070313, 0.13177548217773438, 0.13163827514648438, 0.1318164520263672, 0.1318707275390625, 0.13187277221679689, 0.13203762817382814, 0.13187174987792968, 0.13186968994140624, 0.13186866760253907, 0.13186972045898437, 0.13196592712402344, 0.1320099792480469, 0.1333217315673828, 0.13214207458496094, 0.13328793334960937, 0.13255679321289063, 0.13287628173828125, 0.1319505920410156, 0.13182566833496093, 0.13165158081054687, 0.13177650451660156, 0.13287423706054688, 0.13261311340332033, 0.13195878601074218, 0.1316505584716797, 0.1316495361328125, 0.13174989318847657, 0.13171302795410156, 0.13187686157226564, 0.13190553283691406, 0.13175296020507812, 0.13174989318847657, 0.13191474914550783, 0.13393820190429687, 0.13187989807128905, 0.13174681091308593, 0.1316864013671875, 0.13167205810546875, 0.13172019958496095, 0.13174783325195313, 0.1315952606201172, 0.13168435668945314, 0.13165362548828125, 0.13327769470214842, 0.13195161437988281, 0.1319024963378906, 0.1318594207763672, 0.13180313110351563, 0.13170381164550782, 0.13171507263183593, 0.13190451049804688, 0.13232640075683594, 0.13192294311523436, 0.13172940063476563, 0.1317406768798828, 0.13174578857421876, 0.13170381164550782, 0.13169561767578125, 0.2720942077636719, 0.1323653106689453, 0.1319772186279297, 0.13174887084960937, 0.1318461456298828, 0.13169049072265626, 0.13180825805664062, 0.13170994567871094, 0.13199974060058595, 0.13190348815917968, 0.13172940063476563, 0.13172735595703125, 0.13171612548828124, 0.13159523010253907, 0.13188914489746092, 0.13179904174804688, 0.13198130798339844, 0.13195578002929687, 0.13164845275878906, 0.13183590698242187, 0.13194752502441406, 0.13176524353027344, 0.13180621337890625, 0.13325619506835937, 0.1319086456298828, 0.13182666015625, 0.13163929748535155, 0.13185125732421876, 0.1320396728515625, 0.13438668823242186, 0.13385317993164061, 0.13466111755371094, 0.13423411560058593, 0.133897216796875, 0.1338357696533203, 0.1337620849609375, 0.13375177001953126, 0.1338419189453125, 0.1338470458984375, 0.13377127075195314, 0.13373440551757812, 0.1337507781982422, 0.13362892150878905, 0.13362687683105468, 0.133718017578125, 0.1337139129638672, 0.13433139038085937, 0.1335930938720703, 0.13229158020019532, 0.13321932983398438, 0.1331998748779297, 0.1333289031982422, 0.13277183532714842, 0.13196185302734376, 0.13300531005859376, 0.13356031799316406, 0.13304013061523437, 0.13298892211914062, 0.1336432647705078, 0.13445120239257813, 0.13455258178710938, 0.1344040985107422, 0.13423922729492188, 0.27732583618164064, 0.1339084777832031, 0.13382144165039062, 0.1346867218017578, 0.13416653442382812, 0.13392076110839843, 0.13295001220703126, 0.13354495239257813, 0.13410508728027343, 0.1338173370361328, 0.13408869934082032, 0.1340200958251953, 0.1339043884277344, 0.13345074462890624, 0.1340712890625, 0.133653564453125, 0.1348218231201172, 0.13366886901855468, 0.13357466125488282, 0.1335029754638672, 0.1333729248046875, 0.13342617797851564, 0.1340518341064453, 0.1341563262939453, 0.13418496704101562, 0.13429244995117187, 0.13391769409179688, 0.13419314575195312, 0.13331149291992186, 0.13398121643066407, 0.1340333709716797, 0.13403135681152345, 0.13413682556152343, 0.13416447448730467, 0.1341071319580078, 0.13356748962402343, 0.13255885314941407, 0.1329449005126953, 0.13408767700195312, 0.13382655334472657, 0.13446556091308592, 0.134451171875, 0.1342740478515625, 0.13432524108886718, 0.13397196960449217, 0.133928955078125, 0.13388800048828126, 0.13408154296875, 0.1339463653564453, 0.13414707946777343, 0.1341204528808594, 0.1346938934326172, 0.13456895446777345, 0.13381427001953125, 0.13382553100585937, 0.13402316284179688, 0.13395558166503907, 0.13448602294921874, 0.13234994506835937, 0.13215437316894532, 0.13213081359863282, 0.13221376037597657, 0.13244825744628907, 0.27383807373046876, 0.13208677673339844, 0.1318748779296875, 0.13186143493652344, 0.1319342041015625, 0.1316986846923828, 0.13210009765625, 0.13194956970214844, 0.1318338623046875, 0.13179904174804688, 0.13189427185058594, 0.1324390411376953, 0.1331261444091797, 0.13414399719238282, 0.13440205383300782, 0.13397196960449217, 0.13388394165039064, 0.13371900939941406, 0.13379379272460937, 0.1336944580078125, 0.13435395812988282, 0.133012451171875, 0.1330391082763672, 0.13337496948242186, 0.13211033630371094, 0.1332316131591797, 0.1331374053955078, 0.1329827880859375, 0.13283839416503906, 0.13333401489257812, 0.1330401611328125, 0.1329510040283203, 0.13266636657714845, 0.132642822265625, 0.1325813751220703, 0.1328547821044922, 0.13313536071777343, 0.13244313049316406, 0.13259365844726562, 0.13253018188476562, 0.1329459228515625, 0.13234585571289062, 0.13299200439453124, 0.13250764465332032, 0.13244210815429688, 0.13175296020507812, 0.1317580871582031, 0.1318338623046875, 0.1318666229248047, 0.1317580871582031, 0.13186358642578125, 0.13192803955078125, 0.13181849670410156, 0.13414501953125, 0.13427609252929687, 0.13285580444335937, 0.13174476623535156, 0.1318656005859375, 0.13187277221679689, 0.13214002990722656, 0.13324185180664064, 0.13186151123046874, 0.13166592407226563, 0.27445761108398437, 0.13168025207519532, 0.1318492126464844, 0.13196595764160157, 0.13203353881835939, 0.13289573669433594, 0.13374771118164064, 0.13365350341796875, 0.13377433776855469, 0.13357772827148437, 0.13372621154785155, 0.1335900115966797, 0.1336494140625, 0.13360946655273437, 0.1339095001220703, 0.13365863037109374, 0.13385317993164061, 0.1333053436279297, 0.13203660583496094, 0.13213388061523437, 0.13183897399902345, 0.132063232421875, 0.13191372680664062, 0.13187992858886718, 0.13165875244140626, 0.1317560272216797, 0.13234994506835937, 0.13201510620117188, 0.13194342041015625, 0.13241856384277345, 0.13182566833496093, 0.1321871337890625, 0.13201715087890625, 0.13212979125976562, 0.13185331726074218, 0.13186764526367187, 0.1317969970703125, 0.13172940063476563, 0.13185433959960938, 0.1317959747314453, 0.13192909240722656, 0.1318144073486328, 0.1319772186279297, 0.13174783325195313, 0.13176422119140624, 0.13171916198730468, 0.13182464599609375, 0.1319086151123047, 0.13198745727539063, 0.13206431579589845, 0.13207244873046875, 0.1318757781982422, 0.13179391479492186, 0.13187379455566406, 0.13178060913085937, 0.1318113250732422, 0.13175196838378905, 0.13170889282226564, 0.13177754211425782, 0.13175091552734375, 0.1317611541748047, 0.13186968994140624, 0.1329213409423828]",tokens/s,7.43689477654188,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-14B,Qwen/Qwen-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 550, in from_pretrained + model_class = get_class_from_dynamic_module( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 498, in get_class_from_dynamic_module + final_module = get_cached_module_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 323, in get_cached_module_file + modules_needed = check_imports(resolved_module_file) + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 181, in check_imports + raise ImportError( +ImportError: This modeling file requires the following packages that were not found in your environment: transformers_stream_generator. Run `pip install transformers_stream_generator` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-7.5B,facebook/xglm-7.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,2170.855424,7227.31008,0.0,6580.862976,6226.036224,s,10,5.757403198242186,0.5757403198242188,0.0009922788511109175,0.5756337280273438,0.5767303344726562,0.5773405700683594,0.5778287585449219,"[0.5763906860351562, 0.5754805908203126, 0.5742902221679688, 0.5758182373046875, 0.5751873168945313, 0.574752685546875, 0.575786865234375, 0.5751510620117187, 0.5779508056640625, 0.5765947265625]",tokens/s,444.6449053249566,kWh,6.784125003549789e-06,3.717420948032668e-06,3.101488592300367e-05,4.151643187458613e-05,tokens/kWh,6166233.1862557735,MB,2170.855424,7227.31008,0.0,6580.862976,6478.658048,s,10,336.90312500000005,33.6903125,0.0038578817961585758,33.690537109375,33.6951453125,33.6965375,33.69765125,"[33.6891484375, 33.6979296875, 33.68431640625, 33.69015234375, 33.69138671875, 33.68830859375, 33.690921875, 33.6910078125, 33.6948359375, 33.6851171875]",tokens/s,1.8699737498724598,kWh,0.00039786197528243067,0.0002180628312197768,0.001800652746076575,0.0024165775525787823,tokens/kWh,26069.926840448938,,s,629,341.54916522216826,0.5430034423245914,0.06841490320905963,0.5347440795898437,0.5353432861328125,0.5355950317382813,1.1101828955078124,"[0.5346764526367187, 0.534593505859375, 0.53607421875, 0.5346488037109375, 0.5349376220703125, 0.5348045043945312, 0.5348710327148437, 0.534361083984375, 0.5351505737304687, 0.5344501953125, 0.5342750854492188, 0.5343682861328125, 0.5348853759765625, 0.5342218017578125, 0.5351854248046874, 0.5346846923828125, 0.5351168212890625, 0.5340159912109375, 0.5345996704101562, 0.5340569458007812, 0.5348679809570313, 0.5342392578125, 0.5346948852539063, 0.5346427001953125, 0.5350113525390625, 0.5345361938476563, 0.534276123046875, 0.5349284057617187, 0.5341531982421875, 0.5347072143554688, 0.534413330078125, 0.5346447143554688, 0.5343323974609375, 0.534677490234375, 0.5346836547851562, 0.5348843383789063, 0.5341747436523437, 0.53469287109375, 0.5349488525390625, 0.5349376220703125, 0.5349048461914062, 0.5350922241210937, 0.5349293823242187, 0.5348905029296875, 0.5348187866210937, 0.5352499389648437, 0.5351372680664063, 0.5351874389648438, 0.5344020385742188, 0.5347307739257813, 0.534319091796875, 0.5348607788085937, 0.53452490234375, 0.5352724609375, 0.5355673828125, 0.53477685546875, 0.5353441162109375, 0.5347553100585938, 0.535225341796875, 0.5340364990234375, 0.5348362426757812, 0.53452392578125, 1.1103118896484374, 0.5353318481445313, 0.5341173706054687, 0.5345740966796875, 0.5343201293945312, 0.5346703491210938, 0.5341255493164062, 0.5354373168945312, 0.534645751953125, 0.5345771484375, 0.5340641479492187, 0.535025634765625, 0.5342791748046875, 0.5352877807617188, 0.5341102294921874, 0.5349335327148438, 0.53408154296875, 0.5348935546875, 0.5341614379882812, 0.5348187866210937, 0.5348075561523438, 0.5348945922851562, 0.5341737060546875, 0.5358602294921875, 0.5343109130859375, 0.53522021484375, 0.5344286499023437, 0.5350133666992187, 0.5340877075195313, 0.5351638793945312, 0.5342811889648438, 0.535573486328125, 0.5346611328125, 0.5351854248046874, 0.5346078491210937, 0.53532568359375, 0.5343160400390625, 0.53486181640625, 0.5339535522460938, 0.5352847290039062, 0.5343467407226562, 0.5350491943359375, 0.5341419677734375, 0.5348751220703125, 0.534150146484375, 0.5347573852539063, 0.5344307250976562, 0.5354004516601563, 0.5344286499023437, 0.5352489013671875, 0.5350963134765625, 0.5354454956054687, 0.537660400390625, 0.5361489868164062, 0.53553662109375, 0.5356922607421875, 0.5343662109375, 0.5351168212890625, 0.535394287109375, 0.5363414916992187, 0.5349365844726562, 0.5353850708007812, 0.5346826171875, 1.110350830078125, 0.5346856689453126, 0.5343385620117187, 0.5345433349609375, 0.5344020385742188, 0.5345515747070313, 0.5340989379882812, 0.5348003540039062, 0.5341388549804688, 0.53479833984375, 0.5344071655273438, 0.5343672485351563, 0.5352960205078126, 0.5340405883789062, 0.5349048461914062, 0.5344491577148438, 0.535067626953125, 0.5343518676757812, 0.5349970092773437, 0.534128662109375, 0.5347153930664063, 0.53432421875, 0.5345413208007812, 0.5351311645507812, 0.5347737426757813, 0.534719482421875, 0.5343641357421876, 0.5350553588867187, 0.5344020385742188, 0.5347225341796875, 0.5345587158203124, 0.5351331787109375, 0.5342013549804687, 0.5347911376953125, 0.5345372314453125, 0.5347225341796875, 0.5342412719726563, 0.5348792114257812, 0.5342689208984375, 0.5345525512695313, 0.5345115966796875, 0.5345474853515625, 0.53495703125, 0.53488232421875, 0.535014404296875, 0.5344368896484375, 0.53495703125, 0.534287353515625, 0.5348126831054687, 0.5351044921875, 0.535436279296875, 0.53448193359375, 0.5350891723632812, 0.5339913940429688, 0.5349908447265626, 0.5347307739257813, 0.5351055297851562, 0.5346611328125, 0.5345126342773437, 0.5350236206054687, 0.5345863647460938, 0.5352509155273437, 0.5350174560546875, 1.1105382080078126, 0.5355950317382813, 0.5344696044921875, 0.5347164306640625, 0.5345341186523438, 0.5346948852539063, 0.5343866577148437, 0.535103515625, 0.5347359008789062, 0.5351536865234375, 0.5343518676757812, 0.5350891723632812, 0.5348259887695312, 0.5349007568359375, 0.5342996215820313, 0.5354127197265625, 0.5349365844726562, 0.5348239135742188, 0.5341798095703125, 0.534992919921875, 0.534329345703125, 0.5348515625, 0.5342095336914062, 0.53524072265625, 0.5343549194335937, 0.534803466796875, 0.534345703125, 0.534719482421875, 0.5346047973632813, 0.5349662475585938, 0.5341337890625, 0.5350051879882812, 0.5342566528320313, 0.5347942504882812, 0.5347379150390625, 0.5350942993164063, 0.5344153442382813, 0.5352182006835937, 0.534846435546875, 0.5348956298828125, 0.5342689208984375, 0.5350553588867187, 0.5349601440429688, 0.53513525390625, 0.534561767578125, 0.535309326171875, 0.5342945556640625, 0.5351065673828125, 0.5344860229492188, 0.53475634765625, 0.534240234375, 0.5348341674804687, 0.5342648315429688, 0.534993896484375, 0.5347952880859375, 0.5356380004882813, 0.534824951171875, 0.5348976440429688, 0.5344174194335938, 0.5349017333984375, 0.5350840454101562, 0.5351629028320313, 0.5344470825195312, 1.1098358154296875, 0.5345259399414063, 0.5348362426757812, 0.534181884765625, 0.5347102661132812, 0.534091796875, 0.5347164306640625, 0.534054931640625, 0.5361285400390625, 0.5355950317382813, 0.5359862060546875, 0.5356984252929687, 0.5359247436523438, 0.5353123779296876, 0.5349468383789062, 0.53486181640625, 0.5354373168945312, 0.5347164306640625, 0.5351342163085937, 0.5345413208007812, 0.5349837036132813, 0.53418701171875, 0.5347440795898437, 0.5342218017578125, 0.5345996704101562, 0.5342843017578125, 0.5351854248046874, 0.5345218505859375, 0.5348915405273438, 0.5342116088867187, 0.5350205688476563, 0.534297607421875, 0.534667236328125, 0.5343784790039062, 0.5350491943359375, 0.5344635009765625, 0.535109619140625, 0.5345679321289063, 0.5353430786132812, 0.5343150024414063, 0.5344952392578125, 0.5345228881835937, 0.5347676391601562, 0.53477783203125, 0.534740966796875, 0.5348997192382813, 0.5346795654296875, 0.53429248046875, 0.534645751953125, 0.5347973022460938, 0.5346099243164063, 0.5341378784179688, 0.5352919311523437, 0.5342832641601563, 0.5348894653320313, 0.53421875, 0.5347788696289062, 0.534635498046875, 0.5353768920898437, 0.5347850341796875, 0.5351485595703125, 0.5345423583984374, 0.5350031127929687, 1.1105545654296876, 0.5355919189453126, 0.5343006591796875, 0.5348720703125, 0.53427197265625, 0.5345126342773437, 0.5340037231445313, 0.535520263671875, 0.534739990234375, 0.535130126953125, 0.5343836059570313, 0.5350440673828125, 0.5346948852539063, 0.5351239624023437, 0.5344174194335938, 0.5351260375976562, 0.5345014038085938, 0.5347993774414063, 0.534582275390625, 0.5357168579101562, 0.5343109130859375, 0.5353820190429688, 0.5346990356445313, 0.5351629028320313, 0.5347451171875, 0.5354332275390625, 0.5348679809570313, 0.5348095703125, 0.5343733520507813, 0.5351966552734375, 0.5344102172851563, 0.5353021240234375, 0.5340569458007812, 0.534813720703125, 0.5341829223632812, 0.5347389526367188, 0.5341173706054687, 0.5348935546875, 0.5345310668945312, 0.535056396484375, 0.5342576904296875, 0.5349898071289062, 0.5346007080078125, 0.5354352416992187, 0.53414501953125, 0.5347205200195313, 0.5340692749023438, 0.5347993774414063, 0.5341737060546875, 0.53488330078125, 0.5342361450195312, 0.5347962646484375, 0.53441845703125, 0.53486181640625, 0.5344409790039063, 0.5348505859375, 0.5344778442382813, 0.5346652221679687, 0.5343754272460938, 0.5350758666992188, 0.5344163818359375, 0.5350686645507813, 0.5343856811523438, 1.1107747802734376, 0.5348556518554688, 0.5345413208007812, 0.5346017456054688, 0.5344235229492188, 0.5347625122070313, 0.534202392578125, 0.5346734008789062, 0.534513671875, 0.5345167236328126, 0.5345884399414063, 0.5348331298828125, 0.534603759765625, 0.5347123413085938, 0.535151611328125, 0.5338849487304688, 0.5345567016601562, 0.53408154296875, 0.5351454467773438, 0.5341777954101562, 0.5348731079101563, 0.534076416015625, 0.5348925170898438, 0.5345361938476563, 0.534635498046875, 0.534403076171875, 0.5345423583984374, 0.5348433837890625, 0.5344429931640625, 0.5346826171875, 0.5351076049804687, 0.5349652709960937, 0.5341634521484375, 0.5351157836914062, 0.534329345703125, 0.5346119384765625, 0.5346826171875, 0.5350000610351563, 0.5348444213867187, 0.5346211547851563, 0.5343887329101562, 0.5350768432617188, 0.53532568359375, 0.5349631958007812, 0.5353236694335938, 0.534540283203125, 0.5354454956054687, 0.5352796020507813, 0.5357506713867187, 0.5357987670898438, 0.5363568725585938, 0.5354977416992187, 0.5355079956054688, 0.5346631469726563, 0.5348648681640625, 0.53481982421875, 0.5350850830078125, 0.5346867065429688, 0.534488037109375, 0.5345198364257813, 0.5341511840820312, 0.5348761596679688, 0.5343672485351563, 1.1098511962890625, 0.5344020385742188, 0.5343488159179688, 0.5351004028320312, 0.5345003662109375, 0.5340426025390625, 0.5349335327148438, 0.534129638671875, 0.5345115966796875, 0.534624267578125, 0.5344778442382813, 0.5346898193359375, 0.5347532958984375, 0.5341439819335938, 0.5348208618164062, 0.5342740478515625, 0.534593505859375, 0.5343160400390625, 0.5345218505859375, 0.5341788330078125, 0.5349027709960937, 0.5340282592773438, 0.5352099609375, 0.5348515625, 0.5344050903320312, 0.534667236328125, 0.534846435546875, 0.5345792236328125, 0.5350031127929687, 0.5352366333007812, 0.5349683227539063, 0.5348577270507813, 0.5350656127929687, 0.5350123291015625, 0.5349437255859375, 0.5353164672851562, 0.535109619140625, 0.5345955810546875, 0.5346232299804687, 0.5350543212890625, 0.5351209106445313, 0.5352017822265625, 0.5351526489257813, 0.534813720703125, 0.5347174682617187, 0.53507275390625, 0.5347962646484375, 0.535278564453125, 0.5350348510742188, 0.5344235229492188, 0.5348864135742187, 0.5350287475585938, 0.5347123413085938, 0.5351557006835937, 0.5351854248046874, 0.534824951171875, 0.5352898559570313, 0.5345700073242188, 0.5350369262695313, 0.5348126831054687, 0.5350420532226563, 0.5343068237304688, 0.5352109985351563, 1.1140484619140625, 0.5357752075195312, 0.5347020874023437, 0.5347215576171875, 0.5340579833984375, 0.5347225341796875, 0.5339064331054687, 0.5348648681640625, 0.5343109130859375, 0.5346764526367187, 0.5340579833984375, 0.534751220703125, 0.5342074584960937, 0.53477685546875, 0.5345331420898437, 0.5362677612304687, 0.5344603881835938, 0.535562255859375, 0.534560791015625, 0.5345679321289063, 0.53437646484375, 0.5348444213867187, 0.5341480712890625, 0.5348433837890625, 0.5341439819335938, 0.5348792114257812, 0.5344849853515625, 0.5353707275390625, 0.534392822265625, 0.5349959716796875, 0.534276123046875, 0.5353820190429688, 0.5345740966796875, 0.5356973876953125, 0.534592529296875, 0.5353021240234375, 0.5345812377929687, 0.5350471801757812, 0.5342730102539063, 0.5356431274414063, 0.5344461059570312, 0.5352222900390625, 0.534319091796875, 0.5352243041992187, 0.5342218017578125, 0.5354874877929687, 0.5344635009765625, 0.5353492431640625, 0.5346590576171875, 0.5352969970703125, 0.5345413208007812, 0.5364439086914062, 0.5349744873046876, 0.5354383544921875, 0.5349171142578125, 0.5353850708007812, 0.5352386474609375, 0.5347727661132813, 0.5347440795898437, 0.5347891235351563, 0.53429248046875, 0.5350348510742188, 0.5342208251953126, 1.1111966552734376, 0.5346058349609375, 0.5340743408203125, 0.5344491577148438, 0.5340753784179687, 0.5344573364257813, 0.5351997680664062, 0.5346652221679687, 0.5347676391601562, 0.5348843383789063, 0.5340805053710938, 0.5348229370117188, 0.5342904052734375, 0.5346806030273438, 0.5341122436523438, 0.5346744384765625, 0.5341951904296875, 0.53463037109375, 0.5346017456054688, 0.5344931640625, 0.5348495483398438, 0.5341675415039062, 0.534192138671875, 0.5346129760742188, 0.5353421020507813, 0.5345730590820312, 0.5347010498046875, 0.5343364868164062, 0.53553564453125, 0.5343969116210937, 0.534887451171875, 0.5349539794921875, 0.5349130249023437, 0.5349970092773437, 0.5343682861328125, 0.534813720703125, 0.5341798095703125, 0.5349908447265626, 0.5340886840820313, 0.5348864135742187, 0.53427197265625, 0.5349468383789062, 0.5341091918945312, 0.5351629028320313, 0.5349376220703125, 0.5348556518554688, 0.53496728515625, 0.5342669067382813, 0.5347440795898437, 0.53432421875, 0.53492529296875, 0.5341470947265625, 0.5350174560546875, 0.5343016967773437, 0.5352642822265625, 0.5343908081054688, 0.535689208984375, 0.535041015625, 0.5345894165039062, 0.5347891235351563, 0.5347962646484375, 0.5358428344726562, 0.5355755615234375]",tokens/s,1.8416089513521527,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-12b,EleutherAI/pythia-12b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,,,MB,1585.569792,9676.783616,0.0,9030.336512,8583.573504,s,10,9.573078491210937,0.9573078491210938,0.0009060771854647526,0.9570158996582031,0.9584730651855469,0.958728140258789,0.9589322003173828,"[0.9589832153320312, 0.956114501953125, 0.95637109375, 0.9565009765625, 0.9567921142578125, 0.9570035400390625, 0.9570282592773437, 0.95793017578125, 0.957938232421875, 0.9584163818359375]",tokens/s,267.41658938139295,kWh,1.1295311258296775e-05,6.189367993647465e-06,5.412797259527177e-05,7.161265184721602e-05,tokens/kWh,3574787.323141869,MB,1585.569792,9676.783616,0.0,9030.336512,8872.967168,s,10,567.96033984375,56.796033984375,0.005905783414625482,56.796150390625,56.802221875,56.803097265625,56.803797578125,"[56.78812109375, 56.796453125, 56.795734375, 56.80109765625, 56.800734375, 56.79584765625, 56.80202734375, 56.80397265625, 56.7909765625, 56.785375]",tokens/s,1.1092323808618707,kWh,0.0006704276505549146,0.00036745356669733154,0.0031837375343617275,0.004221618751613973,tokens/kWh,14923.185561442368,,s,629,575.6730700073241,0.9152195071658572,0.113749717069217,0.901496826171875,0.9019905639648437,0.9021755371093749,1.8586116552734375,"[0.9012899780273438, 0.9015009155273438, 0.9013851928710938, 0.9014036254882812, 0.9017272338867187, 0.9014722290039062, 0.9017364501953125, 0.9016832275390625, 0.901676025390625, 0.9014937744140625, 0.9013677978515625, 0.90115380859375, 0.9009315795898437, 0.9009326171875, 0.9006499633789062, 0.9008291625976562, 0.9008558349609375, 0.9008486328125, 0.9011026000976563, 0.9014824829101562, 0.901602294921875, 0.901296142578125, 0.9016278686523438, 0.9010196533203125, 0.901043212890625, 0.9009264526367188, 0.9010216674804687, 0.9008496704101563, 0.901128173828125, 0.9008875732421875, 0.901443603515625, 0.9013667602539063, 0.9012459716796875, 0.90106982421875, 0.9016514282226562, 0.9009028930664063, 0.9010933837890625, 0.9009028930664063, 0.9010616455078125, 0.9009520874023438, 0.901349365234375, 0.9014415283203125, 0.901707763671875, 0.9015367431640625, 0.9014681396484375, 0.9018613891601562, 0.901622802734375, 0.9018931274414063, 0.901707763671875, 0.9014057006835937, 0.901602294921875, 0.90210302734375, 0.9018941650390625, 0.9015654296875, 0.9017302856445313, 0.9016248168945312, 0.9020457153320313, 0.9013043212890625, 0.9014906616210937, 0.9013822021484375, 0.9013165283203125, 0.901349365234375, 1.86075439453125, 0.901201904296875, 0.9012777099609375, 0.9009407958984375, 0.9011107788085938, 0.9012059936523438, 0.901411865234375, 0.9011742553710937, 0.9014906616210937, 0.9008762817382813, 0.9011650390625, 0.9011548461914063, 0.9017159423828125, 0.9009899291992187, 0.9010339965820312, 0.9014087524414063, 0.901095458984375, 0.9012418212890625, 0.9011476440429688, 0.90099609375, 0.901496826171875, 0.9016145629882812, 0.9013197021484375, 0.9020405883789062, 0.9015439453125, 0.9012941284179687, 0.9013165893554688, 0.9014087524414063, 0.9009602661132813, 0.9014149169921875, 0.901086181640625, 0.901359619140625, 0.9010995483398437, 0.9015562133789062, 0.9015613403320313, 0.9021737060546875, 0.9016637573242188, 0.9020088500976563, 0.9014948120117188, 0.90166064453125, 0.901233642578125, 0.901607421875, 0.90140673828125, 0.9016708984375, 0.9019443359375, 0.9020333862304688, 0.90393701171875, 0.9017435913085937, 0.90208154296875, 0.9017467041015625, 0.901749755859375, 0.9017620239257812, 0.9014108276367188, 0.901612548828125, 0.9014630126953125, 0.901897216796875, 0.9016043701171875, 0.9021522216796874, 0.9015429077148438, 0.9017200927734375, 0.9018798217773437, 0.901327880859375, 0.9020098266601563, 1.8579189453125, 0.9009674072265625, 0.9019514770507813, 0.9013217163085937, 0.9013043212890625, 0.9010616455078125, 0.901369873046875, 0.9010882568359375, 0.9010647583007813, 0.9008700561523437, 0.901254150390625, 0.9011005249023437, 0.9012029418945312, 0.901781494140625, 0.9017036743164063, 0.9013206787109375, 0.901180419921875, 0.9015664672851562, 0.9012275390625, 0.901060546875, 0.9013350219726562, 0.9012828369140625, 0.9011179809570312, 0.902530029296875, 0.901454833984375, 0.9017108764648437, 0.9014998779296876, 0.9013156127929688, 0.9013903198242188, 0.901664794921875, 0.9012408447265625, 0.9015879516601563, 0.9012008666992187, 0.9010565185546875, 0.9010083618164062, 0.9016893310546875, 0.9017681884765625, 0.9018941650390625, 0.9012459716796875, 0.9014087524414063, 0.9018050537109376, 0.9015992431640625, 0.9015572509765625, 0.9015726318359375, 0.9014558715820312, 0.9014138793945312, 0.901760009765625, 0.9032969970703125, 0.9014159545898438, 0.9015982055664062, 0.90126953125, 0.9023529052734375, 0.901634033203125, 0.9018787841796875, 0.9014353637695313, 0.9016453247070313, 0.9016279296875, 0.901738525390625, 0.9017559204101563, 0.902043701171875, 0.9015582275390625, 0.9019248657226563, 0.9020723266601562, 1.8581309814453124, 0.9010073852539062, 0.9013156127929688, 0.9008977661132812, 0.9008128051757812, 0.9016401977539062, 0.9013319702148438, 0.9011097412109375, 0.9010739135742187, 0.9013370971679687, 0.9015664672851562, 0.9017160034179688, 0.9023712768554687, 0.9015736083984375, 0.9013422241210938, 0.901080078125, 0.9018327026367188, 0.90148046875, 0.9011190185546875, 0.901212158203125, 0.9018624267578125, 0.9014783935546875, 0.9012479858398438, 0.9013688354492188, 0.90138623046875, 0.9015767211914063, 0.90112109375, 0.901693359375, 0.901707763671875, 0.9020232543945312, 0.901571533203125, 0.9019771118164063, 0.901349365234375, 0.9013718872070312, 0.9014671630859376, 0.9014220581054687, 0.9012838134765625, 0.9017538452148437, 0.9016350708007812, 0.9019105224609375, 0.9021767578125, 0.901813232421875, 0.9018296508789062, 0.9015695190429688, 0.903024658203125, 0.902118408203125, 0.9014558715820312, 0.9017098388671875, 0.901375, 0.901796875, 0.90182861328125, 0.9018388671875, 0.901644287109375, 0.9020845947265625, 0.9015941162109375, 0.9019043579101562, 0.9019207763671875, 0.901602294921875, 0.9014589233398438, 0.9020272827148438, 0.9018408813476563, 0.9020088500976563, 0.9018572998046875, 1.8591446533203124, 0.9011109008789062, 0.9014906005859376, 0.9012500610351563, 0.9013309326171876, 0.9015132446289063, 0.9010267944335938, 0.90169140625, 0.9015234375, 0.9012705078125, 0.90149169921875, 0.90165869140625, 0.90100830078125, 0.9009920043945312, 0.9015848999023437, 0.9015347290039063, 0.9014989013671875, 0.9024532470703125, 0.9021880493164063, 0.9016944580078124, 0.9015951538085938, 0.9013145751953126, 0.90147021484375, 0.9014896850585937, 0.9010872192382813, 0.9012111206054687, 0.9012377319335938, 0.9011046142578125, 0.90134423828125, 0.9021696166992188, 0.9015510864257813, 0.9017293090820313, 0.9014609985351563, 0.9018982543945312, 0.9016985473632813, 0.9015664672851562, 0.9012459716796875, 0.9016145629882812, 0.9012357177734375, 0.9017845458984375, 0.9017640991210938, 0.9023355102539062, 0.9016002807617187, 0.9013688354492188, 0.9013800659179687, 0.9019535522460937, 0.9013340454101563, 0.9017763671875, 0.9015776977539063, 0.90165966796875, 0.9015429077148438, 0.9014620361328125, 0.9018777465820312, 0.9020886840820312, 0.9014589233398438, 0.9016187133789062, 0.901876708984375, 0.903267333984375, 0.9016299438476563, 0.9016832275390625, 0.901592041015625, 0.9018306274414063, 0.9016535034179688, 1.8594969482421875, 0.90172314453125, 0.901623779296875, 0.9014261474609375, 0.9010974731445313, 0.9016514282226562, 0.9012612915039062, 0.9012162475585938, 0.9010811157226563, 0.900874267578125, 0.9016196899414063, 0.9018163452148438, 0.9015050048828125, 0.9010811157226563, 0.9016350708007812, 0.9010995483398437, 0.9011435546875, 0.9012008666992187, 0.9010196533203125, 0.90096337890625, 0.9009110717773438, 0.9015501098632812, 0.9013790893554687, 0.901444580078125, 0.9009868774414063, 0.9019638061523437, 0.9014251708984375, 0.9013463134765625, 0.9010083618164062, 0.9014886474609375, 0.9014528198242188, 0.9012612915039062, 0.902129638671875, 0.9023948974609375, 0.9013688354492188, 0.9017241821289063, 0.90169140625, 0.9015562133789062, 0.9032264404296875, 0.90172509765625, 0.9013514404296875, 0.9015767211914063, 0.9013986206054687, 0.9023374633789063, 0.9012766723632812, 0.90123876953125, 0.9009448852539063, 0.9015643920898437, 0.9012295532226563, 0.9011148681640625, 0.9011896362304688, 0.9011199951171875, 0.9018316650390625, 0.9017835693359375, 0.9014640502929687, 0.9016084594726562, 0.90189208984375, 0.901802978515625, 0.9020620727539063, 0.9019801635742187, 0.9017937622070312, 0.9018091430664062, 0.90191357421875, 1.858798583984375, 0.9011435546875, 0.9014241333007813, 0.90132275390625, 0.9010053100585937, 0.9013289184570312, 0.9013585815429688, 0.9011660766601562, 0.9014292602539062, 0.901317626953125, 0.901381103515625, 0.901228515625, 0.9015480346679687, 0.9016514282226562, 0.9015521240234375, 0.90144873046875, 0.9019647827148437, 0.9018716430664062, 0.9012531127929687, 0.9012572021484375, 0.901212158203125, 0.9014169311523438, 0.901138427734375, 0.9026416625976562, 0.9016320190429687, 0.9017907104492188, 0.9014876098632812, 0.9016504516601562, 0.901855224609375, 0.9019913940429688, 0.9016053466796875, 0.9015654296875, 0.9016565551757812, 0.9016135864257813, 0.901591064453125, 0.9022617797851562, 0.901928955078125, 0.9018091430664062, 0.901897216796875, 0.9014537963867187, 0.9010780029296875, 0.9012725830078125, 0.9011988525390625, 0.9014630126953125, 0.90176416015625, 0.9015735473632812, 0.9017302856445313, 0.9021542358398438, 0.9014528198242188, 0.9019176635742188, 0.9015859375, 0.9015582275390625, 0.901423095703125, 0.9016873168945313, 0.9015879516601563, 0.9017313232421875, 0.9018091430664062, 0.902240234375, 0.9021122436523438, 0.9021419677734375, 0.9018674926757813, 0.9018511352539063, 0.9022258911132812, 1.8597550048828124, 0.9009326171875, 0.90142822265625, 0.901043212890625, 0.9013135375976562, 0.9019094848632813, 0.9014773559570313, 0.9016790771484375, 0.9016719360351563, 0.9012367553710937, 0.9012531127929687, 0.9010974731445313, 0.901064697265625, 0.90119677734375, 0.9015040283203125, 0.901970947265625, 0.90260888671875, 0.9018121948242187, 0.9016063842773437, 0.9018624267578125, 0.9015449829101563, 0.901992431640625, 0.9017988891601563, 0.901696533203125, 0.9012930297851562, 0.9012428588867187, 0.90197607421875, 0.9014159545898438, 0.90151220703125, 0.9018674926757813, 0.9016678466796875, 0.9016135864257813, 0.9015623779296875, 0.9015643920898437, 0.9014937744140625, 0.902150146484375, 0.9015090942382813, 0.9018839111328125, 0.9017354125976562, 0.9015787353515625, 0.9013851928710938, 0.9017507934570312, 0.9015869140625, 0.9017579345703125, 0.901917724609375, 0.9019903564453124, 0.90163916015625, 0.9023303833007813, 0.901960693359375, 0.9023068237304688, 0.9017538452148437, 0.9017886962890626, 0.901560302734375, 0.9016637573242188, 0.9014384765625, 0.9020631103515625, 0.9018910522460938, 0.9014251708984375, 0.9013934326171875, 0.9016893310546875, 0.9019873657226563, 0.9017251586914062, 0.901432373046875, 1.8595020751953124, 0.9010237426757812, 0.9012930297851562, 0.9010022583007813, 0.9012469482421875, 0.9015675048828125, 0.9011814575195313, 0.9016350708007812, 0.9019985961914062, 0.9014097900390625, 0.9013739624023438, 0.9014691772460938, 0.901528564453125, 0.9014323120117187, 0.90157568359375, 0.9015634155273438, 0.9018613891601562, 0.9021675415039062, 0.9016709594726563, 0.9019002685546875, 0.9019852905273438, 0.9018562622070313, 0.900979736328125, 0.90132373046875, 0.901507080078125, 0.9013309936523437, 0.9011680908203125, 0.9011712036132813, 0.9011783447265626, 0.9020948486328125, 0.9016350708007812, 0.9018470458984374, 0.9016832275390625, 0.901180419921875, 0.9011251220703125, 0.901518310546875, 0.9013657836914063, 0.9012398071289063, 0.90106982421875, 0.9011405029296875, 0.901254150390625, 0.9014476928710937, 0.9010718994140625, 0.9012920532226563, 0.90113330078125, 0.9011609497070312, 0.901159912109375, 0.9013043212890625, 0.9009213256835937, 0.9009336547851563, 0.90114453125, 0.9015357666015625, 0.9015296020507813, 0.9014456176757812, 0.9014261474609375, 0.9017907104492188, 0.9016565551757812, 0.9016832275390625, 0.90138623046875, 0.9014343872070313, 0.9012725830078125, 0.9016371459960938, 0.902470703125, 1.8604583740234375, 0.9015695190429688, 0.9012644653320312, 0.9009939575195313, 0.9012254638671875, 0.9013217163085937, 0.9012008666992187, 0.9012828369140625, 0.9010513916015624, 0.9015951538085938, 0.9012674560546875, 0.9009285278320313, 0.9009203491210938, 0.9009633178710937, 0.9009141845703125, 0.9008394165039062, 0.9009039306640625, 0.901094482421875, 0.9009048461914062, 0.9008230590820312, 0.9013002319335938, 0.9019085083007813, 0.9010206909179688, 0.9013463134765625, 0.9014251708984375, 0.9011159057617187, 0.9022166748046875, 0.9010892944335938, 0.900806640625, 0.9012715454101563, 0.9008599243164063, 0.9013524780273438, 0.9015029907226563, 0.9012828369140625, 0.9010565185546875, 0.9016555786132813, 0.9012265014648437, 0.90136474609375, 0.9013237915039063, 0.9012940673828125, 0.90121728515625, 0.9013165893554688, 0.901396484375, 0.9020753784179687, 0.9011875610351563, 0.9012091064453125, 0.9015101318359375, 0.9015930786132812, 0.90141796875, 0.9013585815429688, 0.9015480346679687, 0.9018736572265625, 0.9019412231445313, 0.9015439453125, 0.9019586791992188, 0.9018480834960938, 0.9013289184570312, 0.9018142700195313, 0.9015265502929688, 0.901791748046875, 0.9013944091796875, 0.9017569580078125, 0.9013258056640625]",tokens/s,1.0926340535470893,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-7B,Qwen/Qwen1.5-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,2628.173824,8389.132288,0.0,7742.685184,7007.145472,s,10,5.8011931152343745,0.5801193115234377,0.0010375594919882804,0.5799924316406251,0.5812619140624999,0.5815937561035156,0.5818592297363281,"[0.5800941772460938, 0.5819255981445313, 0.5788403930664062, 0.5789609375, 0.5798906860351563, 0.5787822265625, 0.57971533203125, 0.5806090698242188, 0.5811865234375, 0.5811881713867187]",tokens/s,441.28853309110576,kWh,6.83268827420694e-06,3.744032605603327e-06,3.3321369249667614e-05,4.3898090129477875e-05,tokens/kWh,5831688.787483131,MB,2628.173824,8389.132288,0.0,7742.685184,7283.985408,s,10,337.06498046875,33.706498046875,0.006009418787724905,33.706857421875,33.7119703125,33.715088671875,33.717583359375006,"[33.70903515625, 33.70971875, 33.7039609375, 33.70134375, 33.71127734375, 33.70996875, 33.71820703125, 33.69992578125, 33.69686328125, 33.7046796875]",tokens/s,1.869075805869452,kWh,0.0003979287963388143,0.0002180992035304977,0.0019156344121359354,0.0025316624120052473,tokens/kWh,24884.8344476149,,s,629,341.74955133056636,0.5433220211932693,0.06898772267341372,0.5349846801757813,0.5355790161132813,0.5358112670898437,1.1147944775390626,"[0.5351260375976562, 0.535109619140625, 0.534408203125, 0.5347737426757813, 0.5349273681640625, 0.5355181884765625, 0.5343795166015625, 0.5353154296875, 0.5344747314453125, 0.5348731079101563, 0.5341849365234375, 0.5352171630859375, 0.5348259887695312, 0.5350338745117188, 0.5346682739257812, 0.5355919189453126, 0.5350399780273437, 0.535109619140625, 0.53498779296875, 0.5350379638671875, 0.5342330932617188, 0.534898681640625, 0.5348997192382813, 0.535445556640625, 0.5350767822265625, 0.5357158203125, 0.5351649169921875, 0.53486181640625, 0.5343006591796875, 0.5351065673828125, 0.5353646240234375, 0.5352591552734375, 0.53469091796875, 0.5347808227539063, 0.5347430419921875, 0.5353215942382813, 0.5346262817382812, 0.5358970947265626, 0.5353533325195312, 0.535372802734375, 0.5349212646484375, 0.5354249877929688, 0.5347993774414063, 0.5352724609375, 0.5352263793945312, 0.5359462280273437, 0.5349867553710937, 0.5358069458007813, 0.5356390380859375, 0.5357721557617188, 0.5345842895507813, 0.5350502319335938, 0.5345679321289063, 0.5353901977539063, 0.534408203125, 0.5348423461914062, 0.5349427490234375, 0.5352744750976562, 0.5348515625, 0.5350000610351563, 0.5347225341796875, 0.5350891723632812, 1.117781005859375, 0.5343908081054688, 0.5352744750976562, 0.5350615234375, 0.5358694458007812, 0.5350860595703125, 0.5349007568359375, 0.5344706420898437, 0.5346253051757812, 0.534213623046875, 0.5346007080078125, 0.534765625, 0.5350972900390625, 0.5347195434570312, 0.5356298217773438, 0.5349805908203125, 0.5357711181640625, 0.5348290405273437, 0.5349857177734375, 0.5353421020507813, 0.5351044921875, 0.5350942993164063, 0.5355233154296875, 0.5351280517578125, 0.536111083984375, 0.53477783203125, 0.5354240112304688, 0.5351700439453125, 0.5359093627929687, 0.5350686645507813, 0.5352868041992187, 0.5346743774414062, 0.534993896484375, 0.534455322265625, 0.5347471313476563, 0.5344000244140625, 0.5349601440429688, 0.5342822265625, 0.5353840942382813, 0.5345341186523438, 0.5348598022460938, 0.5351505737304687, 0.5354915771484375, 0.5350184936523438, 0.535066650390625, 0.5345361328125, 0.53806591796875, 0.5347102661132812, 0.535362548828125, 0.5352243041992187, 0.5350154418945312, 0.5345475463867188, 0.5348187255859375, 0.5344368896484375, 0.5351710815429688, 0.5345422973632813, 0.5351874389648438, 0.5350225830078125, 0.5353594970703125, 0.5346467895507813, 0.53593701171875, 0.5349805908203125, 0.535109619140625, 1.11461376953125, 0.5346682739257812, 0.5355612182617188, 0.535363525390625, 0.5350051879882812, 0.5345147094726562, 0.5349580688476563, 0.5345648803710937, 0.5349765014648438, 0.5342883911132813, 0.534782958984375, 0.5348720703125, 0.5348485107421875, 0.5344696044921875, 0.5351680297851562, 0.535014404296875, 0.5347891235351563, 0.5343958740234375, 0.534729736328125, 0.5343323974609375, 0.5347850341796875, 0.534245361328125, 0.5351393432617187, 0.5362525024414062, 0.53484130859375, 0.5343948974609375, 0.53480859375, 0.5345147094726562, 0.535352294921875, 0.5345781860351563, 0.5349120483398437, 0.5347604370117187, 0.5352069091796875, 0.5347593994140625, 0.5351116943359375, 0.5349611206054687, 0.5350625, 0.5345709838867188, 0.535088134765625, 0.5350850830078125, 0.535235595703125, 0.53505126953125, 0.5352868041992187, 0.5351475219726562, 0.5351701049804688, 0.535184326171875, 0.5351956176757813, 0.536447998046875, 0.5354301147460937, 0.53513427734375, 0.5351146850585937, 0.5348946533203125, 0.5352826538085937, 0.5349775390625, 0.5352734985351563, 0.5348341674804687, 0.5351444702148438, 0.5347440185546875, 0.53526220703125, 0.5349683227539063, 0.5353143920898438, 0.5344860229492188, 0.534940673828125, 1.1146219482421875, 0.5341777954101562, 0.5348720703125, 0.5346262817382812, 0.5350717163085937, 0.5346693725585937, 0.5351198120117188, 0.5349754638671875, 0.53484033203125, 0.5343754272460938, 0.5346856689453126, 0.534470703125, 0.5363936767578125, 0.5344901123046875, 0.5354598388671875, 0.5345167236328126, 0.535103515625, 0.5345525512695313, 0.5347041015625, 0.5342689208984375, 0.53492431640625, 0.5343795166015625, 0.5349335327148438, 0.5343723754882812, 0.53469287109375, 0.5351454467773438, 0.5353369750976562, 0.5347051391601563, 0.5347532958984375, 0.5346580200195312, 0.5352058715820313, 0.5351024780273438, 0.5352191772460938, 0.53522021484375, 0.5348075561523438, 0.534297607421875, 0.534724609375, 0.5345833129882812, 0.5349539794921875, 0.5344808959960937, 0.53492529296875, 0.5351454467773438, 0.5350830078125, 0.5344010009765625, 0.5376593627929688, 0.5356707763671875, 0.5352120361328125, 0.5348014526367187, 0.5355078735351563, 0.5344962768554687, 0.5350451049804688, 0.5349385986328125, 0.5351188354492188, 0.5344829711914062, 0.5351536865234375, 0.5346395874023437, 0.5351792602539063, 0.534635498046875, 0.5352366333007812, 0.534640625, 0.5355222778320312, 0.53492431640625, 0.5352960205078126, 1.1157667236328126, 0.5344829711914062, 0.5347286987304688, 0.53443994140625, 0.5350502319335938, 0.5349959716796875, 0.5357506713867187, 0.5345075073242187, 0.5348935546875, 0.5344307250976562, 0.5347952880859375, 0.5344050903320312, 0.5348792114257812, 0.5342566528320313, 0.5350963745117188, 0.5344285888671875, 0.534782958984375, 0.5359862060546875, 0.5356195678710938, 0.53477783203125, 0.5348853759765625, 0.534413330078125, 0.5349529418945312, 0.5347205200195313, 0.5349427490234375, 0.5349754638671875, 0.5349099731445313, 0.5344389038085937, 0.5352949829101562, 0.5348628540039062, 0.5350799560546875, 0.5344368896484375, 0.5352714233398438, 0.5348720703125, 0.5350604858398438, 0.534603759765625, 0.5353707275390625, 0.5352437744140625, 0.535731201171875, 0.5351393432617187, 0.535103515625, 0.5343733520507813, 0.534950927734375, 0.53507275390625, 0.535462890625, 0.5348690185546875, 0.5352427368164062, 0.5344849853515625, 0.5353912353515625, 0.5345740966796875, 0.5349765014648438, 0.5348229370117188, 0.5356441650390625, 0.5349171142578125, 0.5356072998046875, 0.5357066040039062, 0.5356267700195313, 0.5386843872070313, 0.535677978515625, 0.53519970703125, 0.5360302124023437, 0.5351813354492188, 0.5363681030273437, 1.114861572265625, 0.534814697265625, 0.5358141479492188, 0.5348894653320313, 0.535541748046875, 0.5344706420898437, 0.5349130249023437, 0.5350021362304688, 0.5350850830078125, 0.534813720703125, 0.5355775756835938, 0.5344389038085937, 0.53571484375, 0.5351659545898437, 0.5355847778320313, 0.53517822265625, 0.5354332275390625, 0.53538818359375, 0.5351044921875, 0.5344112548828125, 0.5350000610351563, 0.5343201293945312, 0.5348331298828125, 0.5342853393554687, 0.5347276611328124, 0.5348331298828125, 0.534877197265625, 0.534382568359375, 0.5348782348632812, 0.5349212036132812, 0.5350440673828125, 0.5346990356445313, 0.5352007446289062, 0.5345945434570313, 0.5349151000976563, 0.5347184448242187, 0.5349703979492187, 0.5343866577148437, 0.5375672607421875, 0.5345842895507813, 0.535604248046875, 0.5345730590820312, 0.5350553588867187, 0.5347706909179688, 0.5352007446289062, 0.5347727661132813, 0.535562255859375, 0.5352611694335937, 0.5354332275390625, 0.5348392944335938, 0.5353809814453125, 0.5349007568359375, 0.5352796020507813, 0.5347973022460938, 0.535889892578125, 0.535098388671875, 0.5354475708007812, 0.5351219482421875, 0.5354977416992187, 0.5351229248046875, 0.5353697509765625, 0.5347573852539063, 0.5351884765625, 1.11545751953125, 0.534709228515625, 0.5355346069335938, 0.5345218505859375, 0.53505126953125, 0.5347532958984375, 0.5353421020507813, 0.5347676391601562, 0.535130126953125, 0.53502978515625, 0.5353748779296875, 0.5347379150390625, 0.53549462890625, 0.5350656127929687, 0.53538818359375, 0.5347123413085938, 0.53515673828125, 0.5347584228515625, 0.5352509155273437, 0.5351055297851562, 0.5358919677734375, 0.5351751708984375, 0.535287841796875, 0.5365626831054687, 0.5354188842773437, 0.5353696899414062, 0.5352734985351563, 0.5352540283203125, 0.5353543701171875, 0.5349550170898437, 0.5355130615234375, 0.5348792114257812, 0.535568359375, 0.5348311767578126, 0.5353768310546875, 0.5346519165039062, 0.53517822265625, 0.5345648803710937, 0.5352017822265625, 0.535309326171875, 0.5357240600585937, 0.5350317993164062, 0.5355181884765625, 0.5348699951171875, 0.5352345581054687, 0.5348505859375, 0.5359411010742188, 0.5351085815429687, 0.5355888671875, 0.5350133666992187, 0.5354229736328125, 0.5347727661132813, 0.5353400268554688, 0.5348945922851562, 0.5355601806640625, 0.5356461791992188, 0.5350768432617188, 0.5353328857421875, 0.5356503295898437, 0.5349284057617187, 0.53524169921875, 0.53501953125, 0.5352315063476563, 1.1161282958984375, 0.5346826171875, 0.53496630859375, 0.5344050903320312, 0.5346508178710937, 0.5344050903320312, 0.5348168334960938, 0.5343477172851563, 0.5350819702148437, 0.5347593994140625, 0.5352212524414063, 0.534572021484375, 0.5355878295898437, 0.53490380859375, 0.5348966674804687, 0.5346734008789062, 0.5364859008789062, 0.53452392578125, 0.5349498901367188, 0.5348843383789063, 0.5350963134765625, 0.5344942016601563, 0.5349078979492188, 0.5345894165039062, 0.5349498901367188, 0.5344747314453125, 0.53500927734375, 0.5347010498046875, 0.535119873046875, 0.5345147094726562, 0.53530419921875, 0.5354342651367188, 0.535141357421875, 0.5348782348632812, 0.5348505859375, 0.5344890747070312, 0.5351444702148438, 0.53507275390625, 0.535056396484375, 0.5347092895507812, 0.5349498291015625, 0.534508544921875, 0.5350236206054687, 0.5345904541015625, 0.535593994140625, 0.5347225341796875, 0.5352581176757812, 0.5349776000976563, 0.5351546020507812, 0.5346416625976562, 0.5357752075195312, 0.5347604370117187, 0.5349846801757813, 0.5344286499023437, 0.5350154418945312, 0.5346324462890625, 0.5349918823242188, 0.5349908447265626, 0.5351976928710938, 0.5346375732421875, 0.5349151000976563, 0.5348065185546875, 0.5351188354492188, 1.1169342041015624, 0.534445068359375, 0.5351669921875, 0.5345443725585938, 0.53528369140625, 0.534782958984375, 0.5349293823242187, 0.5344174194335938, 0.5348372192382812, 0.5343057861328125, 0.53471435546875, 0.534593505859375, 0.5350717163085937, 0.5345064697265625, 0.5349273681640625, 0.5345054931640625, 0.5349867553710937, 0.535751708984375, 0.5350215454101562, 0.5345771484375, 0.535593994140625, 0.534761474609375, 0.5349488525390625, 0.5347625122070313, 0.5349222412109375, 0.5344276733398438, 0.53481982421875, 0.5344440307617188, 0.5348536376953125, 0.5344583740234375, 0.5348925170898438, 0.5345833129882812, 0.5350532836914063, 0.5344010009765625, 0.5347758178710937, 0.5343866577148437, 0.5347593994140625, 0.534382568359375, 0.5348362426757812, 0.53475634765625, 0.5349376831054687, 0.5345422973632813, 0.5351393432617187, 0.53452490234375, 0.5351270141601563, 0.5344603881835938, 0.5351065673828125, 0.5345259399414063, 0.5350799560546875, 0.5351802978515625, 0.5355919189453126, 0.5348382568359376, 0.5353246459960938, 0.5350021362304688, 0.535816162109375, 0.5346324462890625, 0.5352212524414063, 0.5346836547851562, 0.5353799438476563, 0.5346682739257812, 0.5351454467773438, 0.5346416625976562, 0.53532568359375, 1.116949462890625, 0.5343743896484375, 0.5350113525390625, 0.5348218994140626, 0.5349816284179687, 0.5344235229492188, 0.5348731079101563, 0.534445068359375, 0.534877197265625, 0.53441943359375, 0.5347359008789062, 0.5345535888671875, 0.5347891235351563, 0.5345515747070313, 0.5351004028320312, 0.53458740234375, 0.5349119873046875, 0.5347153930664063, 0.5349099731445313, 0.5344542846679687, 0.5348812866210938, 0.5343672485351563, 0.5352898559570313, 0.5349918823242188, 0.5351209106445313, 0.5350277099609375, 0.5352345581054687, 0.5350748291015625, 0.5352857666015625, 0.5347102661132812, 0.5349519653320313, 0.5349539794921875, 0.5350645751953125, 0.534740966796875, 0.5353809814453125, 0.5346959228515625, 0.5358069458007813, 0.53460888671875, 0.5353052368164063, 0.5350338745117188, 0.5352212524414063, 0.53507275390625, 0.5351802978515625, 0.5347666015625, 0.5349642333984375, 0.53477783203125, 0.535235595703125, 0.5350543212890625, 0.5352120361328125, 0.5350082397460938, 0.5354352416992187, 0.5352120361328125, 0.5354669799804688, 0.535140380859375, 0.535362548828125, 0.534782958984375, 0.5354803466796875, 0.5349232788085938, 0.5352908935546875, 0.535625732421875, 0.5352755126953125, 0.53500732421875, 0.5353860473632812]",tokens/s,1.8405291171591998,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-34B,01-ai/Yi-34B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,,,MB,1767.067648,22129.672192,0.0,21483.225088,20799.168,s,10,28.05056616210938,2.805056616210938,0.0027095714336187707,2.804907958984375,2.8082408447265625,2.808313903808594,2.808372351074219,"[2.805282470703125, 2.808224609375, 2.803196533203125, 2.80234619140625, 2.80277294921875, 2.800419677734375, 2.804533447265625, 2.80722705078125, 2.80817626953125, 2.808386962890625]",tokens/s,91.26375507735885,kWh,3.3081144508388305e-05,1.8129723654910774e-05,0.00015725384802519894,0.000208464716188498,tokens/kWh,1228025.5607789264,MB,1772.261376,22129.672192,0.0,21483.225088,20902.144,s,10,1667.162484375,166.71624843749996,0.015085835868376713,166.71732031250002,166.73279531249997,166.73349140624998,166.73404828124998,"[166.730125, 166.715734375, 166.68440625, 166.724703125, 166.7341875, 166.71721875, 166.708265625, 166.717421875, 166.69778125, 166.732640625]",tokens/s,0.377887581987055,kWh,0.0019678901955319777,0.0010785770549249355,0.009270038554913802,0.012316505805370715,tokens/kWh,5115.087103074991,,s,629,1689.7617126464822,2.686425616290118,0.33327430639158717,2.64620947265625,2.6474256835937497,2.647894970703125,5.4507228125000005,"[2.646453369140625, 2.646373291015625, 2.64631591796875, 2.6454580078125, 2.646531982421875, 2.647617431640625, 2.64826171875, 2.647658447265625, 2.647793701171875, 2.64637646484375, 2.64711376953125, 2.64620947265625, 2.646602783203125, 2.646426513671875, 2.64635498046875, 2.64557568359375, 2.644958251953125, 2.647402587890625, 2.64601904296875, 2.646287353515625, 2.647202880859375, 2.64688525390625, 2.646246337890625, 2.646867919921875, 2.64715478515625, 2.647743408203125, 2.646212646484375, 2.647287841796875, 2.648404052734375, 2.64638671875, 2.64559912109375, 2.647835693359375, 2.648161376953125, 2.6474189453125, 2.647456787109375, 2.6468076171875, 2.646687744140625, 2.64502880859375, 2.646285400390625, 2.646274169921875, 2.64652490234375, 2.64616455078125, 2.64710546875, 2.6460908203125, 2.646740966796875, 2.645139404296875, 2.646036376953125, 2.646411376953125, 2.64627099609375, 2.64452197265625, 2.64525830078125, 2.647901123046875, 2.64601806640625, 2.645980224609375, 2.64719677734375, 2.64639990234375, 2.64538720703125, 2.645042236328125, 2.64540576171875, 2.646120361328125, 2.646118408203125, 2.64612353515625, 5.45506494140625, 2.64526025390625, 2.6462626953125, 2.64690576171875, 2.646042724609375, 2.64546826171875, 2.646393798828125, 2.647192626953125, 2.644919189453125, 2.644935791015625, 2.645579833984375, 2.647086181640625, 2.646035400390625, 2.644955078125, 2.645644287109375, 2.645537841796875, 2.644991943359375, 2.645560302734375, 2.646414306640625, 2.645199951171875, 2.646506591796875, 2.646548583984375, 2.64702880859375, 2.64570166015625, 2.644788330078125, 2.64700732421875, 2.646330322265625, 2.64543017578125, 2.645243896484375, 2.646043701171875, 2.645831787109375, 2.6459453125, 2.64614697265625, 2.64631494140625, 2.64669091796875, 2.646116455078125, 2.647741455078125, 2.6478857421875, 2.645671875, 2.647362548828125, 2.646822998046875, 2.648393798828125, 2.647374755859375, 2.64627294921875, 2.646519775390625, 2.645937255859375, 2.6523720703125, 2.647458740234375, 2.646916015625, 2.6467646484375, 2.644612060546875, 2.645245849609375, 2.64698779296875, 2.646232177734375, 2.645567626953125, 2.645909423828125, 2.64679541015625, 2.64620947265625, 2.645465087890625, 2.646067138671875, 2.6467060546875, 2.6458798828125, 2.6457109375, 5.4513837890625, 2.647235595703125, 2.645803955078125, 2.64654443359375, 2.64496533203125, 2.64464892578125, 2.645158935546875, 2.646266845703125, 2.64560546875, 2.646096923828125, 2.64475537109375, 2.64629052734375, 2.644760498046875, 2.64480859375, 2.6463427734375, 2.64760009765625, 2.645761962890625, 2.645760986328125, 2.646381591796875, 2.646036376953125, 2.64513134765625, 2.646679443359375, 2.6458369140625, 2.6495498046875, 2.646917236328125, 2.644770751953125, 2.645843017578125, 2.644592529296875, 2.6452919921875, 2.6457333984375, 2.64617578125, 2.64542822265625, 2.645474365234375, 2.644977783203125, 2.6475703125, 2.645327880859375, 2.64601611328125, 2.64604052734375, 2.644948974609375, 2.645347412109375, 2.64539453125, 2.64572314453125, 2.645024658203125, 2.64540771484375, 2.64523681640625, 2.64541796875, 2.644529052734375, 2.649079833984375, 2.645666748046875, 2.646816650390625, 2.645528564453125, 2.645729248046875, 2.6472724609375, 2.646464599609375, 2.644356201171875, 2.644589599609375, 2.64574267578125, 2.645088134765625, 2.6443017578125, 2.644094970703125, 2.64595556640625, 2.645572509765625, 2.644812744140625, 5.4503955078125, 2.645255126953125, 2.645544921875, 2.64656884765625, 2.648037353515625, 2.645263427734375, 2.64638671875, 2.6460732421875, 2.6460478515625, 2.645971923828125, 2.645147705078125, 2.646581298828125, 2.6468291015625, 2.64502880859375, 2.645927978515625, 2.64683935546875, 2.64574365234375, 2.645297119140625, 2.64578662109375, 2.645689453125, 2.64529296875, 2.645930908203125, 2.64618798828125, 2.64557373046875, 2.645583984375, 2.64784375, 2.6465341796875, 2.645923828125, 2.648280029296875, 2.647232421875, 2.64726220703125, 2.646688720703125, 2.646456298828125, 2.647128173828125, 2.64785009765625, 2.6451201171875, 2.645867431640625, 2.645412841796875, 2.647185302734375, 2.644790283203125, 2.645506103515625, 2.64781005859375, 2.6459228515625, 2.64651171875, 2.6510693359375, 2.647773193359375, 2.6475908203125, 2.648330322265625, 2.646329345703125, 2.64608349609375, 2.64578759765625, 2.64604150390625, 2.646233154296875, 2.647012451171875, 2.644991943359375, 2.6461533203125, 2.646445068359375, 2.6459423828125, 2.646255615234375, 2.647103515625, 2.64721826171875, 2.6459013671875, 2.646496337890625, 5.45085009765625, 2.646131591796875, 2.647123046875, 2.64625244140625, 2.6469755859375, 2.645900146484375, 2.64694384765625, 2.6482001953125, 2.647193603515625, 2.646340576171875, 2.646576171875, 2.647742431640625, 2.645885986328125, 2.64563916015625, 2.645662841796875, 2.646833251953125, 2.64673291015625, 2.648217529296875, 2.645130126953125, 2.648642578125, 2.64700732421875, 2.6464482421875, 2.645592041015625, 2.646584228515625, 2.64501953125, 2.64555224609375, 2.64648388671875, 2.6452490234375, 2.646088623046875, 2.64539453125, 2.646591552734375, 2.645818359375, 2.6455, 2.645107666015625, 2.647025634765625, 2.64500830078125, 2.647201904296875, 2.6471865234375, 2.647762939453125, 2.6468916015625, 2.646411376953125, 2.645792724609375, 2.647160888671875, 2.64576611328125, 2.645645263671875, 2.646044677734375, 2.647015380859375, 2.64671435546875, 2.645159912109375, 2.646703125, 2.646371337890625, 2.646978515625, 2.6460908203125, 2.648290283203125, 2.646096923828125, 2.646834228515625, 2.6468515625, 2.654116943359375, 2.646571044921875, 2.646265869140625, 2.647033935546875, 2.645507080078125, 2.64506884765625, 5.44818896484375, 2.646950927734375, 2.64728466796875, 2.64570068359375, 2.647045166015625, 2.645780517578125, 2.64650341796875, 2.647185302734375, 2.647396240234375, 2.64726416015625, 2.645887939453125, 2.646220947265625, 2.646308837890625, 2.6452724609375, 2.645916748046875, 2.6456513671875, 2.64530029296875, 2.645833740234375, 2.64532275390625, 2.646921142578125, 2.645572509765625, 2.646182861328125, 2.646026123046875, 2.646593505859375, 2.646531005859375, 2.646508544921875, 2.646131591796875, 2.647275634765625, 2.646612060546875, 2.646274169921875, 2.645583984375, 2.646960205078125, 2.647244873046875, 2.6451845703125, 2.6467060546875, 2.646052001953125, 2.64616650390625, 2.645667724609375, 2.652001220703125, 2.64635693359375, 2.645919677734375, 2.646162353515625, 2.64641845703125, 2.645307373046875, 2.644989013671875, 2.646035400390625, 2.645531494140625, 2.645157958984375, 2.645906494140625, 2.645951416015625, 2.646036376953125, 2.645796875, 2.64467041015625, 2.645960693359375, 2.64549267578125, 2.64652294921875, 2.64587255859375, 2.647626708984375, 2.646477783203125, 2.646447021484375, 2.646128662109375, 2.64745263671875, 2.646246337890625, 5.45296484375, 2.646928466796875, 2.647333984375, 2.646259765625, 2.64629443359375, 2.64618701171875, 2.6448466796875, 2.646443115234375, 2.646686767578125, 2.6447626953125, 2.64540576171875, 2.64494287109375, 2.645835693359375, 2.645675048828125, 2.64589404296875, 2.64648095703125, 2.64646044921875, 2.646432861328125, 2.646467529296875, 2.645669921875, 2.64584814453125, 2.6454580078125, 2.64523486328125, 2.650271728515625, 2.647341064453125, 2.64589306640625, 2.64591064453125, 2.64768505859375, 2.64700830078125, 2.64631201171875, 2.645865478515625, 2.647047119140625, 2.646507568359375, 2.64665087890625, 2.647560302734375, 2.645760986328125, 2.645769287109375, 2.645876708984375, 2.645303466796875, 2.646839111328125, 2.646023193359375, 2.645792724609375, 2.64610400390625, 2.646148193359375, 2.646067138671875, 2.64551416015625, 2.645665771484375, 2.64494482421875, 2.645905517578125, 2.645917724609375, 2.645688232421875, 2.64422705078125, 2.64477587890625, 2.64602001953125, 2.646118408203125, 2.645832763671875, 2.64635888671875, 2.646981689453125, 2.645694580078125, 2.6462197265625, 2.6456484375, 2.647057373046875, 2.646053955078125, 5.45440771484375, 2.64606103515625, 2.64624853515625, 2.645525390625, 2.6462197265625, 2.646288330078125, 2.64601806640625, 2.6458369140625, 2.6465341796875, 2.64682080078125, 2.6460498046875, 2.645517333984375, 2.646379638671875, 2.645887939453125, 2.645572509765625, 2.645886962890625, 2.6497626953125, 2.645772216796875, 2.64593603515625, 2.64641748046875, 2.64673681640625, 2.645414794921875, 2.644895751953125, 2.6456689453125, 2.646624267578125, 2.64646142578125, 2.645294921875, 2.646984619140625, 2.646635498046875, 2.64559619140625, 2.644748291015625, 2.64690576171875, 2.64658642578125, 2.645821533203125, 2.6462392578125, 2.647307373046875, 2.64610205078125, 2.647371826171875, 2.64780078125, 2.64745263671875, 2.646867919921875, 2.64559814453125, 2.646322265625, 2.6461328125, 2.6461328125, 2.6467666015625, 2.64734619140625, 2.647150634765625, 2.6465341796875, 2.646138916015625, 2.645937255859375, 2.64439697265625, 2.645222412109375, 2.646958984375, 2.646992919921875, 2.647083984375, 2.64567822265625, 2.646795166015625, 2.645927978515625, 2.646295654296875, 2.64530224609375, 2.646948974609375, 2.646352783203125, 5.45413623046875, 2.646625244140625, 2.6464501953125, 2.647415771484375, 2.64749462890625, 2.64658837890625, 2.645314453125, 2.647509033203125, 2.647132080078125, 2.64675537109375, 2.646547607421875, 2.646182861328125, 2.64610302734375, 2.6463896484375, 2.646077392578125, 2.64599853515625, 2.646762451171875, 2.64707080078125, 2.646077392578125, 2.64540576171875, 2.646277099609375, 2.64477392578125, 2.645159912109375, 2.6459638671875, 2.64671240234375, 2.646478759765625, 2.645445556640625, 2.646115234375, 2.646445068359375, 2.645536865234375, 2.64454443359375, 2.646118408203125, 2.645505126953125, 2.644704345703125, 2.6460517578125, 2.64701953125, 2.645275634765625, 2.64492431640625, 2.645380126953125, 2.645751708984375, 2.64485888671875, 2.64445654296875, 2.645286865234375, 2.646550537109375, 2.646221923828125, 2.645315673828125, 2.6454169921875, 2.64634375, 2.645382080078125, 2.6473984375, 2.646042724609375, 2.64547119140625, 2.644828125, 2.64538720703125, 2.646898681640625, 2.646388671875, 2.64596484375, 2.6477158203125, 2.64626171875, 2.6452685546875, 2.645350341796875, 2.6460498046875, 2.644905029296875, 5.4557060546875, 2.64730419921875, 2.64740966796875, 2.646350830078125, 2.645677978515625, 2.6465546875, 2.64831689453125, 2.64722119140625, 2.646279052734375, 2.645445556640625, 2.646408203125, 2.646245361328125, 2.646921142578125, 2.64698974609375, 2.646667236328125, 2.646288330078125, 2.64876953125, 2.647458740234375, 2.645729248046875, 2.645494873046875, 2.646150146484375, 2.6453955078125, 2.6452890625, 2.6467861328125, 2.647785400390625, 2.646699951171875, 2.65012744140625, 2.64613671875, 2.646697998046875, 2.6466845703125, 2.646642578125, 2.6476943359375, 2.64681787109375, 2.645788818359375, 2.646158203125, 2.647160888671875, 2.645675048828125, 2.64652490234375, 2.64439306640625, 2.64669189453125, 2.646427734375, 2.64625146484375, 2.646435791015625, 2.645409912109375, 2.646538330078125, 2.64618505859375, 2.6468310546875, 2.647814208984375, 2.645505126953125, 2.645622802734375, 2.6466826171875, 2.646696044921875, 2.6457548828125, 2.6469130859375, 2.6469990234375, 2.64591259765625, 2.64690185546875, 2.64622705078125, 2.64720703125, 2.646657958984375, 2.645370849609375, 2.645916748046875, 2.646266845703125]",tokens/s,0.3722418346281902,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/gemma-2b,google/gemma-2b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/gemma-2b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aa98c-2707b1f61d036e837cb54abf;30c4bd9e-012e-4e4d-85c7-b872135caccb) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/gemma-2b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/gemma-2b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/polyglot-ko-12.8b,EleutherAI/polyglot-ko-12.8b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,,,MB,1740.263424,9941.024768,0.0,9294.577664,8910.102528,s,10,10.642888305664064,1.0642888305664062,0.0008435144191809837,1.06450927734375,1.0651936889648437,1.065327557373047,1.0654346520996094,"[1.065136474609375, 1.06546142578125, 1.063631591796875, 1.063218505859375, 1.06368701171875, 1.0628612060546876, 1.0644658203125, 1.064552734375, 1.0647095947265626, 1.0651639404296875]",tokens/s,240.5362084498799,kWh,1.2557922783825134e-05,6.880571250694628e-06,6.097363211219897e-05,8.041212614671872e-05,tokens/kWh,3183599.4428614704,MB,1740.86144,9941.024768,0.0,9294.577664,9220.867072,s,10,630.7354296875,63.07354296875,0.00490218127220892,63.07426171875,63.081011718750005,63.081064453125,63.081106640625,"[63.0750703125, 63.07307421875, 63.081, 63.075046875, 63.07416015625, 63.0811171875, 63.07436328125, 63.0665078125, 63.06765234375, 63.0674375]",tokens/s,0.9988340124037992,kWh,0.0007445980064074199,0.0004081038876869388,0.003603297882636,0.004755999776730358,tokens/kWh,13246.426189555263,,s,629,639.312124511719,1.016394474581429,0.12643667238363906,1.0011002807617186,1.0017415283203124,1.002017578125,2.065029130859375,"[1.0008678588867188, 1.0004398193359374, 1.001291748046875, 1.0011023559570313, 1.0010675048828126, 1.0010203857421875, 1.0008780517578124, 1.000616943359375, 1.0006917114257812, 1.0008555297851562, 1.0007982177734376, 1.000685546875, 1.0008023071289063, 1.000985595703125, 1.0008780517578124, 1.0010081787109375, 1.0006251220703124, 1.0007255249023437, 1.000806396484375, 1.0009149169921876, 1.0006405029296874, 1.0007285766601564, 1.0012835693359374, 1.0011812133789062, 1.0010265502929687, 1.0012303466796875, 1.0011586303710938, 1.0011637573242187, 1.0010664672851564, 1.0014730224609374, 1.0011238403320313, 1.0014033813476562, 1.0011033325195313, 1.0013900756835938, 1.0010501098632814, 1.0017617797851563, 1.0014638061523438, 1.0016256103515624, 1.0009415893554687, 1.0013726806640626, 1.000985595703125, 1.0008145751953126, 1.001359375, 1.0016143188476563, 1.0011217651367188, 1.0013255615234375, 1.0009302978515624, 1.001133056640625, 1.0009508056640626, 1.001533447265625, 1.0019143676757813, 1.0015897827148437, 1.0020361938476563, 1.0020464477539062, 1.00206591796875, 1.0012640991210937, 1.0014044189453124, 1.001734130859375, 1.0015846557617187, 1.00234033203125, 1.0014207763671874, 1.0013255615234375, 2.067092529296875, 1.0008361206054688, 1.0011576538085938, 1.0006384887695312, 1.001079833984375, 1.0007879638671875, 1.0012293090820312, 1.0010501098632814, 1.001322509765625, 1.0013265991210938, 1.0013900756835938, 1.000838134765625, 1.001486328125, 1.0010675048828126, 1.0011484375, 1.0016091918945313, 1.001449462890625, 1.0011023559570313, 1.0005330200195313, 1.0008074340820312, 1.0010623779296874, 1.0006220703125, 1.0009927978515625, 1.0007716064453125, 1.001026611328125, 1.0008770141601562, 1.0014136352539063, 1.0009508056640626, 1.0009508056640626, 1.00097021484375, 1.001290771484375, 1.00075830078125, 1.0016583862304687, 1.0007131958007813, 1.0006619873046876, 1.0005339965820312, 1.0010706176757813, 1.0007859497070313, 1.0006456298828126, 1.0009036865234375, 1.00139111328125, 1.0010776977539062, 1.0015672607421875, 1.0008135375976563, 1.0011463623046875, 1.000784912109375, 1.0035189819335937, 1.00113818359375, 1.0015252685546876, 1.0012507934570312, 1.0016573486328124, 1.0014893798828124, 1.001533447265625, 1.0010839233398436, 1.0011525268554688, 1.0011566162109375, 1.0015396118164062, 1.0014955444335938, 1.0011074829101563, 1.0009037475585938, 1.0016132202148438, 1.0009343872070313, 1.002029052734375, 2.06504150390625, 1.0005913696289062, 1.0010398559570313, 1.00097021484375, 1.0005196533203125, 1.0015057983398437, 1.0011084594726563, 1.0007337036132813, 1.0012190551757814, 1.0008473510742188, 1.0012252197265625, 1.0008402099609375, 1.0011617431640625, 1.0010194091796876, 1.0008719482421875, 1.0009927978515625, 1.0008237915039062, 1.00075927734375, 1.0011688842773439, 1.000806396484375, 1.0008606567382812, 1.0010132446289062, 1.0011740112304688, 1.00221337890625, 1.0005995483398438, 1.000721435546875, 1.00090673828125, 1.0007039794921875, 1.001343994140625, 1.001470947265625, 1.0014852905273437, 1.00143408203125, 1.0017413330078124, 1.002018798828125, 1.0018693237304688, 1.0017587280273437, 1.0011954956054687, 1.0007203979492187, 1.0011064453125, 1.00105419921875, 1.0013839111328124, 1.0012108764648437, 1.001881591796875, 1.0012119140625, 1.0016245727539062, 1.0019154052734376, 1.0013501586914062, 1.0026547241210937, 1.0016737060546874, 1.0011678466796874, 1.00153857421875, 1.0014085083007813, 1.0018191528320313, 1.001275390625, 1.0014157104492187, 1.0018908081054687, 1.0011791381835937, 1.0009927978515625, 1.0019522705078125, 1.0013368530273437, 1.0016307373046875, 1.0016552734375, 1.001829345703125, 2.064997314453125, 1.001175048828125, 1.0011033935546876, 1.0011688232421876, 1.001164794921875, 1.0007736206054687, 1.001064453125, 1.0007973022460936, 1.00076025390625, 1.0006077270507812, 1.0007183227539063, 1.0011361083984376, 1.0011351318359376, 1.0006005859375, 1.0008688354492188, 1.000642578125, 1.00075830078125, 1.0010562744140625, 1.001421875, 1.0009476318359376, 1.000995849609375, 1.0008811645507814, 1.001438232421875, 1.000816650390625, 1.0011361083984376, 1.0006569213867187, 1.0008370971679688, 1.0008350830078125, 1.00099072265625, 1.0007787475585936, 1.0014464111328125, 1.0013634643554687, 1.0017720336914062, 1.001069580078125, 1.001523193359375, 1.0014351196289062, 1.002029052734375, 1.001865234375, 1.0020095825195312, 1.0007572631835937, 1.0010081176757812, 1.0010040283203125, 1.000943603515625, 1.001080810546875, 1.0031155395507811, 1.000857666015625, 1.00139208984375, 1.001006103515625, 1.0015139770507813, 1.0011719970703126, 1.0016010131835937, 1.0014443359375, 1.0011443481445312, 1.0013952026367188, 1.0011340942382811, 1.0010603637695312, 1.0010203857421875, 1.0011033935546876, 1.0015077514648438, 1.00125390625, 1.002071044921875, 1.0015109252929688, 1.0014412841796876, 2.065207275390625, 1.0004561767578124, 1.0008197021484375, 1.0008104858398437, 1.0015027465820312, 1.0008872680664063, 1.0009497680664063, 1.0020311279296874, 1.0013726806640626, 1.0015405883789064, 1.00103271484375, 1.0013511962890624, 1.0011668701171874, 1.0011986083984374, 1.00210791015625, 1.0007920532226562, 1.0011351318359376, 1.0023803100585937, 1.0009579467773437, 1.0010460205078124, 1.0012252197265625, 1.0007500610351563, 1.001091064453125, 1.0007879638671875, 1.0017576904296874, 1.0010828857421874, 1.0014474487304688, 1.0011094970703125, 1.0007890014648437, 1.0009169921875, 1.0007787475585936, 1.0004838256835937, 1.0008545532226563, 1.0006609497070313, 1.0009682006835938, 1.0007777099609374, 1.0011300048828125, 1.00124365234375, 1.000796142578125, 1.0009343872070313, 1.0009251708984375, 1.000722412109375, 1.0007971801757813, 1.0006896362304687, 1.0014924926757813, 1.0007080688476562, 1.0007705688476562, 1.0013204345703124, 1.001218017578125, 1.0010511474609376, 1.0009517822265626, 1.0008607177734374, 1.0009199829101563, 1.00119140625, 1.0017146606445313, 1.0010194091796876, 1.0013358154296874, 1.003936767578125, 1.001802734375, 1.0014443359375, 1.0012477416992187, 1.0015396118164062, 1.0018938598632812, 2.064691162109375, 1.0003916625976563, 1.0009784545898437, 1.0010337524414064, 1.0008944702148437, 1.00107470703125, 1.0008616943359374, 1.001354248046875, 1.00143408203125, 1.0012088623046875, 1.0020802612304687, 1.0009651489257811, 1.001312255859375, 1.0014505004882812, 1.0007481079101563, 1.0010009155273438, 1.0012723388671876, 1.0009712524414063, 1.0011033325195313, 1.0011371459960938, 1.0015723266601562, 1.0016163940429688, 1.0008381958007813, 1.0017801513671876, 1.0015897827148437, 1.0014259033203126, 1.0019799194335937, 1.0012518310546874, 1.0014515380859375, 1.0014484252929687, 1.001759765625, 1.00107568359375, 1.0014505004882812, 1.0014423217773438, 1.001112548828125, 1.0011678466796874, 1.0010286254882812, 1.0009210815429688, 1.0033643798828125, 1.0010767211914062, 1.001469970703125, 1.0009139404296874, 1.001143310546875, 1.00132763671875, 1.0011791381835937, 1.001169921875, 1.0012620849609375, 1.0009456787109374, 1.0010347290039063, 1.00092724609375, 1.0015047607421874, 1.00096923828125, 1.0009251708984375, 1.0013870239257812, 1.0013132934570312, 1.0014105834960938, 1.0011658325195312, 1.0012948608398438, 1.0014935302734376, 1.0009978637695311, 1.0015344848632812, 1.001290771484375, 1.0020157470703126, 2.0664228515625, 1.001302001953125, 1.000827880859375, 1.0009938354492187, 1.0006937255859376, 1.0006784057617188, 1.000911865234375, 1.0006487426757813, 1.0011852416992189, 1.0008043823242188, 1.0012477416992187, 1.0013726806640626, 1.0013101806640625, 1.0008023071289063, 1.0008514404296875, 1.0008248291015625, 1.0011658325195312, 1.001006103515625, 1.0012498168945312, 1.0011207885742188, 1.0012610473632813, 1.0012415771484375, 1.0010951538085937, 1.0023833618164062, 1.0010685424804688, 1.0005288696289063, 1.000853515625, 1.000543212890625, 1.0008811645507814, 1.0009784545898437, 1.0013153076171875, 1.0010859375, 1.0009425659179687, 1.0009343872070313, 1.000748046875, 1.000453125, 1.0013388671875, 1.00075927734375, 1.000826904296875, 1.00143310546875, 1.0011207885742188, 1.0015027465820312, 1.001175048828125, 1.0009395141601563, 1.0011279296875, 1.0010224609375, 1.0013798217773437, 1.0013460693359375, 1.00193896484375, 1.00166552734375, 1.0014197998046874, 1.0018252563476562, 1.0011893920898438, 1.0011268920898437, 1.0015764770507813, 1.0016574096679687, 1.0018969116210938, 1.0011760864257813, 1.0016829223632813, 1.0013952026367188, 1.0015078125, 1.0017791748046876, 1.001365478515625, 2.065933349609375, 1.00049609375, 1.0008944702148437, 1.00080126953125, 1.0008135375976563, 1.0008340454101563, 1.0008822021484376, 1.001302001953125, 1.0008237915039062, 1.0006292724609376, 1.0011105346679687, 1.0010654907226562, 1.0013255615234375, 1.0006692504882813, 1.0007797241210938, 1.0010091552734375, 1.0015211791992187, 1.0007367553710937, 1.0007234497070312, 1.0005473022460938, 1.0006712036132812, 1.00077978515625, 1.0011351318359376, 1.001279541015625, 1.0010480346679687, 1.0010921020507813, 1.000892333984375, 1.001080810546875, 1.00067431640625, 1.0010296020507812, 1.0010767822265625, 1.0005595703125, 1.0013173828125, 1.0010501098632814, 1.0013409423828126, 1.0012006225585937, 1.0011443481445312, 1.0012139282226562, 1.0008053588867187, 1.0007408447265624, 1.0009384765625, 1.0005883178710937, 1.0010715942382813, 1.0009569091796875, 1.0010900268554688, 1.0009027099609376, 1.000849365234375, 1.0009651489257811, 1.0008524780273438, 1.001027587890625, 1.0014033813476562, 1.0012498168945312, 1.00135009765625, 1.001059326171875, 1.0014095458984376, 1.0020269775390624, 1.0020638427734374, 1.0020095825195312, 1.0013235473632813, 1.0012037353515626, 1.0014484252929687, 1.0009476928710936, 1.0012733154296876, 2.066282470703125, 1.0010859375, 1.0004767456054688, 1.0010264892578125, 1.0008135375976563, 1.0007572631835937, 1.0011146240234374, 1.0009620361328124, 1.0013931274414063, 1.0009630737304687, 1.0009375, 1.0007900390625, 1.0008340454101563, 1.0009886474609375, 1.0007900390625, 1.0007756958007812, 1.0007982177734376, 1.0019952392578124, 1.0009794311523437, 1.0006138916015626, 1.0008309936523438, 1.0008893432617187, 1.0010368041992188, 1.0011248779296875, 1.0005238037109374, 1.0006343383789063, 1.0008145751953126, 1.0007039794921875, 1.0007521362304688, 1.0009896850585938, 1.0010880126953126, 1.0012303466796875, 1.00091796875, 1.0013450317382813, 1.000784912109375, 1.0007367553710937, 1.0009825439453126, 1.0009108276367187, 1.001238525390625, 1.0008944702148437, 1.0009989013671876, 1.0009886474609375, 1.0009651489257811, 1.0011422729492188, 1.0012374877929688, 1.0012569580078126, 1.00099072265625, 1.0008780517578124, 1.0009682006835938, 1.0007080688476562, 1.001248779296875, 1.0015518798828125, 1.0010582885742187, 1.0014033813476562, 1.001016357421875, 1.0009476318359376, 1.0012354736328124, 1.0020833129882813, 1.0019143676757813, 1.0018539428710938, 1.0019850463867188, 1.0015999755859375, 1.0015641479492188, 2.0665712890625, 1.0007060546875, 1.0006599731445311, 1.00056884765625, 1.000685546875, 1.0005678100585937, 1.0009722900390625, 1.000806396484375, 1.0007203979492187, 1.0008207397460938, 1.0005872802734375, 1.0005545043945312, 1.00069384765625, 1.0008534545898438, 1.0013716430664064, 1.0008811645507814, 1.0013511962890624, 1.0014893798828124, 1.0010726318359375, 1.0017157592773438, 1.0011299438476562, 1.00086376953125, 1.00071728515625, 1.0005729370117187, 1.0008995971679688, 1.0005995483398438, 1.0024898681640626, 1.0008678588867188, 1.000722412109375, 1.0012149658203124, 1.0008237915039062, 1.00071728515625, 1.00101220703125, 1.0010562744140625, 1.0010572509765625, 1.00170751953125, 1.00140234375, 1.000974365234375, 1.0007828369140626, 1.0009139404296874, 1.0011300048828125, 1.0010675048828126, 1.0013562622070313, 1.000953857421875, 1.0009722900390625, 1.0011146240234374, 1.0015631103515625, 1.000975341796875, 1.0017495727539063, 1.0013224487304688, 1.0012518310546874, 1.001439208984375, 1.0012415771484375, 1.0010685424804688, 1.0010286254882812, 1.00105419921875, 1.0016696166992187, 1.0011002807617186, 1.0011443481445312, 1.0017423095703124, 1.0011033325195313, 1.0011924438476563, 1.0012406005859376]",tokens/s,0.9838699688049951,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-125m,,cuda,0,42,,,,,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,,MB,1310.69952,921.174016,0.0,274.726912,220.646912,s,10,0.36048070526123044,0.03604807052612304,0.0010977874344766398,0.03588955307006836,0.036477619552612306,0.03779876155853271,0.03885567516326904,"[0.03911990356445313, 0.03553478240966797, 0.03469347381591797, 0.035817951202392576, 0.036184032440185546, 0.035873409271240234, 0.03543529510498047, 0.03591993713378906, 0.03590569686889648, 0.03599622344970703]",tokens/s,7101.628360787961,kWh,4.1762897090252873e-07,2.2884077887720709e-07,9.141624902616015e-07,1.5606322400413373e-06,tokens/kWh,164036083.2179266,MB,1311.0272,921.174016,0.0,274.726912,250.723328,s,10,21.866080322265628,2.186608032226563,0.01854961703991556,2.191078857421875,2.2041080810546876,2.2073608032226564,2.2099629809570316,"[2.1803974609375, 2.17111474609375, 2.154228759765625, 2.200171142578125, 2.210613525390625, 2.181986572265625, 2.161680419921875, 2.20071923828125, 2.20338525390625, 2.201783203125]",tokens/s,28.81174818325753,kWh,2.578141039891902e-05,1.4128932798644376e-05,5.28597971955373e-05,9.27701403931007e-05,tokens/kWh,679097.8188999841,,s,629,22.151173107147223,0.03521649142630718,0.004356343174673542,0.03486105728149414,0.035241778564453125,0.03557929000854492,0.07092830078125002,"[0.03474124908447265, 0.03529420852661133, 0.03632339096069336, 0.03650860977172852, 0.036083713531494144, 0.037152767181396484, 0.03645132827758789, 0.036311038970947264, 0.03472588729858399, 0.0354252815246582, 0.03531980895996094, 0.03479657745361328, 0.03465929412841797, 0.03616358566284179, 0.035585025787353515, 0.03515903854370117, 0.03486105728149414, 0.034651134490966795, 0.034900993347167966, 0.03449958419799805, 0.03328102493286133, 0.03349708938598633, 0.033560577392578124, 0.03324620819091797, 0.033328128814697267, 0.033255424499511715, 0.033754112243652344, 0.033223743438720706, 0.03327481460571289, 0.033438720703125, 0.0335175666809082, 0.03376332855224609, 0.03339468765258789, 0.03345305633544922, 0.03355033493041992, 0.034203647613525394, 0.03342233657836914, 0.03335372924804687, 0.0336732177734375, 0.03367116928100586, 0.03360870361328125, 0.03362508773803711, 0.03436441421508789, 0.03604377746582031, 0.03504127883911133, 0.03473920059204102, 0.0348037109375, 0.03472895812988281, 0.03472895812988281, 0.03479865646362305, 0.03479955291748047, 0.03479449462890625, 0.03474739074707031, 0.03478220748901367, 0.0347064323425293, 0.034907135009765625, 0.03474431991577148, 0.03473715209960938, 0.03485388946533203, 0.034683902740478514, 0.035019775390625, 0.0350750732421875, 0.0711178207397461, 0.0347955207824707, 0.034948097229003904, 0.03439616012573242, 0.03447808074951172, 0.03496755218505859, 0.034920448303222655, 0.0348590087890625, 0.034699264526367186, 0.0346163215637207, 0.03491430282592774, 0.03490304183959961, 0.03508019256591797, 0.034423809051513675, 0.03472793579101562, 0.03449446487426758, 0.03476377487182617, 0.03476582336425781, 0.03468902587890625, 0.03387494277954101, 0.0342210578918457, 0.034008094787597656, 0.03374998474121094, 0.03369062423706055, 0.033459201812744144, 0.03354937744140625, 0.03469715118408203, 0.03489484786987305, 0.03487027359008789, 0.03510067367553711, 0.03503513717651367, 0.03503923034667969, 0.03469619369506836, 0.03455590438842773, 0.03517852783203125, 0.03441353607177734, 0.03464704132080078, 0.03492659378051758, 0.034506752014160154, 0.03476070404052734, 0.03487846374511719, 0.03493478393554687, 0.03499212646484375, 0.034945056915283206, 0.035192798614501954, 0.03450265502929688, 0.03448524856567383, 0.03455487823486328, 0.034563072204589845, 0.03503104019165039, 0.03474431991577148, 0.03377872085571289, 0.03339875030517578, 0.03340800094604492, 0.03347558212280274, 0.03381350326538086, 0.033650688171386715, 0.033535999298095705, 0.033334270477294925, 0.03363123321533203, 0.03390771102905273, 0.03361996841430664, 0.03386880111694336, 0.06886093139648437, 0.03341823959350586, 0.03395993423461914, 0.03393228912353516, 0.03374387359619141, 0.03333529663085937, 0.0334919662475586, 0.03423027038574219, 0.03382476806640625, 0.03374387359619141, 0.0336824951171875, 0.033724353790283206, 0.03363532638549805, 0.033941505432128906, 0.03381760025024414, 0.033503231048583985, 0.03344998550415039, 0.033301502227783206, 0.033445888519287106, 0.033957889556884766, 0.03387801742553711, 0.03366092681884766, 0.03360153579711914, 0.03381043243408203, 0.033825790405273434, 0.03360870361328125, 0.03370703887939453, 0.03348271942138672, 0.03366502380371094, 0.03378995132446289, 0.03319193649291992, 0.03369267272949219, 0.03357900619506836, 0.033358848571777344, 0.0341104621887207, 0.03495731353759766, 0.03543961715698242, 0.03482316970825195, 0.034435073852539064, 0.034557952880859374, 0.034900993347167966, 0.034678783416748044, 0.03517542266845703, 0.03521023941040039, 0.035046398162841795, 0.03481702423095703, 0.03453235244750977, 0.03470131301879883, 0.03486412811279297, 0.03474739074707031, 0.03477196884155274, 0.03457843017578125, 0.034781185150146485, 0.03486310577392578, 0.035125247955322264, 0.03486617660522461, 0.03496243286132812, 0.03503513717651367, 0.03523891067504883, 0.03469823837280273, 0.0339128303527832, 0.03414220809936523, 0.035117088317871095, 0.0711659164428711, 0.03465318298339844, 0.03465727996826172, 0.03458355331420898, 0.034283519744873044, 0.034304000854492187, 0.03577958297729492, 0.03595161437988281, 0.03544473648071289, 0.03500543975830078, 0.03521843338012695, 0.03495935821533203, 0.0349194221496582, 0.0354252815246582, 0.03515903854370117, 0.0349306869506836, 0.03492454528808594, 0.036211711883544925, 0.03564441680908203, 0.03498086547851562, 0.034841598510742186, 0.034941951751708986, 0.03537612915039062, 0.034994174957275394, 0.0350382080078125, 0.03486105728149414, 0.03484672164916992, 0.03494297790527344, 0.03524095916748047, 0.03526041412353516, 0.03520614242553711, 0.03488153457641602, 0.03502796936035156, 0.03500543975830078, 0.03404185485839844, 0.033903617858886716, 0.03378073501586914, 0.03472281646728516, 0.03411251068115234, 0.03379097747802735, 0.03348787307739258, 0.0343633918762207, 0.03492659378051758, 0.03500339126586914, 0.03493788909912109, 0.03485590362548828, 0.034895870208740236, 0.034948097229003904, 0.03518771362304687, 0.034479103088378905, 0.03459174346923828, 0.035253280639648436, 0.034963424682617185, 0.034976768493652347, 0.03491843032836914, 0.034911201477050784, 0.035046398162841795, 0.03507199859619141, 0.03536383819580078, 0.03499929428100586, 0.03523379135131836, 0.03531161499023437, 0.03527679824829102, 0.07160934448242187, 0.0352624626159668, 0.03535769653320313, 0.03481087875366211, 0.03554921722412109, 0.034968544006347656, 0.03489491271972656, 0.034890689849853516, 0.03483238220214844, 0.034993152618408206, 0.035932159423828124, 0.03508838272094727, 0.03496755218505859, 0.03476377487182617, 0.03488358306884766, 0.03527884674072266, 0.03521228790283203, 0.035040256500244144, 0.03476684951782227, 0.03499622344970703, 0.03482316970825195, 0.03518259048461914, 0.03503411102294922, 0.036178943634033206, 0.039564289093017575, 0.035097599029541016, 0.03457843017578125, 0.03505561447143555, 0.034756607055664065, 0.0350013427734375, 0.03477811050415039, 0.035004417419433595, 0.03501772689819336, 0.0348671989440918, 0.03473100662231445, 0.0349224967956543, 0.03482931137084961, 0.03578777694702148, 0.0350115852355957, 0.03483443069458008, 0.03486412811279297, 0.03486207962036133, 0.034938880920410156, 0.03505561447143555, 0.035004417419433595, 0.034887680053710936, 0.034531326293945314, 0.03493580627441406, 0.03510067367553711, 0.03492454528808594, 0.034887680053710936, 0.03578879928588867, 0.034974720001220705, 0.03519180679321289, 0.034850879669189455, 0.03495315170288086, 0.035060768127441404, 0.034809825897216796, 0.03377766418457031, 0.03550620651245117, 0.034968544006347656, 0.034915328979492184, 0.03489177703857422, 0.07134719848632813, 0.03493273544311523, 0.03501465606689453, 0.03539148712158203, 0.03512319946289062, 0.034813953399658204, 0.03508019256591797, 0.03486515045166016, 0.035079166412353514, 0.03493580627441406, 0.034939903259277344, 0.034925567626953126, 0.03743027114868164, 0.034576385498046876, 0.03397836685180664, 0.03377356719970703, 0.03403878402709961, 0.03371724700927734, 0.03380223846435547, 0.03486310577392578, 0.03486412811279297, 0.03496243286132812, 0.03475046539306641, 0.03492659378051758, 0.03499622344970703, 0.034783233642578126, 0.033949695587158206, 0.033876991271972655, 0.03427123260498047, 0.03477811050415039, 0.03472076797485352, 0.034127872467041014, 0.03329740905761719, 0.0333834228515625, 0.03379814529418945, 0.03392409515380859, 0.03363532638549805, 0.03386982345581055, 0.03498086547851562, 0.034928638458251955, 0.03496345520019531, 0.03495423889160156, 0.03523891067504883, 0.03508838272094727, 0.034769920349121096, 0.03489279937744141, 0.03482624053955078, 0.034955265045166016, 0.03508838272094727, 0.03486412811279297, 0.03486515045166016, 0.03471769714355469, 0.034560001373291016, 0.034579456329345705, 0.03481292724609375, 0.03479244613647461, 0.034802688598632815, 0.03475254440307617, 0.03517948913574219, 0.03433884811401367, 0.033702880859375, 0.03359539031982422, 0.034111488342285154, 0.07044096374511719, 0.03437670516967774, 0.03483238220214844, 0.0348671989440918, 0.03470438385009766, 0.035230720520019534, 0.034816001892089846, 0.03467366409301758, 0.0347770881652832, 0.03501055908203125, 0.03429580688476563, 0.03443609619140625, 0.03509964752197266, 0.03524505615234375, 0.0347248649597168, 0.034753536224365236, 0.03496345520019531, 0.03496448135375976, 0.034770942687988284, 0.034799617767333986, 0.03484467315673828, 0.03481190490722656, 0.034353153228759765, 0.034062335968017575, 0.03349094390869141, 0.033584129333496096, 0.033587200164794925, 0.034141185760498044, 0.0347740478515625, 0.0345906867980957, 0.0344268798828125, 0.03489484786987305, 0.03420979309082031, 0.03364352035522461, 0.03377356719970703, 0.033686527252197264, 0.033290241241455076, 0.0337520637512207, 0.03401932907104492, 0.03371417617797851, 0.03375513458251953, 0.033691646575927735, 0.03370809555053711, 0.03350316619873047, 0.03325439834594727, 0.03360255813598633, 0.033860607147216795, 0.03376230239868164, 0.033691646575927735, 0.033783809661865234, 0.033797119140625, 0.03425177764892578, 0.03381452941894531, 0.033791999816894534, 0.033675262451171875, 0.03391385650634766, 0.034165760040283204, 0.03502899169921875, 0.03486822509765625, 0.034702335357666016, 0.035046398162841795, 0.03521535873413086, 0.034825214385986326, 0.07150080108642579, 0.03484364700317383, 0.034351104736328124, 0.03453747177124023, 0.035064830780029296, 0.03482931137084961, 0.03483238220214844, 0.03492659378051758, 0.034974720001220705, 0.03501465606689453, 0.034869247436523435, 0.034830337524414064, 0.034277374267578126, 0.034797569274902344, 0.03501363372802734, 0.03530342483520508, 0.0351016960144043, 0.03491635131835937, 0.03483238220214844, 0.03484672164916992, 0.03495116806030273, 0.03500646209716797, 0.03499827194213867, 0.03480678558349609, 0.03486207962036133, 0.03505152130126953, 0.034784255981445314, 0.034783233642578126, 0.034835456848144535, 0.03493580627441406, 0.034770942687988284, 0.035037185668945314, 0.035019775390625, 0.0342476806640625, 0.03505049514770508, 0.034315265655517575, 0.03477913665771484, 0.034791454315185544, 0.0353023681640625, 0.03562700653076172, 0.03510374450683594, 0.03493580627441406, 0.035156993865966796, 0.03526144027709961, 0.03488256072998047, 0.034933761596679686, 0.03517337417602539, 0.0348590087890625, 0.0352911376953125, 0.035888126373291016, 0.03522048187255859, 0.03506073760986328, 0.03483955383300781, 0.03486105728149414, 0.03498400115966797, 0.03498387145996094, 0.03485494232177734, 0.03438998413085938, 0.03479244613647461, 0.03507712173461914, 0.03500339126586914, 0.03471257781982422, 0.03476684951782227, 0.07122022247314454, 0.03482316970825195, 0.03523276901245117, 0.034941951751708986, 0.03477196884155274, 0.034991104125976565, 0.03495731353759766, 0.034939903259277344, 0.03488153457641602, 0.034900993347167966, 0.0347955207824707, 0.0350300178527832, 0.03487948989868164, 0.03520614242553711, 0.035297279357910154, 0.03513651275634765, 0.03494297790527344, 0.0349378547668457, 0.034948097229003904, 0.03488051223754883, 0.03481497573852539, 0.03454873657226563, 0.03495935821533203, 0.035062782287597655, 0.035043327331542966, 0.0349306869506836, 0.03497881698608398, 0.034909183502197266, 0.03493791961669922, 0.03483539199829101, 0.034869247436523435, 0.035031105041503904, 0.03493983840942383, 0.03483340835571289, 0.03487744140625, 0.03489484786987305, 0.03516108703613281, 0.03484364700317383, 0.034969600677490234, 0.035146751403808595, 0.035194881439208986, 0.03479244613647461, 0.03557068634033203, 0.03522969436645508, 0.035337215423583986, 0.03480473709106445, 0.03436236953735351, 0.033745918273925785, 0.03496755218505859, 0.03541401672363281, 0.03512934494018555, 0.035163135528564454, 0.0348487663269043, 0.03500339126586914, 0.03484467315673828, 0.03498495864868164, 0.03576115036010742, 0.03517030334472656, 0.03490816116333008, 0.034825214385986326, 0.0347586555480957, 0.03516723251342774, 0.035053569793701174, 0.07151315307617187, 0.03519071960449219, 0.03483852767944336, 0.03482419204711914, 0.03502592086791992, 0.03529216003417969, 0.035125247955322264, 0.034872318267822264, 0.034939903259277344, 0.0347658576965332, 0.03481699371337891, 0.03497267150878906, 0.03488665771484375, 0.034958335876464845, 0.035007488250732424, 0.03504537582397461, 0.03489382553100586, 0.03520716857910156, 0.034988033294677735, 0.034874366760253905, 0.03459481430053711, 0.03430297470092773, 0.0346879997253418, 0.03478732681274414, 0.03508736038208008, 0.0354856948852539, 0.034981952667236325, 0.0348732795715332, 0.03489382553100586, 0.03475558471679688, 0.03478732681274414, 0.03488358306884766, 0.03473920059204102, 0.03476172637939453, 0.034685951232910156, 0.035111934661865234, 0.03508736038208008, 0.03526553726196289, 0.03492147064208984, 0.03475763320922851, 0.034767871856689454, 0.03482316970825195, 0.03498092651367188, 0.03539142227172851, 0.03607961654663086, 0.03499008178710938, 0.03491123199462891, 0.034990142822265625, 0.03488249588012695, 0.03484467315673828, 0.03486207962036133, 0.0349378547668457, 0.034953216552734374, 0.03488051223754883, 0.03490816116333008, 0.0349378547668457, 0.03494604873657227, 0.03508838272094727, 0.03482726287841797, 0.03486003112792969, 0.03481702423095703, 0.034729984283447264, 0.0349224967956543]",tokens/s,28.395787300179112,,,main,False,False,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-13b,facebook/opt-13b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,2935.840768,9548.857344,0.0,8902.41024,8265.321472,s,10,10.732376098632812,1.0732376098632814,0.0018882051518454717,1.0727980346679686,1.0744577026367188,1.076272918701172,1.0777250915527343,"[1.078088134765625, 1.0709527587890626, 1.071809814453125, 1.072758056640625, 1.0723739013671876, 1.0717447509765625, 1.0728380126953125, 1.07379296875, 1.0740543212890625, 1.07396337890625]",tokens/s,238.53058972897117,kWh,1.2652489013142056e-05,6.933077765861525e-06,5.8725796980607646e-05,7.831136375961123e-05,tokens/kWh,3269001.939307702,MB,2940.198912,9548.857344,0.0,8902.41024,8556.582912,s,10,636.0469335937499,63.60469335937499,0.008753951962243356,63.60478125,63.61318203125,63.615917187499996,63.6181053125,"[63.59721875, 63.61257421875, 63.61096484375, 63.6103046875, 63.61865234375, 63.5982421875, 63.6112421875, 63.58923828125, 63.5992578125, 63.59923828125]",tokens/s,0.9904929443499021,kWh,0.0007508853809701072,0.00041155162046943583,0.003498196520777186,0.00466063352221673,tokens/kWh,13517.475617785843,,s,629,644.6920160522462,1.0249475612913292,0.12743236447478887,1.0094981079101562,1.01038447265625,1.0106333129882812,2.0813571875,"[1.0103838500976563, 1.011209228515625, 1.0106275634765625, 1.010282470703125, 1.0090455322265626, 1.00902294921875, 1.0092001342773438, 1.0092001342773438, 1.0087454833984375, 1.0091202392578125, 1.0090209350585937, 1.0097459106445312, 1.009122314453125, 1.0094325561523438, 1.0090762329101564, 1.008990234375, 1.009164306640625, 1.0090936279296876, 1.0088233032226563, 1.009649658203125, 1.0089727783203124, 1.0095328979492189, 1.009112060546875, 1.0095298461914062, 1.0089553833007812, 1.0094817504882811, 1.0095001831054689, 1.0089779052734376, 1.0089512939453125, 1.0092308349609376, 1.0093475952148439, 1.0095062866210938, 1.0100172729492187, 1.0097244262695313, 1.0091735229492187, 1.0095360107421876, 1.0095134887695312, 1.0093209838867188, 1.009239013671875, 1.00967529296875, 1.0096640014648437, 1.00973974609375, 1.0092687377929688, 1.009269775390625, 1.0091714477539062, 1.0091786499023438, 1.0094950561523437, 1.0094622802734374, 1.0091345825195313, 1.009670166015625, 1.0100039672851562, 1.0096732177734375, 1.0092994384765626, 1.0093557739257812, 1.0091253662109374, 1.00938134765625, 1.0091427612304686, 1.0090465087890625, 1.0089072875976564, 1.0098462524414062, 1.010165771484375, 1.0093465576171874, 2.08412060546875, 1.0087465209960937, 1.0093834228515626, 1.0092031860351562, 1.0096558227539063, 1.0095739135742188, 1.0104360961914063, 1.0100264892578126, 1.0097346801757812, 1.0098964233398438, 1.0101534423828125, 1.00982373046875, 1.009934326171875, 1.0099998779296875, 1.0098063354492188, 1.0099056396484376, 1.0100838623046875, 1.0100254516601563, 1.00965478515625, 1.0096107788085937, 1.0098421630859375, 1.0097664184570312, 1.0100264892578126, 1.0098104248046875, 1.0099118041992188, 1.0101432495117189, 1.010134033203125, 1.0106337280273439, 1.0090772705078126, 1.0097838134765624, 1.0102476806640626, 1.0099251098632813, 1.0096466064453125, 1.00933837890625, 1.0092185668945313, 1.0092902221679687, 1.0095257568359375, 1.00948583984375, 1.0100172729492187, 1.010408447265625, 1.0100695190429687, 1.0098995361328125, 1.0105394897460938, 1.00971826171875, 1.0095902709960938, 1.0096414794921875, 1.009565673828125, 1.00931787109375, 1.0096066284179688, 1.0097407836914063, 1.009343505859375, 1.0096947021484375, 1.0099138793945313, 1.009455078125, 1.0092472534179688, 1.0096680908203124, 1.0096250610351563, 1.0100101318359376, 1.0093414306640625, 1.0092892456054687, 1.009311767578125, 1.009322998046875, 1.0097203369140626, 2.081078369140625, 1.009328125, 1.0090977172851563, 1.0092052612304687, 1.0094520263671876, 1.0091028442382812, 1.0091796264648438, 1.0086420288085938, 1.0094274291992187, 1.0089891967773437, 1.0090137329101563, 1.0094161987304688, 1.0096568603515625, 1.0095421142578125, 1.009850341796875, 1.0090895385742187, 1.0094059448242187, 1.009059814453125, 1.0094509887695313, 1.0088161010742187, 1.008974853515625, 1.0091888427734375, 1.0095595703125, 1.009016845703125, 1.0097346801757812, 1.0089810180664063, 1.00931787109375, 1.0090147705078125, 1.0099138793945313, 1.0100213623046874, 1.0104381713867188, 1.010629638671875, 1.010713623046875, 1.0101360473632812, 1.0105651245117186, 1.0090690307617187, 1.0092144775390626, 1.009344482421875, 1.01054052734375, 1.0098954467773438, 1.0101473388671875, 1.010071533203125, 1.0109757690429688, 1.0099885864257812, 1.010386962890625, 1.0099199829101562, 1.0098933715820313, 1.0100418701171876, 1.0103040161132812, 1.0101207275390625, 1.0105702514648438, 1.0102599487304686, 1.00997119140625, 1.0098646850585937, 1.0102732543945312, 1.0096066284179688, 1.00906494140625, 1.0095984497070312, 1.0102210693359375, 1.0098585815429688, 1.0101514282226562, 1.010039794921875, 1.0104360961914063, 2.082943115234375, 1.009787841796875, 1.0099476318359375, 1.0097244262695313, 1.0104217529296875, 1.0104340209960938, 1.0096947021484375, 1.0098309326171875, 1.009544189453125, 1.0100009155273437, 1.01024560546875, 1.0097223510742188, 1.0100695190429687, 1.0102835083007813, 1.009902587890625, 1.0104699096679688, 1.0107535400390626, 1.0102671508789063, 1.010640869140625, 1.0102630615234376, 1.0101463012695313, 1.0099834594726562, 1.0103336791992188, 1.0099384155273436, 1.0103685302734375, 1.0097254638671875, 1.0096906127929688, 1.0098125, 1.0103756713867187, 1.0102271728515626, 1.010640869140625, 1.0090096435546876, 1.0095994873046874, 1.0097633056640625, 1.0101729125976562, 1.0090198974609375, 1.0095718383789063, 1.0087034912109376, 1.009523681640625, 1.00916943359375, 1.0090352783203125, 1.0087833862304687, 1.0096363525390626, 1.0092359619140625, 1.0092656860351563, 1.0091233520507812, 1.00946533203125, 1.008932861328125, 1.009027099609375, 1.0091468505859376, 1.0092257080078124, 1.0089072875976564, 1.008837646484375, 1.0091827392578125, 1.0094110717773437, 1.0090751953125, 1.0094848022460938, 1.0091868286132812, 1.009480712890625, 1.0097039184570313, 1.0093670654296876, 1.0094059448242187, 1.0094376831054688, 2.081429443359375, 1.009227783203125, 1.0094479370117186, 1.010234375, 1.0106552124023438, 1.010882568359375, 1.0105589599609375, 1.0087034912109376, 1.0110679321289062, 1.0107822265625, 1.0108876953125, 1.0104412231445312, 1.0089584350585938, 1.0093711547851563, 1.0090792846679688, 1.0109276123046875, 1.0108334350585937, 1.0103521118164063, 1.0105712890625, 1.0102968139648438, 1.0089482421875, 1.0090025024414062, 1.0100203247070312, 1.0104258422851562, 1.009818603515625, 1.0096087036132813, 1.0098165893554687, 1.0094642944335936, 1.0101575927734374, 1.0103203735351562, 1.0114457397460936, 1.0106500854492189, 1.010361328125, 1.010103271484375, 1.0099415283203126, 1.0093311767578126, 1.0099465942382813, 1.0090475463867188, 1.0093302001953126, 1.0094981079101562, 1.0090823974609375, 1.0102753295898437, 1.0096322631835937, 1.0092564697265625, 1.0091519775390625, 1.0096640014648437, 1.009281005859375, 1.0090864868164062, 1.0091448364257813, 1.0099537963867187, 1.0093250732421875, 1.0095186157226563, 1.0094940185546875, 1.00933837890625, 1.0094080200195312, 1.0091744995117187, 1.0092451782226564, 1.0095390625, 1.0101217041015624, 1.01006640625, 1.0101780395507813, 1.009744873046875, 1.0093660278320313, 2.08117138671875, 1.0091130981445313, 1.009100830078125, 1.0086492309570312, 1.0088734741210938, 1.0091294555664063, 1.0090188598632812, 1.0089410400390626, 1.009344482421875, 1.0095523681640626, 1.0093916015625, 1.0090680541992187, 1.00940185546875, 1.009154052734375, 1.0091970825195313, 1.0093629150390624, 1.0088673095703125, 1.0088601684570313, 1.009817626953125, 1.009006591796875, 1.0100746459960936, 1.009349609375, 1.0093670654296876, 1.0091519775390625, 1.0093035278320313, 1.0091581420898437, 1.0092021484375, 1.0091058959960937, 1.01058251953125, 1.0097705078125, 1.0106183471679688, 1.0098779907226563, 1.0096998291015624, 1.0104105224609374, 1.0094213256835938, 1.009344482421875, 1.0096097412109375, 1.0097407836914063, 1.0099261474609376, 1.009217529296875, 1.0095718383789063, 1.0091837158203125, 1.0093905639648437, 1.0092267456054687, 1.0091878662109375, 1.0096301879882812, 1.0098125, 1.0094929809570312, 1.009860595703125, 1.0093219604492187, 1.0101544799804687, 1.009301513671875, 1.0094714965820313, 1.009460205078125, 1.0093588256835937, 1.0097899780273438, 1.0095984497070312, 1.0096742553710938, 1.010693115234375, 1.0096957397460937, 1.010155517578125, 1.010134033203125, 1.00997119140625, 2.082757568359375, 1.0091796264648438, 1.0095452270507812, 1.0092155151367188, 1.0096373901367188, 1.00949609375, 1.0099005737304687, 1.0090813598632813, 1.009080322265625, 1.008932861328125, 1.0091448364257813, 1.008964599609375, 1.00949609375, 1.0093455200195312, 1.0097909545898438, 1.0094417724609375, 1.00999267578125, 1.0092533569335937, 1.009428466796875, 1.0093148193359376, 1.0098770141601563, 1.0099015502929687, 1.0105159912109376, 1.0107003173828124, 1.0108948364257813, 1.009523681640625, 1.010176025390625, 1.009207275390625, 1.0096517333984374, 1.0096107788085937, 1.010208740234375, 1.0101217041015624, 1.0100725708007812, 1.0099415283203126, 1.01014013671875, 1.009311767578125, 1.0097100830078125, 1.0091714477539062, 1.0097858276367186, 1.0092830810546876, 1.0094745483398437, 1.009259521484375, 1.0102118530273438, 1.0100234375, 1.0097469482421875, 1.010239501953125, 1.0100633544921875, 1.0095165405273439, 1.0095984497070312, 1.0094940185546875, 1.0099507446289062, 1.0094642944335936, 1.0101094360351563, 1.0097879028320313, 1.0109890747070311, 1.0111426391601563, 1.0094315795898436, 1.0090772705078126, 1.009227783203125, 1.0092001342773438, 1.009723388671875, 1.00982373046875, 1.0097694702148436, 2.08317333984375, 1.0093322143554688, 1.00935986328125, 1.0087188720703124, 1.0093055419921875, 1.008848876953125, 1.0092216186523437, 1.0090895385742187, 1.0095001831054689, 1.0094622802734374, 1.0091847534179688, 1.0095789794921874, 1.0093875122070313, 1.0095748901367188, 1.0098646850585937, 1.0092850952148438, 1.0101329956054688, 1.0093475952148439, 1.0094888916015625, 1.00940185546875, 1.0091079711914062, 1.0090505981445312, 1.0093823852539063, 1.0089502563476562, 1.009269775390625, 1.0090301513671875, 1.00952783203125, 1.0098544921875, 1.0100684814453125, 1.0095472412109374, 1.0094541015625, 1.0092247314453124, 1.0096537475585938, 1.0091100463867186, 1.0094458618164062, 1.0091038818359375, 1.0096301879882812, 1.0089943237304688, 1.0096271362304687, 1.009623046875, 1.0094694213867188, 1.009554443359375, 1.0093568115234375, 1.0092083129882812, 1.0091427612304686, 1.0093660278320313, 1.010423828125, 1.0095114135742187, 1.0090711059570312, 1.0094817504882811, 1.009312744140625, 1.0092687377929688, 1.009132568359375, 1.0089052124023437, 1.0089922485351563, 1.0089625854492188, 1.0091192016601562, 1.00922265625, 1.0094295043945312, 1.0091376342773437, 1.0093772583007812, 1.0091202392578125, 1.0095810546875, 2.08344580078125, 1.0095042724609375, 1.0109685668945312, 1.0099476928710938, 1.01062548828125, 1.0095011596679688, 1.0104494018554688, 1.008996337890625, 1.0095462646484374, 1.008911376953125, 1.00917041015625, 1.0089707641601562, 1.00910693359375, 1.0093240356445312, 1.0093403930664062, 1.0089093017578126, 1.009713134765625, 1.0089676513671875, 1.0090844116210937, 1.0088786010742188, 1.00897998046875, 1.0090782470703126, 1.0091878662109375, 1.0094356689453126, 1.009923095703125, 1.0092708129882813, 1.0097356567382811, 1.008616455078125, 1.0090946655273438, 1.0091868286132812, 1.0094305419921874, 1.0093619384765624, 1.0100254516601563, 1.0098093872070313, 1.0101801147460938, 1.0097049560546876, 1.010398193359375, 1.0095513305664063, 1.0100848388671875, 1.0106326904296874, 1.011041259765625, 1.010619384765625, 1.0100930786132813, 1.0094878540039063, 1.0097684326171874, 1.0093609008789062, 1.0095360107421876, 1.009154052734375, 1.0090618896484376, 1.009249267578125, 1.009196044921875, 1.0090895385742187, 1.0093352661132813, 1.0091714477539062, 1.0094796752929687, 1.0091049194335937, 1.0094694213867188, 1.0094202880859375, 1.009227783203125, 1.0093311767578126, 1.0093629150390624, 1.0093004760742188, 1.0094376831054688, 2.083451904296875, 1.0096097412109375, 1.0098380737304689, 1.009165283203125, 1.0100848388671875, 1.0104432373046874, 1.0090741577148437, 1.0095851440429688, 1.0092666625976563, 1.0088263549804688, 1.0094428100585937, 1.0090045166015624, 1.0094745483398437, 1.0092830810546876, 1.0091581420898437, 1.0089246826171876, 1.0092728271484375, 1.009238037109375, 1.0095165405273439, 1.0092728271484375, 1.009455078125, 1.00973876953125, 1.009924072265625, 1.0094458618164062, 1.0094541015625, 1.0086737670898438, 1.009306640625, 1.009659912109375, 1.0100695190429687, 1.0101934204101561, 1.0098739013671876, 1.0097745971679688, 1.0097152099609374, 1.0097776489257813, 1.0097192993164064, 1.0093025512695313, 1.0096865234375, 1.009438720703125, 1.0101616821289063, 1.0098831176757812, 1.0098104248046875, 1.0097673950195312, 1.01049853515625, 1.0099834594726562, 1.0104944458007812, 1.0097633056640625, 1.0092708129882813, 1.009201171875, 1.0093056030273437, 1.0092236938476562, 1.0090127563476563, 1.0091929321289062, 1.00919091796875, 1.0090670166015625, 1.0091888427734375, 1.008996337890625, 1.0091161499023438, 1.0092789916992186, 1.0097572021484376, 1.0099056396484376, 1.0091837158203125, 1.0094663696289063, 1.0093506469726563]",tokens/s,0.9756596705690017,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/recurrentgemma-7b,google/recurrentgemma-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 404 Client Error: Not Found for url: https://huggingface.co/google/recurrentgemma-7b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1823, in _raise_on_head_call_error + raise head_call_error + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 352, in hf_raise_for_status + raise RepositoryNotFoundError(message, response) from e +huggingface_hub.utils._errors.RepositoryNotFoundError: 404 Client Error. (Request ID: Root=1-664aaa90-0a22c30224ea2bc804eea870;cc92f1c3-cd4c-4719-ba14-ff7660a5b350) + +Repository Not Found for url: https://huggingface.co/google/recurrentgemma-7b/resolve/main/config.json. +Please make sure you specified the correct `repo_id` and `repo_type`. +If you are trying to access a private or gated repo, make sure you are authenticated. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 422, in cached_file + raise EnvironmentError( +OSError: google/recurrentgemma-7b is not a local folder and is not a valid model identifier listed on 'https://huggingface.co/models' +If this is a private repository, make sure to pass a token having permission to this repo either by logging in with `huggingface-cli login` or by passing `token=` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-410m,EleutherAI/pythia-410m,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,1390.702592,1340.604416,0.0,694.157312,598.499328,s,10,0.6226632919311523,0.062266329193115234,0.002317832326909016,0.06196879959106445,0.06357897109985351,0.06612861976623535,0.06816833869934083,"[0.06867826843261719, 0.06029948806762695, 0.06301238250732422, 0.06221964645385742, 0.06201440048217773, 0.06192319869995117, 0.06178224182128906, 0.06027980804443359, 0.06031568145751953, 0.06213817596435547]",tokens/s,4111.3713192571795,kWh,7.181983443623039e-07,3.9353970671072603e-07,2.1138383914060147e-06,3.2255764424790445e-06,tokens/kWh,79365658.99621,MB,1391.030272,1340.604416,0.0,694.157312,659.032576,s,10,37.91430688476563,3.791430688476563,0.033463632163375424,3.808232421875,3.8238944580078122,3.8254893432617187,3.8267652514648436,"[3.8235400390625, 3.807294921875, 3.809169921875, 3.815871337890625, 3.827084228515625, 3.813744140625, 3.745643798828125, 3.73081494140625, 3.753345703125, 3.7877978515625]",tokens/s,16.616418754924958,kWh,4.459946840148583e-05,2.4442959093412353e-05,0.00012145201046119378,0.00019049443795609197,tokens/kWh,330718.3174267859,,s,629,38.408798217773445,0.06106327220631707,0.007411049818688859,0.06058700942993164,0.06105620422363281,0.061356236267089845,0.1198044711303711,"[0.05935715103149414, 0.059200511932373044, 0.06025625610351563, 0.06658048248291015, 0.06115430450439453, 0.06075187301635742, 0.06081740951538086, 0.06116556930541992, 0.06187519836425781, 0.060846080780029295, 0.0618015022277832, 0.061259742736816405, 0.06081024169921875, 0.06082559967041016, 0.06123110580444336, 0.06096486282348633, 0.06082559967041016, 0.06090752029418945, 0.06103142547607422, 0.05951590347290039, 0.05919027328491211, 0.060930049896240235, 0.06138163375854492, 0.06130278396606445, 0.0615464973449707, 0.061396991729736325, 0.06103859329223633, 0.060919807434082034, 0.06081228637695312, 0.06087680053710937, 0.06095257568359375, 0.06107648086547852, 0.06047641754150391, 0.05975551986694336, 0.060652542114257815, 0.060800033569335936, 0.06089315032958984, 0.060319744110107425, 0.06068940734863281, 0.06150451278686524, 0.06105190277099609, 0.061037567138671874, 0.059990016937255856, 0.060765182495117184, 0.06077849578857422, 0.06057881546020508, 0.060082176208496096, 0.0589219856262207, 0.059184127807617185, 0.05976166534423828, 0.060783649444580076, 0.06094025421142578, 0.06090854263305664, 0.060641281127929686, 0.05964799880981445, 0.05914112091064453, 0.05936537551879883, 0.06067507171630859, 0.06081740951538086, 0.060158977508544924, 0.06099660873413086, 0.05982624053955078, 0.11980588531494141, 0.06112768173217773, 0.06088294219970703, 0.060576766967773435, 0.06073548889160156, 0.06021529769897461, 0.06116966247558594, 0.06180044937133789, 0.06070483016967773, 0.05943494415283203, 0.059440128326416014, 0.060862464904785155, 0.060211200714111325, 0.05941452789306641, 0.06035456085205078, 0.060521472930908204, 0.06083174514770508, 0.06059929656982422, 0.060902400970458986, 0.06084096145629883, 0.06026649475097656, 0.05905920028686523, 0.059622398376464845, 0.06106521606445312, 0.06076313781738281, 0.05981081771850586, 0.05929369735717773, 0.059216896057128904, 0.061050880432128904, 0.06076620864868164, 0.059224063873291016, 0.06019583892822265, 0.06055321502685547, 0.060796928405761716, 0.061064193725585934, 0.06083071899414062, 0.060793857574462894, 0.05939507293701172, 0.058877952575683595, 0.05913702392578125, 0.05904793548583984, 0.05907865524291992, 0.06065049743652344, 0.06088294219970703, 0.06102732849121094, 0.06089215850830078, 0.06101504135131836, 0.060728321075439455, 0.06069760131835938, 0.060909599304199216, 0.06107542419433594, 0.060761089324951174, 0.06076416015625, 0.060911617279052734, 0.06071091079711914, 0.06042009735107422, 0.06087372970581055, 0.06067302322387695, 0.06060851287841797, 0.060984321594238285, 0.06090956878662109, 0.06082867050170898, 0.060943359375, 0.12434432220458984, 0.059063297271728515, 0.05924863815307617, 0.059273216247558595, 0.059202560424804686, 0.0593807373046875, 0.05934284973144531, 0.06072115325927734, 0.060693504333496094, 0.06077439880371094, 0.0608798713684082, 0.05987430572509766, 0.05933158493041992, 0.05966438293457031, 0.05923430252075195, 0.06037811279296875, 0.060916736602783204, 0.06084505462646484, 0.060793857574462894, 0.06016204833984375, 0.06031257629394531, 0.060813312530517576, 0.06093008041381836, 0.06087472152709961, 0.06037811279296875, 0.06069964981079102, 0.06112972640991211, 0.06072524642944336, 0.060851200103759766, 0.06091468811035156, 0.0610334701538086, 0.0608901138305664, 0.06094847869873047, 0.06083071899414062, 0.060846080780029295, 0.06088806533813477, 0.06078566360473633, 0.06029625701904297, 0.06034220886230469, 0.06101094436645508, 0.06097919845581055, 0.061350910186767575, 0.05945548629760742, 0.05929779052734375, 0.059101184844970706, 0.059858943939208986, 0.05947903823852539, 0.05919948959350586, 0.059469825744628904, 0.06209024047851563, 0.06109183883666992, 0.06083071899414062, 0.06089318466186523, 0.06092083358764649, 0.06089113616943359, 0.06088601684570313, 0.06026956939697266, 0.06099763107299805, 0.06088499069213867, 0.06075494384765625, 0.06154547119140625, 0.0613570556640625, 0.06097919845581055, 0.12348928070068359, 0.060916736602783204, 0.060777473449707034, 0.0607303695678711, 0.06054297637939453, 0.06096694564819336, 0.06082352066040039, 0.060835838317871094, 0.06073855972290039, 0.060864513397216796, 0.05947903823852539, 0.059015167236328124, 0.059270145416259766, 0.05909404754638672, 0.05935817718505859, 0.05992550277709961, 0.06107955169677735, 0.061110271453857425, 0.06092083358764649, 0.06084505462646484, 0.06058803176879883, 0.06083891296386719, 0.06079487991333008, 0.060827648162841794, 0.060832767486572265, 0.06084403228759765, 0.06105702209472656, 0.06056345748901367, 0.06058598327636719, 0.06076006317138672, 0.06095462417602539, 0.0609617919921875, 0.06127206420898437, 0.05955788803100586, 0.06099967956542969, 0.06095667266845703, 0.06068633651733398, 0.060887039184570314, 0.06139187240600586, 0.061454334259033204, 0.06054912185668945, 0.06096281433105469, 0.06029414367675781, 0.06135500717163086, 0.062328830718994144, 0.06116147232055664, 0.0603770866394043, 0.05916262435913086, 0.059469825744628904, 0.0593438720703125, 0.05933055877685547, 0.05895884704589844, 0.05933670425415039, 0.06046003341674805, 0.06081024169921875, 0.06076620864868164, 0.061451263427734375, 0.06086348724365234, 0.06087168121337891, 0.06075699234008789, 0.060631038665771485, 0.0605296630859375, 0.06076006317138672, 0.12325580596923828, 0.060826625823974606, 0.06067612838745117, 0.06101910400390625, 0.06079283142089844, 0.06072115325927734, 0.06085529708862305, 0.06092902374267578, 0.0608983039855957, 0.060955646514892575, 0.06091059112548828, 0.060889087677001956, 0.06080716705322266, 0.059873279571533204, 0.05963673782348633, 0.06089932632446289, 0.061224960327148435, 0.06137139129638672, 0.06110105514526367, 0.06112870407104492, 0.06075699234008789, 0.0609249267578125, 0.0607457275390625, 0.06121267318725586, 0.060843006134033206, 0.06098124694824219, 0.061001728057861325, 0.06096588897705078, 0.06096895980834961, 0.06099558258056641, 0.06101708984375, 0.06083174514770508, 0.06094643020629883, 0.06118297576904297, 0.06099148941040039, 0.06102937698364258, 0.06078771209716797, 0.06082457733154297, 0.059865089416503904, 0.060216320037841796, 0.059087871551513675, 0.05941452789306641, 0.06077439880371094, 0.06028083038330078, 0.06068121719360352, 0.06069760131835938, 0.06096691131591797, 0.06012416076660156, 0.06075494384765625, 0.06013849639892578, 0.06086963272094727, 0.060609535217285154, 0.06098739242553711, 0.06088294219970703, 0.06077337646484375, 0.06087782287597656, 0.060805118560791016, 0.06075801467895508, 0.060829696655273435, 0.060837886810302735, 0.060200958251953124, 0.060911617279052734, 0.06101094436645508, 0.1231247329711914, 0.060744705200195315, 0.06079897689819336, 0.0600002555847168, 0.059115520477294924, 0.0603351058959961, 0.06076620864868164, 0.060727294921875, 0.060744705200195315, 0.060527614593505856, 0.06128844833374023, 0.06094438552856445, 0.060677120208740234, 0.06176768112182617, 0.062110721588134764, 0.0609249267578125, 0.060832767486572265, 0.060432384490966794, 0.06098636627197265, 0.06112972640991211, 0.060895233154296874, 0.062491649627685546, 0.060980224609375, 0.06099558258056641, 0.06110310363769531, 0.06090956878662109, 0.060767230987548826, 0.06076211166381836, 0.06077542495727539, 0.06094847869873047, 0.06072115325927734, 0.06075392150878906, 0.060911617279052734, 0.06179123306274414, 0.06146355056762695, 0.06151174545288086, 0.05984864044189453, 0.0592097282409668, 0.059241470336914064, 0.0594442253112793, 0.05915238571166992, 0.05912063980102539, 0.059140094757080076, 0.05916159820556641, 0.05927219009399414, 0.059033599853515625, 0.059902976989746094, 0.06051942443847656, 0.060739585876464844, 0.06076313781738281, 0.060609535217285154, 0.059551742553710936, 0.05947699356079102, 0.06035257720947266, 0.060338111877441404, 0.06081126403808594, 0.06072524642944336, 0.060757022857666015, 0.06074979019165039, 0.060472320556640625, 0.06065459060668945, 0.06086656188964844, 0.06003609466552735, 0.12312576293945313, 0.06074060821533203, 0.06058700942993164, 0.060955646514892575, 0.060897281646728516, 0.060668926239013675, 0.05904793548583984, 0.059138046264648435, 0.0591984977722168, 0.059074527740478514, 0.05915545654296875, 0.05922918319702149, 0.05921279907226563, 0.05931417465209961, 0.05925068664550781, 0.05924665451049805, 0.05927315139770508, 0.062034942626953124, 0.06089625549316406, 0.05932646560668945, 0.05910630416870117, 0.05935615921020508, 0.059224063873291016, 0.0590909423828125, 0.05909299087524414, 0.05916159820556641, 0.05918105697631836, 0.05901004791259765, 0.05914828872680664, 0.05935718536376953, 0.059216896057128904, 0.05914214324951172, 0.05914214324951172, 0.05903564834594727, 0.05917388916015625, 0.059701248168945314, 0.05926399993896484, 0.05921177673339844, 0.059167774200439456, 0.05920764923095703, 0.05906227111816406, 0.05908992004394531, 0.059063297271728515, 0.05934694290161133, 0.05926707077026367, 0.05919027328491211, 0.05903462219238281, 0.0592988166809082, 0.05978316879272461, 0.06010367965698242, 0.05926502227783203, 0.059202560424804686, 0.05930092620849609, 0.05915334320068359, 0.059085823059082034, 0.059358207702636716, 0.05939199829101562, 0.05905929565429688, 0.05903247833251953, 0.05921382522583008, 0.059084800720214846, 0.05944319915771484, 0.059240447998046876, 0.11974553680419922, 0.059086910247802736, 0.05899875259399414, 0.05906224060058594, 0.059066368103027345, 0.059218944549560545, 0.05904281616210937, 0.05910220718383789, 0.05922304153442383, 0.059017215728759766, 0.05900185775756836, 0.05915750503540039, 0.059101184844970706, 0.05916262435913086, 0.05922099304199219, 0.059033599853515625, 0.05897625732421875, 0.05926604843139648, 0.059210750579833986, 0.059121662139892575, 0.05920358276367187, 0.05924556732177735, 0.059066368103027345, 0.05899776077270508, 0.05931008148193359, 0.06076620864868164, 0.059312126159667966, 0.059284481048583984, 0.059046913146972656, 0.05910425567626953, 0.059017215728759766, 0.05894041442871094, 0.05906227111816406, 0.05915955352783203, 0.059063297271728515, 0.05913600158691406, 0.05902233505249024, 0.059099136352539064, 0.059030529022216796, 0.05913497543334961, 0.0589035530090332, 0.05907353591918945, 0.059066368103027345, 0.058962944030761716, 0.059486209869384764, 0.059453441619873044, 0.058998783111572264, 0.058987518310546876, 0.05903564834594727, 0.05900185775756836, 0.05933977508544922, 0.05908070373535156, 0.059009025573730466, 0.06041702270507812, 0.05994598388671875, 0.058993663787841794, 0.05916262435913086, 0.059085823059082034, 0.05953638458251953, 0.05912473678588867, 0.05950054550170898, 0.060255264282226564, 0.05989884948730469, 0.11980083465576172, 0.059218944549560545, 0.05915238571166992, 0.059079681396484375, 0.05905715179443359, 0.059115520477294924, 0.05936025619506836, 0.0590643196105957, 0.05912575912475586, 0.05918207931518555, 0.06097919845581055, 0.059486209869384764, 0.05931622314453125, 0.059044864654541014, 0.05903363037109375, 0.059218910217285155, 0.059025409698486325, 0.05917491149902344, 0.05916364669799805, 0.05909401702880859, 0.0590489616394043, 0.05904076766967774, 0.059096065521240235, 0.059121662139892575, 0.05902438354492188, 0.05902131271362305, 0.05898854446411133, 0.05910015869140625, 0.059300865173339844, 0.05925785446166992, 0.06097612762451172, 0.060662784576416016, 0.060878849029541014, 0.06074060821533203, 0.06073446273803711, 0.06080412673950195, 0.06049481582641601, 0.06058700942993164, 0.060581886291503906, 0.0608092155456543, 0.059243518829345705, 0.059063297271728515, 0.05907251358032226, 0.05892403030395508, 0.05899776077270508, 0.05919744110107422, 0.05904793548583984, 0.059481086730957033, 0.05923942565917969, 0.05914419174194336, 0.05912985610961914, 0.059145217895507814, 0.05928550338745117, 0.059261951446533206, 0.05939199829101562, 0.05910323333740235, 0.05878988647460937, 0.059832321166992185, 0.061055999755859375, 0.060690433502197265, 0.06087174224853516, 0.06086547088623047, 0.060862464904785155, 0.12235059356689452, 0.05892607879638672, 0.05911859130859375, 0.05902950286865234, 0.05892403030395508, 0.058929153442382816, 0.05973811340332031, 0.059853824615478515, 0.058992641448974606, 0.059038719177246096, 0.05915545654296875, 0.05913190460205078, 0.058912769317626956, 0.05937152099609375, 0.060902400970458986, 0.06070272064208984, 0.06099353790283203, 0.060695552825927736, 0.059014144897460936, 0.05908172988891602, 0.05928140640258789, 0.059154430389404294, 0.06019583892822265, 0.06073241424560547, 0.0606668815612793, 0.060614654541015625, 0.06086252975463867, 0.06079072189331055, 0.06075801467895508, 0.06062899017333984, 0.06069247817993164, 0.060826625823974606, 0.061156352996826174, 0.06073241424560547, 0.06075392150878906, 0.060655616760253904, 0.06082252883911133, 0.06071705627441406, 0.060818431854248046, 0.060819454193115234, 0.06071091079711914, 0.06072012710571289, 0.0607723503112793, 0.061059070587158204, 0.06127308654785156, 0.060447742462158206, 0.05913907241821289, 0.05907251358032226, 0.06083174514770508, 0.06084505462646484, 0.06077644729614258, 0.060709888458251954, 0.0603054084777832, 0.059133953094482425, 0.05898342514038086, 0.06038323211669922, 0.06078668975830078, 0.059170848846435545, 0.060978145599365235, 0.06088601684570313, 0.06031568145751953, 0.05907247924804687, 0.05908889770507812]",tokens/s,16.3764561555309,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-32B,Qwen/Qwen1.5-32B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,4421.566464,24111.480832,0.0,23465.033728,21691.057664,s,10,26.201972412109374,2.6201972412109376,0.0014392080916773324,2.620363525390625,2.621498388671875,2.622221728515625,2.622800400390625,"[2.62030126953125, 2.620416259765625, 2.620310791015625, 2.621337646484375, 2.620593017578125, 2.619945556640625, 2.622945068359375, 2.618053466796875, 2.6175732421875, 2.62049609375]",tokens/s,97.70256833095829,kWh,3.092599951558643e-05,1.6948578058490964e-05,0.00015162839908040572,0.00019950297665448312,tokens/kWh,1283188.874135765,MB,4421.566464,24111.480832,0.0,23465.033728,21890.213376,s,10,1554.3346406249998,155.43346406249998,0.013784362336340894,155.428859375,155.4548328125,155.45787734375,155.46031296875,"[155.44153125, 155.42890625, 155.45415625, 155.426875, 155.41909375, 155.416984375, 155.4288125, 155.42403125, 155.460921875, 155.433328125]",tokens/s,0.4053181236099044,kWh,0.0018348197203377882,0.0010056435987538315,0.008919082301926796,0.011759545621018413,tokens/kWh,5357.349852650515,,s,629,1575.449455322266,2.504689118159405,0.3111804707409938,2.46706884765625,2.4683868164062504,2.4691709960937502,5.086480859375,"[2.467560546875, 2.468384765625, 2.46702587890625, 2.467464111328125, 2.46763427734375, 2.468833251953125, 2.467610595703125, 2.46744580078125, 2.467310546875, 2.4678994140625, 2.46620263671875, 2.466724853515625, 2.46737109375, 2.46790234375, 2.466193359375, 2.46736279296875, 2.4661279296875, 2.467852294921875, 2.469380126953125, 2.468475830078125, 2.467101806640625, 2.46930322265625, 2.4674384765625, 2.46818505859375, 2.46793115234375, 2.468096923828125, 2.466873291015625, 2.466653076171875, 2.466144287109375, 2.467576904296875, 2.466427978515625, 2.46687744140625, 2.467862548828125, 2.468832275390625, 2.46677001953125, 2.46746533203125, 2.4666767578125, 2.467177490234375, 2.4664228515625, 2.466610107421875, 2.466058349609375, 2.46723388671875, 2.470784912109375, 2.466979736328125, 2.466400146484375, 2.4673740234375, 2.467078125, 2.466891845703125, 2.46717041015625, 2.467664794921875, 2.4676455078125, 2.467311767578125, 2.4667421875, 2.46812060546875, 2.467968017578125, 2.467040283203125, 2.466512939453125, 2.4665068359375, 2.467083251953125, 2.466418701171875, 2.465967041015625, 2.466157470703125, 5.087494140625, 2.466189208984375, 2.466884521484375, 2.46639501953125, 2.46648828125, 2.466607177734375, 2.467287109375, 2.46839599609375, 2.46702587890625, 2.467313720703125, 2.467323974609375, 2.467330078125, 2.466697265625, 2.46641455078125, 2.4657080078125, 2.46809912109375, 2.46658984375, 2.467029052734375, 2.46753076171875, 2.468769775390625, 2.467987548828125, 2.468222900390625, 2.4690595703125, 2.469578857421875, 2.46847900390625, 2.4680283203125, 2.46656298828125, 2.46801513671875, 2.46639013671875, 2.4670986328125, 2.46615234375, 2.46773974609375, 2.4667216796875, 2.46727685546875, 2.4663193359375, 2.46763525390625, 2.46688671875, 2.467874755859375, 2.46723779296875, 2.46824853515625, 2.466218017578125, 2.466986083984375, 2.46668505859375, 2.4669931640625, 2.4665302734375, 2.465965087890625, 2.468173828125, 2.467091552734375, 2.4666787109375, 2.466904052734375, 2.466720703125, 2.466922607421875, 2.46651806640625, 2.466711669921875, 2.46662158203125, 2.4666767578125, 2.466606201171875, 2.466504638671875, 2.46702294921875, 2.467114990234375, 2.46698291015625, 2.466469970703125, 2.46719384765625, 5.0873857421875, 2.46711181640625, 2.4676669921875, 2.4682373046875, 2.46677294921875, 2.465900634765625, 2.466285400390625, 2.46675244140625, 2.4686142578125, 2.467852294921875, 2.468147216796875, 2.469138427734375, 2.468021240234375, 2.46839501953125, 2.467322998046875, 2.46803857421875, 2.4694794921875, 2.467284912109375, 2.467267578125, 2.46729736328125, 2.471918701171875, 2.4674765625, 2.467620849609375, 2.468518798828125, 2.46965966796875, 2.468581298828125, 2.467287109375, 2.4676025390625, 2.46909130859375, 2.4678798828125, 2.466585693359375, 2.46625390625, 2.468111328125, 2.466431884765625, 2.4665302734375, 2.46625390625, 2.46766796875, 2.46662646484375, 2.46632763671875, 2.466314208984375, 2.468085693359375, 2.467567626953125, 2.47077490234375, 2.466901123046875, 2.46776416015625, 2.466997314453125, 2.468117431640625, 2.471785400390625, 2.468166748046875, 2.46658154296875, 2.467800048828125, 2.4673310546875, 2.46755029296875, 2.466908203125, 2.4666357421875, 2.4657724609375, 2.4668466796875, 2.46647509765625, 2.46641455078125, 2.4653935546875, 2.46607861328125, 2.465657958984375, 2.46717041015625, 5.08862353515625, 2.46618017578125, 2.46677001953125, 2.465965087890625, 2.4666962890625, 2.46651806640625, 2.466512939453125, 2.4657919921875, 2.46704833984375, 2.466091064453125, 2.4665322265625, 2.466198486328125, 2.466378662109375, 2.4663388671875, 2.466697265625, 2.4661689453125, 2.467324951171875, 2.46691748046875, 2.46706884765625, 2.46692041015625, 2.466785400390625, 2.467284912109375, 2.467230712890625, 2.4668681640625, 2.4700732421875, 2.46681396484375, 2.466840576171875, 2.46628857421875, 2.4691845703125, 2.469140380859375, 2.468798583984375, 2.46765869140625, 2.4666748046875, 2.46800390625, 2.466842529296875, 2.46696044921875, 2.46841552734375, 2.46970166015625, 2.468170654296875, 2.467095458984375, 2.46616259765625, 2.468423583984375, 2.46754296875, 2.467493896484375, 2.469568603515625, 2.467686279296875, 2.467119140625, 2.46765576171875, 2.4675615234375, 2.46753076171875, 2.46786962890625, 2.467504150390625, 2.466217041015625, 2.46670751953125, 2.466275390625, 2.46649951171875, 2.465642578125, 2.4665927734375, 2.466114501953125, 2.46563134765625, 2.466107421875, 2.466840576171875, 2.46594775390625, 5.0872607421875, 2.466586669921875, 2.467085205078125, 2.466087890625, 2.46624658203125, 2.468030517578125, 2.466228271484375, 2.46631201171875, 2.465871826171875, 2.466620361328125, 2.466470947265625, 2.4659375, 2.465469482421875, 2.467073974609375, 2.467386474609375, 2.466375732421875, 2.46649658203125, 2.469547119140625, 2.46750634765625, 2.466642822265625, 2.46601123046875, 2.466673583984375, 2.466423828125, 2.466227294921875, 2.466232421875, 2.467124267578125, 2.466873291015625, 2.46641162109375, 2.466154541015625, 2.46636962890625, 2.466863037109375, 2.46698291015625, 2.467325927734375, 2.46662744140625, 2.4677744140625, 2.466999267578125, 2.467263427734375, 2.466723876953125, 2.46856201171875, 2.46679150390625, 2.466046875, 2.46635107421875, 2.46673828125, 2.467541015625, 2.4671416015625, 2.467443603515625, 2.466511962890625, 2.467031982421875, 2.46706787109375, 2.4669921875, 2.468369384765625, 2.467056640625, 2.46666455078125, 2.467033203125, 2.468075439453125, 2.468128662109375, 2.466417724609375, 2.471729248046875, 2.467071044921875, 2.466891845703125, 2.46647509765625, 2.466440185546875, 2.46868798828125, 5.08691357421875, 2.466028564453125, 2.4677333984375, 2.467010498046875, 2.467205078125, 2.4664013671875, 2.4672451171875, 2.4667392578125, 2.46585546875, 2.466406494140625, 2.46685693359375, 2.4661708984375, 2.4663828125, 2.46592822265625, 2.466884521484375, 2.466239501953125, 2.46583203125, 2.465574951171875, 2.46738330078125, 2.467092529296875, 2.4664677734375, 2.467158935546875, 2.467737548828125, 2.4660244140625, 2.4675830078125, 2.46628662109375, 2.467053466796875, 2.4666142578125, 2.46684375, 2.466788330078125, 2.467851318359375, 2.46658056640625, 2.468338623046875, 2.467275634765625, 2.46681201171875, 2.466661376953125, 2.466303955078125, 2.46763720703125, 2.473092041015625, 2.46727685546875, 2.467389404296875, 2.467167236328125, 2.467786865234375, 2.4663837890625, 2.4663388671875, 2.465721435546875, 2.467124267578125, 2.4679638671875, 2.46670751953125, 2.46626416015625, 2.467136474609375, 2.466255859375, 2.4674990234375, 2.466595947265625, 2.466490478515625, 2.4665712890625, 2.46690087890625, 2.466754638671875, 2.4670341796875, 2.467210205078125, 2.467547119140625, 2.467244140625, 2.466747314453125, 5.09081201171875, 2.467222412109375, 2.46835205078125, 2.467926025390625, 2.4673740234375, 2.467116943359375, 2.46747021484375, 2.467093505859375, 2.466975830078125, 2.46742626953125, 2.4675400390625, 2.467567626953125, 2.4670546875, 2.467200927734375, 2.467986328125, 2.46847900390625, 2.468297607421875, 2.466345947265625, 2.467420166015625, 2.468263916015625, 2.46740283203125, 2.466755615234375, 2.467157958984375, 2.471689208984375, 2.467622802734375, 2.4668671875, 2.467239013671875, 2.46835205078125, 2.467516357421875, 2.46651904296875, 2.466069580078125, 2.46681591796875, 2.466124755859375, 2.465919921875, 2.465881103515625, 2.46696240234375, 2.46622607421875, 2.46639208984375, 2.465967041015625, 2.46714990234375, 2.466946044921875, 2.466486328125, 2.465672119140625, 2.466734130859375, 2.466515869140625, 2.46664404296875, 2.4666142578125, 2.4672265625, 2.466231201171875, 2.46652001953125, 2.4661728515625, 2.46692041015625, 2.466157470703125, 2.46626611328125, 2.46582470703125, 2.466809814453125, 2.467745849609375, 2.4677080078125, 2.466193359375, 2.4672470703125, 2.467043212890625, 2.47162060546875, 2.4657490234375, 5.08459130859375, 2.46683544921875, 2.46649755859375, 2.467766357421875, 2.467400634765625, 2.467850341796875, 2.467080078125, 2.46782470703125, 2.466931640625, 2.46681298828125, 2.466908203125, 2.467493896484375, 2.466670654296875, 2.46761767578125, 2.467306396484375, 2.46744873046875, 2.46968310546875, 2.467335205078125, 2.466291748046875, 2.46672900390625, 2.4669912109375, 2.468328369140625, 2.466957275390625, 2.46684375, 2.467378173828125, 2.4669912109375, 2.467179443359375, 2.46784521484375, 2.468820068359375, 2.466711669921875, 2.465966064453125, 2.465700927734375, 2.4670791015625, 2.46738427734375, 2.467485595703125, 2.4666328125, 2.4676455078125, 2.466926513671875, 2.46613720703125, 2.46624755859375, 2.4671220703125, 2.46669921875, 2.465594482421875, 2.46590771484375, 2.466947021484375, 2.466668701171875, 2.466886474609375, 2.46719189453125, 2.46757080078125, 2.466482177734375, 2.46595068359375, 2.46605712890625, 2.467263427734375, 2.466189208984375, 2.467197998046875, 2.46523095703125, 2.466417724609375, 2.466122802734375, 2.467242919921875, 2.466663330078125, 2.46723388671875, 2.466820068359375, 2.472427490234375, 5.0853681640625, 2.46607568359375, 2.46763818359375, 2.466788330078125, 2.467725341796875, 2.466460693359375, 2.467099609375, 2.46757177734375, 2.46760546875, 2.466769775390625, 2.46808056640625, 2.46717236328125, 2.468547607421875, 2.46778369140625, 2.4672236328125, 2.466810791015625, 2.467812255859375, 2.468862060546875, 2.467946533203125, 2.467600341796875, 2.46891015625, 2.46693994140625, 2.46805810546875, 2.46829052734375, 2.469150634765625, 2.468052978515625, 2.4687646484375, 2.467143798828125, 2.467037109375, 2.466817138671875, 2.468148193359375, 2.467322998046875, 2.467407958984375, 2.468336669921875, 2.468115478515625, 2.4671845703125, 2.4663388671875, 2.4676884765625, 2.467588134765625, 2.466388916015625, 2.466503662109375, 2.466964599609375, 2.4676494140625, 2.466975830078125, 2.466547607421875, 2.46734033203125, 2.467566650390625, 2.467642333984375, 2.466969482421875, 2.468360107421875, 2.467146728515625, 2.466821044921875, 2.466783203125, 2.46778466796875, 2.46734130859375, 2.467453857421875, 2.467220458984375, 2.468514892578125, 2.467493896484375, 2.4675439453125, 2.467715087890625, 2.47457275390625, 2.468958251953125, 5.0886162109375, 2.467864501953125, 2.469477294921875, 2.468490234375, 2.4670166015625, 2.467751953125, 2.46676171875, 2.466572265625, 2.467084228515625, 2.467091552734375, 2.4673525390625, 2.4678388671875, 2.46633154296875, 2.467618896484375, 2.46672900390625, 2.466831298828125, 2.4672685546875, 2.467834716796875, 2.46698193359375, 2.467683349609375, 2.46766796875, 2.46748974609375, 2.46799560546875, 2.4667822265625, 2.4664052734375, 2.467306396484375, 2.469897216796875, 2.46721630859375, 2.46658251953125, 2.46739453125, 2.467197998046875, 2.468125732421875, 2.466786376953125, 2.467189697265625, 2.467324951171875, 2.467589111328125, 2.46772314453125, 2.468556884765625, 2.466018310546875, 2.466438232421875, 2.467070068359375, 2.466093017578125, 2.466754638671875, 2.46706884765625, 2.46609521484375, 2.467577880859375, 2.46765771484375, 2.468675537109375, 2.46646875, 2.466482177734375, 2.466712646484375, 2.465977294921875, 2.46544384765625, 2.466545654296875, 2.466736083984375, 2.467407958984375, 2.466906005859375, 2.466165771484375, 2.46822900390625, 2.4675185546875, 2.466716552734375, 2.4658984375, 2.466937744140625]",tokens/s,0.3992511456810495,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-4.5B,facebook/xglm-4.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,3047.755776,4521.984,0.0,3875.536896,3575.121408,s,10,3.925874938964844,0.3925874938964844,0.0032003814064900295,0.39131011962890627,0.3974839904785156,0.39779729309082035,0.39804793518066406,"[0.398110595703125, 0.3898521728515625, 0.39046603393554685, 0.39330859375, 0.38967181396484374, 0.38878271484375, 0.39181430053710936, 0.3956484069824219, 0.39741436767578125, 0.3908059387207031]",tokens/s,652.0839404718809,kWh,4.606836531152074e-06,2.5243247530642504e-06,1.9845015875998338e-05,2.697617716021466e-05,tokens/kWh,9489854.640247436,MB,3047.755776,4521.984,0.0,3875.536896,3800.975872,s,10,231.19119335937498,23.119119335937498,0.024652283207228468,23.1104677734375,23.1560455078125,23.16491630859375,23.17201294921875,"[23.104484375, 23.106837890625, 23.11409765625, 23.1257734375, 23.09865234375, 23.091490234375, 23.10330859375, 23.173787109375, 23.15407421875, 23.1186875]",tokens/s,2.7250172934602097,kWh,0.0002730506428477602,0.00014965489141425045,0.0011581440098478151,0.0015808495441098258,tokens/kWh,39851.989858702975,,s,629,234.351453125,0.3725778269077901,0.04661763996043978,0.3668070373535156,0.3683815368652344,0.36919970703125,0.7562686254882812,"[0.3660308532714844, 0.3664025573730469, 0.36736306762695314, 0.3682928771972656, 0.3671849060058594, 0.366424072265625, 0.36691455078125, 0.36642098999023437, 0.3666820983886719, 0.36670053100585936, 0.3671910400390625, 0.3671593017578125, 0.3675473937988281, 0.367383544921875, 0.36641281127929687, 0.3659335632324219, 0.366244873046875, 0.3652311096191406, 0.3665745849609375, 0.3662633056640625, 0.3673968505859375, 0.36597760009765623, 0.3667660827636719, 0.3654379577636719, 0.3664527282714844, 0.3673671569824219, 0.36636468505859376, 0.3656898498535156, 0.3663196105957031, 0.3655997314453125, 0.36726272583007813, 0.36568780517578126, 0.3668695068359375, 0.36599508666992187, 0.3665059204101562, 0.3655762023925781, 0.36666366577148435, 0.36643328857421875, 0.3669176330566406, 0.3657113647460937, 0.36776141357421877, 0.3660533752441406, 0.36670053100585936, 0.36581787109375, 0.36618853759765624, 0.3657318420410156, 0.36655206298828125, 0.3656663818359375, 0.36818527221679687, 0.36655206298828125, 0.3670487060546875, 0.366587890625, 0.3682109375, 0.3672012939453125, 0.3683788757324219, 0.3662612609863281, 0.36835430908203126, 0.3673917541503906, 0.3714969482421875, 0.36665652465820314, 0.3664425048828125, 0.36596124267578123, 0.7560806274414062, 0.36733746337890627, 0.36763238525390624, 0.3667384338378906, 0.3690987548828125, 0.3665008544921875, 0.3659049072265625, 0.3675392150878906, 0.36642611694335936, 0.36708966064453125, 0.36646810913085937, 0.3664025573730469, 0.3674142761230469, 0.36601959228515624, 0.36743270874023437, 0.36704461669921873, 0.36616806030273436, 0.36657357788085937, 0.3658526611328125, 0.3659263916015625, 0.3657093200683594, 0.3664773254394531, 0.3665899658203125, 0.3660308532714844, 0.36575845336914065, 0.3663247375488281, 0.3659059143066406, 0.36675994873046874, 0.3660185546875, 0.3673917541503906, 0.36654180908203127, 0.36684698486328127, 0.367072265625, 0.3671255187988281, 0.36628070068359375, 0.3660789794921875, 0.36613528442382814, 0.3671644287109375, 0.3661455383300781, 0.36799591064453124, 0.36629299926757813, 0.36816485595703125, 0.36593869018554687, 0.3678883972167969, 0.3660892028808594, 0.3684013977050781, 0.36708865356445314, 0.36752896118164063, 0.36656741333007814, 0.3682099304199219, 0.36648345947265626, 0.36751565551757814, 0.3659735107421875, 0.3680624694824219, 0.3659202575683594, 0.3667906494140625, 0.36607489013671873, 0.3667957763671875, 0.36641485595703127, 0.36750439453125, 0.36619058227539064, 0.36707431030273435, 0.36645068359375, 0.7576248168945312, 0.3653406677246094, 0.3671552124023437, 0.3671224365234375, 0.3667906494140625, 0.3663052673339844, 0.36621209716796876, 0.36776141357421877, 0.3668439025878906, 0.36665139770507815, 0.3658373107910156, 0.36659506225585936, 0.36775115966796873, 0.36711117553710937, 0.36632986450195315, 0.36725247192382815, 0.36715213012695314, 0.36897998046875, 0.3676968994140625, 0.3678535766601562, 0.366065673828125, 0.367072265625, 0.36664523315429687, 0.3668899841308594, 0.366271484375, 0.3664445495605469, 0.3664271240234375, 0.366392333984375, 0.36568576049804685, 0.36705279541015623, 0.36768154907226563, 0.36760678100585936, 0.36624383544921874, 0.3671849060058594, 0.36741018676757814, 0.36703640747070315, 0.36581375122070314, 0.36656536865234374, 0.3663052673339844, 0.36876492309570313, 0.366455810546875, 0.3671715698242187, 0.36698419189453124, 0.3677542419433594, 0.36644046020507814, 0.36671795654296874, 0.36616293334960937, 0.36698828125, 0.3662899169921875, 0.36686746215820315, 0.3674449768066406, 0.36675994873046874, 0.366519287109375, 0.3665111083984375, 0.36600628662109375, 0.36778289794921876, 0.3663800354003906, 0.36820684814453125, 0.3660175476074219, 0.36745419311523436, 0.3680143432617187, 0.3678023681640625, 0.3672965087890625, 0.76101123046875, 0.36763442993164064, 0.3688806457519531, 0.3682867126464844, 0.3672862854003906, 0.36792831420898436, 0.36681729125976564, 0.36853555297851565, 0.36648446655273437, 0.36719000244140626, 0.3671961669921875, 0.3679938659667969, 0.3675965576171875, 0.3683921813964844, 0.3668418579101563, 0.3667855224609375, 0.36585470581054685, 0.36641998291015626, 0.3668428649902344, 0.3661527099609375, 0.3681187744140625, 0.3663144836425781, 0.367393798828125, 0.36705484008789063, 0.36674969482421876, 0.36767642211914064, 0.3666595764160156, 0.36638516235351565, 0.3659837341308594, 0.3676375122070312, 0.36701901245117186, 0.36602264404296875, 0.36724429321289065, 0.36727194213867187, 0.3677163391113281, 0.36646194458007814, 0.3668899841308594, 0.36752896118164063, 0.3668070373535156, 0.36688385009765623, 0.3675699157714844, 0.36781158447265627, 0.3655679931640625, 0.3665203247070313, 0.3669176330566406, 0.3664250793457031, 0.36553421020507815, 0.3678494567871094, 0.367056884765625, 0.3667271728515625, 0.3659151306152344, 0.36611276245117186, 0.3676180419921875, 0.3684751281738281, 0.3661475830078125, 0.3675002746582031, 0.3667466125488281, 0.36979815673828126, 0.3675893859863281, 0.3668500366210937, 0.36594073486328127, 0.3664721984863281, 0.3660421142578125, 0.7563417358398438, 0.36537957763671874, 0.3666237487792969, 0.3666851806640625, 0.36812799072265623, 0.3660308532714844, 0.3663882141113281, 0.3670702209472656, 0.3671562194824219, 0.3671490478515625, 0.3683932189941406, 0.36664523315429687, 0.3658895263671875, 0.36612710571289064, 0.36652340698242186, 0.3672749938964844, 0.3655301208496094, 0.366635009765625, 0.3657994384765625, 0.36625100708007813, 0.36786483764648437, 0.3671715698242187, 0.36642816162109376, 0.3672166442871094, 0.3667108459472656, 0.36772549438476565, 0.36702822875976565, 0.367720458984375, 0.3670384521484375, 0.3671142272949219, 0.3670169677734375, 0.369112060546875, 0.36612506103515624, 0.3668643798828125, 0.3657052307128906, 0.3660328979492187, 0.3655086059570313, 0.3667589111328125, 0.3669944458007813, 0.36692889404296875, 0.3655577697753906, 0.3665080261230469, 0.36726373291015624, 0.36662066650390623, 0.3658455505371094, 0.36661654663085935, 0.365918212890625, 0.368047119140625, 0.3663114318847656, 0.36607794189453124, 0.36666778564453123, 0.36623870849609375, 0.36541131591796877, 0.36611993408203125, 0.36540826416015626, 0.36666470336914064, 0.3660861511230469, 0.3677306823730469, 0.3666360168457031, 0.36710296630859374, 0.366376953125, 0.3661721496582031, 0.36556698608398436, 0.75503515625, 0.36811160278320315, 0.3666186218261719, 0.3668746337890625, 0.36598681640625, 0.36689202880859373, 0.3665131530761719, 0.36666983032226563, 0.3660205993652344, 0.3664025573730469, 0.3666462707519531, 0.3671009216308594, 0.3655587768554687, 0.3668203430175781, 0.36630117797851564, 0.36666366577148435, 0.36585574340820315, 0.36627865600585935, 0.3661414489746094, 0.36637799072265625, 0.36607998657226565, 0.3658045349121094, 0.3679395751953125, 0.36621517944335935, 0.36576153564453123, 0.3663124389648437, 0.3655833740234375, 0.3661219787597656, 0.3678760986328125, 0.36634521484375, 0.36614862060546877, 0.3665377197265625, 0.3658229675292969, 0.36720025634765624, 0.3659571228027344, 0.36605746459960936, 0.36572467041015627, 0.36643429565429686, 0.3677470703125, 0.3671142272949219, 0.36663092041015627, 0.36711935424804687, 0.3663636474609375, 0.36681625366210935, 0.3660943298339844, 0.36676812744140624, 0.36628582763671874, 0.3674306640625, 0.3657963562011719, 0.36687359619140625, 0.36616293334960937, 0.3663595581054688, 0.3658577880859375, 0.367494140625, 0.3659980773925781, 0.3674224548339844, 0.3666606140136719, 0.3666483154296875, 0.36678964233398437, 0.367278076171875, 0.3672596435546875, 0.36691558837890625, 0.3659898986816406, 0.7581112060546875, 0.36696063232421877, 0.36686541748046875, 0.3666217041015625, 0.3667189636230469, 0.36658688354492186, 0.36612606811523435, 0.3663329162597656, 0.36610763549804687, 0.3663811340332031, 0.3662243347167969, 0.3669329833984375, 0.3673456726074219, 0.3678023681640625, 0.36764877319335937, 0.3677501525878906, 0.3669678039550781, 0.3668746337890625, 0.3683153991699219, 0.36921240234375, 0.368362548828125, 0.36775827026367186, 0.3663052673339844, 0.3672483825683594, 0.36686746215820315, 0.36620184326171873, 0.36589157104492187, 0.36683877563476563, 0.36737841796875, 0.36710400390625, 0.3661424560546875, 0.36658074951171876, 0.365517822265625, 0.36632781982421875, 0.36605645751953125, 0.366856201171875, 0.36679168701171877, 0.3667793884277344, 0.3662469177246094, 0.36760064697265626, 0.36884786987304685, 0.3667568664550781, 0.3658803100585937, 0.36786996459960936, 0.3664742431640625, 0.3666083984375, 0.36686541748046875, 0.367140869140625, 0.3666790466308594, 0.36664422607421876, 0.365802490234375, 0.36675277709960935, 0.365348876953125, 0.36615576171875, 0.3656755065917969, 0.36651724243164063, 0.3666483154296875, 0.3665623168945312, 0.3659253845214844, 0.36593869018554687, 0.3652372436523437, 0.36640359497070313, 0.3656540222167969, 0.7624765625, 0.36612710571289064, 0.3665633239746094, 0.36761907958984374, 0.36593048095703123, 0.36634112548828124, 0.3662571411132812, 0.3672842102050781, 0.3657646179199219, 0.368606201171875, 0.36768359375, 0.3676794738769531, 0.36858981323242185, 0.3683133544921875, 0.369396728515625, 0.36714599609375, 0.36760577392578125, 0.3668357238769531, 0.3677470703125, 0.3677235107421875, 0.36642098999023437, 0.36621209716796876, 0.3658362731933594, 0.36801739501953123, 0.3667783813476562, 0.3668715515136719, 0.3696394348144531, 0.369375244140625, 0.3691806640625, 0.3698810729980469, 0.37041253662109375, 0.36909466552734377, 0.3689021301269531, 0.3667169189453125, 0.36602264404296875, 0.3657953186035156, 0.3676252136230469, 0.36923904418945314, 0.3685191650390625, 0.3708651428222656, 0.3690188903808594, 0.370050048828125, 0.36955239868164064, 0.366551025390625, 0.3658076171875, 0.3684198303222656, 0.3685857238769531, 0.3687065734863281, 0.36557516479492186, 0.3663943786621094, 0.3689195556640625, 0.3698083801269531, 0.3692380065917969, 0.3681546325683594, 0.3669350280761719, 0.36759039306640623, 0.3672842102050781, 0.3702108154296875, 0.3688243103027344, 0.36716030883789064, 0.36670465087890625, 0.36789248657226564, 0.3689768981933594, 0.7654307861328125, 0.367162353515625, 0.369691650390625, 0.3691734924316406, 0.370017333984375, 0.368702392578125, 0.36607794189453124, 0.36709375, 0.3667589111328125, 0.36747161865234373, 0.3679764404296875, 0.3669626770019531, 0.3658486022949219, 0.36790267944335936, 0.3670026245117187, 0.36758526611328124, 0.3668479919433594, 0.36762625122070314, 0.36738970947265626, 0.36922470092773435, 0.36768768310546873, 0.36928103637695314, 0.3660902404785156, 0.36697601318359374, 0.36616705322265625, 0.3675197448730469, 0.36707635498046876, 0.3673385009765625, 0.3668746337890625, 0.36747161865234373, 0.36656536865234374, 0.36796722412109373, 0.36610662841796876, 0.3665213317871094, 0.365781005859375, 0.36651007080078124, 0.36767025756835936, 0.3668213806152344, 0.36538470458984373, 0.3673118591308594, 0.365907958984375, 0.3671910400390625, 0.3668971557617188, 0.3673313293457031, 0.3669053344726563, 0.3664783935546875, 0.36691961669921874, 0.36783718872070315, 0.36801126098632814, 0.36738970947265626, 0.36602264404296875, 0.36634521484375, 0.3659909057617188, 0.36675787353515626, 0.3704688720703125, 0.37160037231445314, 0.370060302734375, 0.3701022644042969, 0.36780850219726563, 0.36857342529296877, 0.3668899841308594, 0.36739788818359376, 0.3669186706542969, 0.7624959716796875, 0.366529541015625, 0.367541259765625, 0.3666298828125, 0.367783935546875, 0.36849972534179687, 0.36857550048828125, 0.36766720581054685, 0.3665428466796875, 0.367025146484375, 0.3665684509277344, 0.3679231872558594, 0.3658014831542969, 0.3674449768066406, 0.36611276245117186, 0.367678466796875, 0.3667189636230469, 0.3687383117675781, 0.36729037475585935, 0.3665848388671875, 0.36641177368164063, 0.36767333984375, 0.3674972229003906, 0.36638516235351565, 0.3664025573730469, 0.36736306762695314, 0.3676334228515625, 0.3689072570800781, 0.3667712097167969, 0.3672862854003906, 0.3658874816894531, 0.36724429321289065, 0.3662274475097656, 0.3670732727050781, 0.3663523864746094, 0.3670425720214844, 0.3662489624023437, 0.36671282958984375, 0.3670978698730469, 0.3670241394042969, 0.3659970703125, 0.3673231506347656, 0.3663739013671875, 0.3662264404296875, 0.36703436279296875, 0.36681729125976564, 0.3661414489746094, 0.3668623352050781, 0.366129150390625, 0.367130615234375, 0.36584756469726565, 0.3665489807128906, 0.3668070373535156, 0.3673395080566406, 0.3674347534179688, 0.3669698486328125, 0.36617010498046876, 0.3672535095214844, 0.36757708740234374, 0.3671152648925781, 0.3667189636230469, 0.3670374450683594, 0.3659479064941406]",tokens/s,2.684002986166677,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,TencentARC/Mistral_Pro_8B_v0.1,TencentARC/Mistral_Pro_8B_v0.1,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,2593.296384,7298.613248,0.0,6652.166144,6323.352576,s,10,7.731728576660157,0.7731728576660156,0.0036666872570226623,0.7712686767578125,0.7772910278320313,0.7798124450683593,0.7818295788574219,"[0.7823338623046875, 0.776730712890625, 0.7708938598632813, 0.7714732666015625, 0.7710640869140625, 0.770478515625, 0.770045166015625, 0.773399658203125, 0.7746173706054688, 0.7706920776367188]",tokens/s,331.10319052428935,kWh,9.090474362556752e-06,4.981150633102515e-06,4.3684586229685554e-05,5.775621122534482e-05,tokens/kWh,4432423.709394238,MB,2593.296384,7298.613248,0.0,6652.166144,6382.565888,s,10,458.08653125,45.808653125,0.015714942284677954,45.80371484375,45.83385546875,45.83744140625,45.84031015625,"[45.8103125, 45.8045234375, 45.84102734375, 45.79598828125, 45.78718359375, 45.80095703125, 45.79937890625, 45.80290625, 45.8111953125, 45.83305859375]",tokens/s,1.3752860148079282,kWh,0.0005408116745630392,0.00029641227384910964,0.0025237221621385317,0.0033609461105506802,tokens/kWh,18744.721851454397,,s,629,464.3046874389652,0.7381632550698964,0.09178720852369766,0.7269785766601562,0.7282177978515625,0.72855,1.49778814453125,"[0.728890380859375, 0.7286569213867188, 0.728468505859375, 0.727841796875, 0.7261204223632812, 0.7262678833007813, 0.7260436401367187, 0.7263180541992188, 0.7261552734375, 0.7261777954101563, 0.727593994140625, 0.7266488037109375, 0.726371337890625, 0.7267962646484375, 0.7268167724609375, 0.72713623046875, 0.7261091918945313, 0.7263775024414062, 0.726086669921875, 0.7265126953125, 0.7268792114257813, 0.7265433349609375, 0.7264358520507812, 0.7269488525390625, 0.7277793579101562, 0.7264532470703124, 0.7258746948242187, 0.7262843017578124, 0.726619140625, 0.726530029296875, 0.726930419921875, 0.7266365356445312, 0.7263467407226563, 0.7267451171875, 0.7262003173828125, 0.7265435180664063, 0.7259226684570312, 0.7268444213867188, 0.727319580078125, 0.72618798828125, 0.725907470703125, 0.7261306762695312, 0.7267153930664062, 0.7260835571289063, 0.7304304809570312, 0.7278233642578125, 0.7278356323242188, 0.7286978759765625, 0.7283251342773438, 0.7282206420898437, 0.7269744873046875, 0.7285032958984375, 0.728458251953125, 0.7281141967773438, 0.7277547607421875, 0.727568359375, 0.7287777099609375, 0.727846923828125, 0.7280547485351563, 0.72765234375, 0.7284254760742187, 0.7282565307617187, 1.504901123046875, 0.7284664306640625, 0.7290664672851562, 0.7276646118164063, 0.7266221923828124, 0.7263283081054688, 0.7270154418945313, 0.7276861572265625, 0.7269110717773437, 0.7259544677734375, 0.7266682739257813, 0.7265515747070312, 0.7271710815429687, 0.7266058349609374, 0.7279667358398437, 0.7284111328125, 0.7286005859375, 0.726703125, 0.7264102172851562, 0.7261973876953125, 0.7271340942382812, 0.72724072265625, 0.7264429931640625, 0.726245361328125, 0.7261767578125, 0.7261071166992188, 0.72650244140625, 0.726097900390625, 0.7263776245117187, 0.726574951171875, 0.728237060546875, 0.7274116821289063, 0.7262125854492187, 0.7267153930664062, 0.7282913208007813, 0.726614013671875, 0.7263006591796874, 0.7260282592773437, 0.726560791015625, 0.7266262817382813, 0.7263662109375, 0.7263231811523437, 0.7260938110351562, 0.727815185546875, 0.7281008911132812, 0.727647216796875, 0.7297269897460937, 0.726920166015625, 0.72753564453125, 0.7265023803710937, 0.726255615234375, 0.7262545776367187, 0.7261071166992188, 0.72612353515625, 0.726687744140625, 0.726403076171875, 0.7265781860351562, 0.7276553955078126, 0.728158203125, 0.7277864990234375, 0.7276748657226563, 0.7281663818359375, 0.7285872802734376, 1.497443359375, 0.726253662109375, 0.7272754516601563, 0.7280455932617188, 0.7274752197265625, 0.7277711181640625, 0.728121337890625, 0.7287439575195312, 0.7278479614257812, 0.7272069091796876, 0.728195068359375, 0.7279093627929687, 0.7278653564453125, 0.7270317993164063, 0.7276226806640625, 0.7280025634765624, 0.7273707275390625, 0.7268731079101562, 0.7277240600585938, 0.7278991088867187, 0.728922119140625, 0.7292661743164063, 0.7271915283203125, 0.72674609375, 0.72665087890625, 0.7293204345703125, 0.7278919677734375, 0.7269775390625, 0.7271495971679688, 0.7265064697265625, 0.7276011352539062, 0.7278345947265625, 0.7281715087890624, 0.7285504150390625, 0.72888525390625, 0.7278458862304688, 0.7282175903320313, 0.7279083251953125, 0.7274700927734375, 0.72768408203125, 0.7280230102539063, 0.727125, 0.7273963623046875, 0.7267492065429687, 0.7279011840820313, 0.7279144897460937, 0.7274905395507812, 0.7290408935546875, 0.7272591552734375, 0.7271188354492187, 0.7262003173828125, 0.7260712890625, 0.7269284057617188, 0.7274291381835938, 0.7278981323242187, 0.7281622924804687, 0.727773193359375, 0.72789501953125, 0.7277240600585938, 0.727568359375, 0.727462890625, 0.72740966796875, 0.7279739379882812, 1.49783544921875, 0.7277803344726562, 0.7284193115234375, 0.7274598388671875, 0.727204833984375, 0.7272642822265625, 0.7271331787109375, 0.7269990234375, 0.72734619140625, 0.726697998046875, 0.7275581665039063, 0.7281787109375, 0.7283179321289063, 0.7270953369140625, 0.7263682250976562, 0.7266856689453125, 0.7272109985351562, 0.7262371826171875, 0.7263467407226563, 0.726561767578125, 0.7266099243164063, 0.7263775024414062, 0.7269427490234375, 0.7270051879882813, 0.7267901611328125, 0.7268003540039063, 0.7271588134765625, 0.7267839965820313, 0.7270287475585937, 0.7264204711914063, 0.7269273681640624, 0.7266437377929688, 0.7264727172851563, 0.7263098754882813, 0.7269898071289063, 0.72724072265625, 0.7265023803710937, 0.726582275390625, 0.727125, 0.7272601318359375, 0.7271116943359375, 0.7265310668945313, 0.7264296875, 0.7270942993164062, 0.72810498046875, 0.7268444213867188, 0.7269007568359375, 0.72627197265625, 0.7266754760742188, 0.7263273315429688, 0.726993896484375, 0.72631298828125, 0.727208984375, 0.7275560913085938, 0.7275110473632812, 0.7266375732421875, 0.7264818725585938, 0.7270154418945313, 0.726582275390625, 0.7271209106445312, 0.7262269287109375, 0.7262637939453125, 0.7265955810546875, 1.49766650390625, 0.7265485229492188, 0.7276533203125, 0.7271311645507812, 0.726666259765625, 0.7262494506835937, 0.7270901489257813, 0.7279677734375, 0.7269109497070313, 0.7260007934570313, 0.726369140625, 0.7265126342773438, 0.7262740478515625, 0.7264921875, 0.7263375244140625, 0.7265730590820313, 0.727419921875, 0.7267584228515624, 0.726582275390625, 0.7271137084960938, 0.7265505981445313, 0.726699951171875, 0.72631298828125, 0.7264921875, 0.7262166748046875, 0.7264942016601562, 0.727056396484375, 0.7262740478515625, 0.726329345703125, 0.72700927734375, 0.7266652221679688, 0.7271157836914063, 0.7266324462890625, 0.7262699584960938, 0.7262648315429687, 0.7263733520507812, 0.7264257202148438, 0.7263466796875, 0.7261910400390625, 0.726476806640625, 0.7270922241210938, 0.7264163818359375, 0.726287353515625, 0.726835205078125, 0.7277772827148438, 0.7272254638671874, 0.7271760864257812, 0.7273492431640625, 0.7281674194335938, 0.7272499389648438, 0.7273533325195313, 0.7268812866210937, 0.7270850830078125, 0.7263784790039063, 0.7263406372070312, 0.7266806030273437, 0.7265914916992188, 0.7289251708984374, 0.7269837036132812, 0.7268054809570312, 0.7262802124023438, 0.7263119506835938, 0.727593994140625, 1.498050537109375, 0.7272069091796876, 0.7269785766601562, 0.7263672485351562, 0.7267686157226563, 0.7265392456054688, 0.7266918334960938, 0.7279226684570312, 0.7270952758789062, 0.7273717651367188, 0.7269324951171875, 0.7273584594726562, 0.7267799072265625, 0.7266355590820313, 0.7265770874023437, 0.7268864135742188, 0.7267072143554687, 0.7268515625, 0.72646875, 0.7266506958007812, 0.7269765014648437, 0.7271884765625, 0.7277824096679687, 0.7268259887695312, 0.7264942016601562, 0.7269846801757812, 0.7266631469726562, 0.7266221923828124, 0.7265433959960937, 0.7265709228515626, 0.7269048461914063, 0.726640625, 0.7268331298828125, 0.7268157348632812, 0.7275376586914063, 0.7284869384765625, 0.72732568359375, 0.72707275390625, 0.7283681030273438, 0.7273052368164062, 0.727035888671875, 0.726771728515625, 0.7267860717773438, 0.7267205200195312, 0.7269447631835938, 0.7271106567382812, 0.7269212036132813, 0.7263958740234375, 0.7277291259765625, 0.7267225341796875, 0.7280814208984375, 0.726513671875, 0.72703076171875, 0.7273072509765625, 0.7264839477539062, 0.72660888671875, 0.7266047973632812, 0.7268126831054688, 0.7263416137695312, 0.7265904541015625, 0.727103515625, 0.72654541015625, 0.7273564453125, 1.4995804443359375, 0.7264389038085938, 0.72692529296875, 0.726814697265625, 0.7266785278320312, 0.7263928833007812, 0.7267000122070313, 0.726929443359375, 0.7269508666992187, 0.7267072143554687, 0.7272652587890625, 0.727456787109375, 0.7268905029296875, 0.7272182006835938, 0.7271659545898438, 0.7272396850585937, 0.727488525390625, 0.727024658203125, 0.7267123413085937, 0.7267573852539062, 0.72673486328125, 0.7267174682617188, 0.7264542846679688, 0.7265056762695312, 0.7268145141601563, 0.7267010498046875, 0.7269017333984376, 0.7276697387695312, 0.7267901611328125, 0.7262535400390625, 0.7270850830078125, 0.7274118041992188, 0.7274280395507813, 0.7267235717773437, 0.727041015625, 0.7266365356445312, 0.7266611328125, 0.7265853271484375, 0.7269519653320312, 0.7284623413085938, 0.7270697021484375, 0.7270390014648438, 0.7264204711914063, 0.7266826171875, 0.7274915771484375, 0.726845458984375, 0.72681982421875, 0.72749365234375, 0.72808447265625, 0.727667724609375, 0.726656005859375, 0.7265833129882813, 0.726640625, 0.7277496337890625, 0.7269703979492188, 0.7271680297851563, 0.7267593994140625, 0.7272683715820313, 0.727320556640625, 0.7269324951171875, 0.7269560546875, 0.7263651733398437, 0.7267870483398438, 1.500390380859375, 0.7265679321289062, 0.7275233154296875, 0.7270062255859375, 0.7268106079101563, 0.7261430053710938, 0.7260712890625, 0.7270625, 0.7281285400390625, 0.728479736328125, 0.7286661376953125, 0.7285464477539062, 0.7271463623046875, 0.7275120849609376, 0.7265228881835938, 0.7265628051757812, 0.7272130737304687, 0.7273912353515625, 0.7264358520507812, 0.7261306762695312, 0.7267123413085937, 0.7264491577148438, 0.72669287109375, 0.726719482421875, 0.7263631591796875, 0.7265198364257812, 0.7262894287109375, 0.7263252563476562, 0.7271505737304688, 0.7277117309570312, 0.7283435668945313, 0.7273421020507812, 0.7282186279296875, 0.7270390014648438, 0.7294136352539062, 0.7282667236328125, 0.7285493774414062, 0.7274598388671875, 0.7267921752929688, 0.7266416625976563, 0.72656591796875, 0.72635595703125, 0.7266129760742187, 0.7267891235351562, 0.7269498901367187, 0.7263416137695312, 0.72686181640625, 0.7261245727539063, 0.7262761840820312, 0.7264869995117188, 0.7264901123046875, 0.7260047607421874, 0.726487060546875, 0.7265208129882812, 0.7264603881835937, 0.7263908081054687, 0.7275448608398437, 0.727146484375, 0.72680859375, 0.7264603881835937, 0.7277178955078125, 0.72709326171875, 0.728501220703125, 1.5012095947265625, 0.72654541015625, 0.7268945922851563, 0.7272919311523437, 0.7272028198242187, 0.7270809326171875, 0.7276093139648437, 0.7282923583984375, 0.7271884765625, 0.7268485717773437, 0.726763427734375, 0.7262822265625, 0.7274660034179687, 0.7266007080078125, 0.72632421875, 0.7261010131835938, 0.726930419921875, 0.7262833251953125, 0.727236572265625, 0.7263109130859375, 0.727277587890625, 0.729017333984375, 0.7280557861328125, 0.7287122192382812, 0.726213623046875, 0.72635595703125, 0.7278028564453125, 0.7271577758789063, 0.7265413208007813, 0.7267686157226563, 0.7265740966796875, 0.7276277465820312, 0.727436279296875, 0.7270174560546875, 0.7273318481445312, 0.72755712890625, 0.72652392578125, 0.7278878784179688, 0.7272489013671875, 0.7272007446289063, 0.727320556640625, 0.7268229370117187, 0.7260794677734375, 0.7274660034179687, 0.726408203125, 0.726381591796875, 0.7260671997070313, 0.7263037719726563, 0.7268433837890625, 0.726957275390625, 0.7271473388671875, 0.7270225830078125, 0.7275509643554687, 0.7284777221679688, 0.7270584106445312, 0.7272191772460938, 0.727357421875, 0.7268218994140625, 0.7266734008789062, 0.726677490234375, 0.7279493408203125, 0.7279093627929687, 0.7280087280273437, 1.50259716796875, 0.7271526489257812, 0.727667724609375, 0.7279851684570312, 0.7274424438476562, 0.7262618408203125, 0.7266088256835938, 0.7269324951171875, 0.7273543701171875, 0.7273164672851562, 0.7271823120117188, 0.7268035278320313, 0.7278939208984375, 0.7264112548828126, 0.7268864135742188, 0.726513671875, 0.7265802001953126, 0.72677685546875, 0.7268116455078125, 0.7273564453125, 0.72827392578125, 0.72766259765625, 0.7272315063476562, 0.7262218017578125, 0.726414306640625, 0.7278561401367187, 0.728342529296875, 0.7276656494140625, 0.7273707275390625, 0.7264389038085938, 0.7276830444335938, 0.7273564453125, 0.7274208984375, 0.7277598876953125, 0.72755712890625, 0.7271044921875, 0.7269417114257812, 0.7284859008789063, 0.728322021484375, 0.727647216796875, 0.7293214721679687, 0.7274178466796875, 0.7273226318359375, 0.7271823120117188, 0.72783154296875, 0.7280087280273437, 0.7273554077148437, 0.7271423950195313, 0.7276287841796875, 0.7275499267578125, 0.7273082885742187, 0.7275765991210937, 0.7275704345703125, 0.7288955078125, 0.7285176391601562, 0.7280199584960938, 0.727841796875, 0.72796875, 0.7280548095703125, 0.7271168212890625, 0.7272642822265625, 0.7274669799804687, 0.7284869384765625]",tokens/s,1.3547138700439787,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-564M,facebook/xglm-564M,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,1510.83008,1619.525632,0.0,973.078528,855.737856,s,10,0.7501975097656249,0.07501975097656251,0.003494001966293861,0.0744185905456543,0.07687975234985352,0.0807021327972412,0.08376003715515137,"[0.08452451324462891, 0.0724151382446289, 0.07226691436767578, 0.07602333068847657, 0.07370025634765626, 0.072499267578125, 0.07526127624511719, 0.07513692474365234, 0.07233955383300782, 0.07603033447265625]",tokens/s,3412.4346811012333,kWh,8.781473283414488e-07,4.811843592162606e-07,2.329853715734e-06,3.6891854032917087e-06,tokens/kWh,69392012.60299407,MB,1510.83008,1619.525632,0.0,973.078528,915.411456,s,10,44.923669921875,4.492366992187501,0.05186865648995985,4.49665234375,4.532765380859375,4.566419799804687,4.593343334960937,"[4.4801484375, 4.46079150390625, 4.396806640625, 4.52528662109375, 4.43945751953125, 4.50497265625, 4.49965625, 4.4936484375, 4.52282763671875, 4.60007421875]",tokens/s,14.023787484317472,kWh,5.2254403288717625e-05,2.8638463246623156e-05,0.00013152905892685864,0.00021242192546219938,tokens/kWh,296579.5544076776,,s,629,45.51549240112303,0.07236167313374095,0.008872551278845286,0.07199948883056641,0.07308881988525391,0.07345274963378906,0.1417435186767578,"[0.07422156524658204, 0.07438028717041016, 0.07465984344482422, 0.07427993774414063, 0.070793212890625, 0.07357952117919922, 0.07368294525146485, 0.07249817657470703, 0.07192473602294921, 0.0720373764038086, 0.07181107330322266, 0.07195545959472656, 0.07189401245117187, 0.07177318572998047, 0.07202201843261719, 0.07179264068603515, 0.07144448089599609, 0.07195033264160157, 0.07206502532958985, 0.0718704605102539, 0.07257292938232422, 0.07203839874267579, 0.07174246215820312, 0.07193804931640625, 0.0720373764038086, 0.07203123474121094, 0.07179058837890626, 0.07161753845214844, 0.07207730865478515, 0.06936883544921875, 0.06907904052734375, 0.06904524993896484, 0.06900326538085938, 0.0691599349975586, 0.0692305908203125, 0.0704000015258789, 0.07194111633300782, 0.07188582611083984, 0.07171788787841797, 0.07179673767089843, 0.0721244125366211, 0.07180902099609375, 0.07232717132568359, 0.06903193664550782, 0.06900838470458984, 0.0690483169555664, 0.07060889434814453, 0.07242137908935548, 0.06946304321289062, 0.06903091430664063, 0.06926335906982421, 0.06909337615966797, 0.06909645080566407, 0.06918144226074219, 0.06912204742431641, 0.0692152328491211, 0.06930022430419921, 0.06917120361328125, 0.06931763458251954, 0.06909951782226563, 0.06903091430664063, 0.06910975646972656, 0.14174925231933594, 0.07115980529785157, 0.07259852600097656, 0.07286988830566406, 0.07273677062988282, 0.07269068908691406, 0.06918656158447266, 0.07026585388183594, 0.07330303955078125, 0.06960230255126953, 0.06916505432128907, 0.0694824981689453, 0.07036313629150391, 0.0726292495727539, 0.07242649841308593, 0.07292723083496094, 0.07256371307373047, 0.0725749740600586, 0.07288422393798828, 0.0727930908203125, 0.06923571014404296, 0.07285657501220703, 0.07292825317382813, 0.07260671997070313, 0.06921318054199219, 0.06922649383544922, 0.06944051361083985, 0.06901862335205078, 0.06910873413085937, 0.0694814682006836, 0.06917938995361328, 0.0691435546875, 0.06913843536376953, 0.06960230255126953, 0.07092428588867188, 0.07321907043457031, 0.07265586853027343, 0.07284429168701172, 0.07265996551513672, 0.07249919891357422, 0.0727388153076172, 0.06851993560791016, 0.06896537780761719, 0.06924390411376953, 0.06928793334960938, 0.06909951782226563, 0.06925926208496094, 0.06904115295410156, 0.06911590576171875, 0.07245005035400391, 0.07256063842773437, 0.07257087707519531, 0.0727162857055664, 0.0725555191040039, 0.07245414733886718, 0.06913638305664062, 0.06932685089111328, 0.06927565002441406, 0.06910566711425781, 0.06935040283203125, 0.06927974700927735, 0.06926643371582031, 0.06896947479248047, 0.1415485382080078, 0.06951423645019532, 0.06943334197998047, 0.06925619506835938, 0.06917324829101562, 0.06899097442626953, 0.06922752380371094, 0.06912000274658203, 0.06899199676513672, 0.06969446563720703, 0.07301529693603516, 0.06934323120117188, 0.06960537719726563, 0.07306034851074218, 0.07241011047363281, 0.07290675354003906, 0.07274700927734375, 0.0726476821899414, 0.07265996551513672, 0.0726824951171875, 0.07269068908691406, 0.06931763458251954, 0.06954188537597657, 0.06927257537841797, 0.06938623809814454, 0.06953062438964844, 0.06952960205078125, 0.06926233673095702, 0.06965862274169922, 0.06954803466796874, 0.06898995208740234, 0.06898892974853515, 0.0695767059326172, 0.06904729461669921, 0.06909030151367188, 0.0691230697631836, 0.06926131439208984, 0.06881587219238282, 0.06911795043945312, 0.0690145263671875, 0.069106689453125, 0.06901760101318359, 0.06876467132568359, 0.06921625518798828, 0.06918246459960938, 0.06903091430664063, 0.0689797134399414, 0.06893158721923828, 0.0690544662475586, 0.06931865692138672, 0.06913433837890624, 0.0691619873046875, 0.06918758392333985, 0.06922956848144532, 0.06901248168945312, 0.06939955139160156, 0.0690708465576172, 0.06908415985107422, 0.06909747314453125, 0.06877798461914063, 0.0688721923828125, 0.06958284759521484, 0.07407615661621093, 0.14876364135742187, 0.0725524444580078, 0.06927769470214844, 0.07208243560791015, 0.07297433471679687, 0.0731176986694336, 0.07251251220703125, 0.0712273941040039, 0.07263846588134766, 0.07023104095458985, 0.07284735870361328, 0.07289651489257812, 0.07278694152832031, 0.07275827026367188, 0.0733675537109375, 0.07290982055664062, 0.07269888305664063, 0.06938419342041016, 0.07068057250976563, 0.07282994842529297, 0.07284838104248047, 0.07352217864990235, 0.07396147155761719, 0.07278694152832031, 0.0727224349975586, 0.07267123413085938, 0.07291801452636719, 0.07321497344970704, 0.07286579132080079, 0.0730408935546875, 0.06961766052246093, 0.07272755432128906, 0.07313715362548828, 0.07269990539550782, 0.07287091064453124, 0.07342694091796875, 0.07279821014404297, 0.07299993896484375, 0.07290777587890625, 0.07300505828857422, 0.07294976043701172, 0.07384166717529297, 0.07301222229003906, 0.07304806518554688, 0.07300096130371093, 0.06921318054199219, 0.06943846130371094, 0.07083724975585938, 0.07355596923828125, 0.07319039916992187, 0.06939443206787109, 0.06938111877441407, 0.06927155303955078, 0.0692520980834961, 0.06911283111572265, 0.06944153594970703, 0.06942515563964843, 0.06953266906738281, 0.06910975646972656, 0.06917017364501953, 0.06945382690429687, 0.0691568603515625, 0.0722903060913086, 0.1428500518798828, 0.07174861145019532, 0.07272755432128906, 0.07288422393798828, 0.07220531463623046, 0.07277977752685547, 0.07296819305419922, 0.07285759735107422, 0.0730439682006836, 0.06917120361328125, 0.06948044586181641, 0.06915174102783203, 0.06912921905517579, 0.06931148529052734, 0.06940672302246094, 0.06928179168701172, 0.07211110687255859, 0.07298047637939453, 0.07278079986572265, 0.07271218872070312, 0.07268966674804687, 0.06976306915283204, 0.0731504669189453, 0.07382220458984375, 0.07313919830322266, 0.07456563568115235, 0.0726456298828125, 0.07287091064453124, 0.07290777587890625, 0.072595458984375, 0.06927974700927735, 0.06974668884277344, 0.06926131439208984, 0.06974668884277344, 0.0696094741821289, 0.0693565444946289, 0.06932173156738282, 0.06926541137695312, 0.06931763458251954, 0.06929203033447266, 0.06896025848388672, 0.0688875503540039, 0.0692838363647461, 0.06925107574462891, 0.06969344329833985, 0.06929510498046874, 0.06936064147949218, 0.06916607666015626, 0.06940364837646484, 0.06927667236328125, 0.0688721923828125, 0.06915071868896484, 0.06949273681640625, 0.06910361480712891, 0.06893260955810547, 0.06888857269287109, 0.06916710662841796, 0.06919065856933594, 0.06913433837890624, 0.06897663879394532, 0.06965760040283203, 0.06907596588134765, 0.06904524993896484, 0.14172877502441406, 0.06887628936767579, 0.06923980712890625, 0.06865408325195313, 0.06912102508544922, 0.06938317108154297, 0.06938521575927735, 0.06910771179199218, 0.06913843536376953, 0.06909951782226563, 0.07147007751464844, 0.07286374664306641, 0.07196057891845703, 0.07204454040527344, 0.07199948883056641, 0.07233126068115234, 0.07187763214111328, 0.07217971038818359, 0.07222169494628906, 0.07219097900390625, 0.07204966735839843, 0.07197491455078125, 0.07207833862304687, 0.07198822021484375, 0.0723394546508789, 0.07189299011230468, 0.0720547866821289, 0.07195750427246093, 0.0719288330078125, 0.07224626922607422, 0.0721981430053711, 0.0721295394897461, 0.07224729919433594, 0.07194624328613282, 0.07236812591552734, 0.07330508422851563, 0.07221247863769531, 0.07189094543457031, 0.07187763214111328, 0.0722503662109375, 0.07257292938232422, 0.07226982116699218, 0.07204761505126953, 0.07212134552001953, 0.07196876525878906, 0.07214694213867187, 0.07219404602050782, 0.07190940856933593, 0.07228310394287109, 0.0723609619140625, 0.07222476959228516, 0.07254732513427735, 0.07211724853515625, 0.07198617553710937, 0.07232102203369141, 0.0694302749633789, 0.06944563293457032, 0.06930226898193359, 0.06912000274658203, 0.07197286224365235, 0.07240499114990234, 0.07262515258789062, 0.07231488037109375, 0.1474877471923828, 0.07206195068359375, 0.07203635406494141, 0.0718704605102539, 0.07204147338867188, 0.07205785369873047, 0.07204249572753907, 0.07156735992431641, 0.0718551025390625, 0.07209369659423828, 0.07219404602050782, 0.07199334716796875, 0.07205683135986328, 0.0719974365234375, 0.07223910522460937, 0.07229440307617188, 0.0720373764038086, 0.07192985534667969, 0.07207218933105469, 0.07226060485839844, 0.07216844940185548, 0.07213158416748047, 0.07198822021484375, 0.07197695922851563, 0.07198515319824218, 0.07197901153564454, 0.07212748718261719, 0.07194111633300782, 0.07195136260986328, 0.07207936096191406, 0.07236505889892578, 0.07226982116699218, 0.06943231964111328, 0.06955929565429687, 0.069607421875, 0.06953164672851563, 0.07322624206542969, 0.07243981170654297, 0.07254937744140624, 0.07196979522705078, 0.07131033325195313, 0.07205785369873047, 0.07247872161865235, 0.07209471893310547, 0.07210291290283204, 0.06919782257080079, 0.06972108459472656, 0.06934835052490235, 0.06924082946777343, 0.06914765167236328, 0.06955519866943359, 0.06937907409667969, 0.06931148529052734, 0.06938829040527343, 0.06970674896240234, 0.06969548797607422, 0.06920089721679687, 0.07203942108154297, 0.07206809234619141, 0.07200972747802735, 0.07209779357910157, 0.07210291290283204, 0.07222681427001953, 0.14453248596191406, 0.06921011352539062, 0.0692326431274414, 0.06925312042236328, 0.06934220886230469, 0.07179571533203125, 0.07198617553710937, 0.07202098846435546, 0.07208243560791015, 0.07199436950683594, 0.07232921600341796, 0.07199027252197265, 0.07135231781005859, 0.07213772583007813, 0.07198207855224609, 0.07201996612548828, 0.07227597045898437, 0.0715315170288086, 0.07204863739013671, 0.0719319076538086, 0.07195442962646484, 0.0720148468017578, 0.07350169372558593, 0.07210291290283204, 0.07235584259033204, 0.07218790435791016, 0.07257804870605469, 0.07216844940185548, 0.07202713775634766, 0.07208550262451172, 0.07201689910888671, 0.0721786880493164, 0.07204557037353515, 0.07207218933105469, 0.0721070098876953, 0.0719482879638672, 0.07234662628173828, 0.07192985534667969, 0.07208345794677734, 0.07192371368408203, 0.07228108978271484, 0.07198925018310547, 0.07188787078857421, 0.07216639709472657, 0.07209062194824219, 0.07224217224121093, 0.07236914825439453, 0.07234559631347656, 0.07200768280029297, 0.07227903747558594, 0.07189913940429687, 0.06916403198242188, 0.06938009643554688, 0.0693903350830078, 0.06937190246582031, 0.06929714965820312, 0.06908108520507812, 0.06912102508544922, 0.06940876770019531, 0.06922752380371094, 0.06929100799560547, 0.06920191955566406, 0.06866534423828125, 0.14497279357910156, 0.07190630340576172, 0.07215411376953125, 0.07193702697753906, 0.07205785369873047, 0.07186431884765625, 0.07196672058105469, 0.07226060485839844, 0.07261901092529296, 0.07256473541259766, 0.0723609619140625, 0.07218585968017578, 0.07166259002685547, 0.07175373077392579, 0.0723773422241211, 0.0719626235961914, 0.07179878234863281, 0.072015869140625, 0.0722001953125, 0.07199641418457031, 0.07210495758056641, 0.07236300659179687, 0.07228006744384766, 0.07196672058105469, 0.07183257293701172, 0.07211519622802734, 0.07188480377197265, 0.07188070678710938, 0.07184076690673828, 0.07201689910888671, 0.07219404602050782, 0.07205068969726562, 0.07195442962646484, 0.07184076690673828, 0.07215513610839844, 0.07230873870849609, 0.07222067260742188, 0.07214284515380859, 0.07292928314208984, 0.07242649841308593, 0.07191654205322266, 0.07287398529052734, 0.07126732635498047, 0.07227391815185547, 0.07001599884033204, 0.07193292999267578, 0.07222681427001953, 0.07207218933105469, 0.07356108856201173, 0.07293440246582031, 0.07190835571289063, 0.07186943817138672, 0.06945689392089843, 0.06953164672851563, 0.06899814605712891, 0.06878720092773437, 0.06898278045654296, 0.06849638366699219, 0.06950809478759766, 0.07029452514648438, 0.07316889953613281, 0.07317708587646485, 0.07281254577636719, 0.14895103454589845, 0.07272550201416016, 0.07273779296875, 0.07310028839111328, 0.07281459045410156, 0.07278899383544922, 0.0730245132446289, 0.07308595275878907, 0.07223705291748046, 0.072880126953125, 0.07315660858154296, 0.07311974334716796, 0.07316172790527344, 0.07296717071533203, 0.07294464111328125, 0.07245823669433593, 0.07288217926025391, 0.07276441955566407, 0.07286784362792968, 0.07301734161376953, 0.07289548492431641, 0.07284735870361328, 0.07271321868896484, 0.07318630218505859, 0.07307878112792969, 0.07253504180908203, 0.07293644714355468, 0.07297843170166016, 0.07267430114746094, 0.07276748657226563, 0.07329894256591797, 0.07337881469726562, 0.0731668472290039, 0.07399219512939453, 0.07357746887207031, 0.07269580841064453, 0.0729169921875, 0.07252992248535156, 0.07297126770019531, 0.07306240081787109, 0.07285964965820313, 0.07280947113037109, 0.07383859252929688, 0.07331942749023437, 0.07286271667480469, 0.07300812530517578, 0.07352012634277344, 0.07395225524902344, 0.073385986328125, 0.07318630218505859, 0.07355289459228516, 0.07336959838867188, 0.07346995544433593, 0.0732938232421875, 0.07325389099121093, 0.07335321807861328, 0.07319347381591797, 0.0742266845703125, 0.0733337631225586, 0.0729722900390625, 0.0729200668334961, 0.07297740936279297, 0.0695920639038086]",tokens/s,13.819470400466983,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-350m,facebook/opt-350m,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,1308.95872,1044.905984,0.0,398.45888,290.479104,s,10,0.7298112030029297,0.07298112030029298,0.0019203205238493336,0.07210595321655273,0.07568494338989258,0.07613730201721192,0.07649918891906739,"[0.07658966064453125, 0.07404208374023437, 0.0716014404296875, 0.07261046600341797, 0.07435145568847656, 0.07145033264160157, 0.07128550720214843, 0.07136418914794922, 0.07093164825439453, 0.07558441925048828]",tokens/s,3507.7565121862385,kWh,8.692573962940111e-07,4.763112833442766e-07,2.233504646279191e-06,3.579073325917479e-06,tokens/kWh,71526894.44672821,MB,1309.261824,1044.905984,0.0,398.45888,337.28256,s,10,45.10692724609375,4.510692724609375,0.02785183557667033,4.507988037109375,4.524690771484375,4.5543905517578125,4.578150375976563,"[4.4999814453125, 4.51792333984375, 4.49663525390625, 4.50728076171875, 4.5086953125, 4.47609033203125, 4.5180908203125, 4.4837802734375, 4.514359375, 4.58409033203125]",tokens/s,13.966812604255988,kWh,5.335015537424221e-05,2.9239079429191457e-05,0.0001294696562717161,0.00021205889107514978,tokens/kWh,297087.2840114681,,s,629,45.688359863281214,0.07263650216737877,0.008731184173797845,0.07209983825683594,0.0731078628540039,0.07340298156738281,0.14137104125976563,"[0.07185408020019532, 0.07194009399414063, 0.07217356872558593, 0.07269068908691406, 0.07360717010498047, 0.0735498275756836, 0.07288934326171875, 0.07259142303466797, 0.07237420654296875, 0.07188172912597657, 0.07213568115234376, 0.0719452133178711, 0.07213875579833984, 0.07207730865478515, 0.07207014465332032, 0.0719257583618164, 0.06970162963867188, 0.06967910766601562, 0.0696995849609375, 0.07190425872802735, 0.07216028594970703, 0.07225545501708984, 0.07087206268310547, 0.06890290832519531, 0.06927776336669922, 0.06949574279785156, 0.06937395477294922, 0.06961663818359375, 0.06992694091796875, 0.07223088073730469, 0.07267635345458984, 0.07262620544433594, 0.07189091491699219, 0.07197593688964844, 0.07230976104736328, 0.07216025543212891, 0.07246540832519531, 0.07132262420654296, 0.07190016174316406, 0.07246745300292969, 0.07053619384765625, 0.06935552215576171, 0.06963097381591797, 0.07147932434082031, 0.06940361785888671, 0.0716410903930664, 0.07243264007568359, 0.07219302368164063, 0.07221145629882812, 0.07042355346679688, 0.06942002868652344, 0.06980608367919922, 0.06986752319335937, 0.07069696044921875, 0.0722165756225586, 0.0733306884765625, 0.07286374664306641, 0.07210086059570313, 0.07077069091796875, 0.07022796630859375, 0.07062940979003907, 0.07200457763671875, 0.14657945251464843, 0.07262617492675781, 0.07222579193115235, 0.07230054473876953, 0.07205580902099609, 0.07209779357910157, 0.07216537475585938, 0.07003648376464844, 0.07229849243164063, 0.07284019470214843, 0.07227597045898437, 0.07243981170654297, 0.07234457397460937, 0.07199948883056641, 0.07222169494628906, 0.07268147277832031, 0.07375667572021484, 0.07214591979980468, 0.07202201843261719, 0.07222271728515625, 0.07206604766845703, 0.0720025634765625, 0.07187251281738281, 0.0722913589477539, 0.072248291015625, 0.07194624328613282, 0.07201999664306641, 0.07205987548828124, 0.07217356872558593, 0.0712837142944336, 0.06940569305419922, 0.06932991790771484, 0.06942310333251953, 0.06952140808105468, 0.06934835052490235, 0.06939238739013671, 0.06940774536132813, 0.06954188537597657, 0.06909951782226563, 0.0694824981689453, 0.06964530944824218, 0.06932991790771484, 0.07233843231201172, 0.07286271667480469, 0.06985113525390625, 0.06934528350830078, 0.07116902160644531, 0.0729722900390625, 0.0716072998046875, 0.07257907104492188, 0.0729917449951172, 0.07322112274169922, 0.07285657501220703, 0.07319859313964844, 0.07252787017822265, 0.07272038269042969, 0.07294668579101563, 0.0729886703491211, 0.07285350036621094, 0.07272959899902344, 0.07295283508300782, 0.07268556976318359, 0.07238143920898438, 0.1415720977783203, 0.0693934097290039, 0.06967814636230468, 0.06940972900390625, 0.06931148529052734, 0.06929817962646484, 0.06965049743652343, 0.0701173095703125, 0.07237427520751953, 0.07200054168701171, 0.07237423706054688, 0.07208857727050781, 0.07215309143066406, 0.07220838165283203, 0.07202713775634766, 0.0719810562133789, 0.07221862030029297, 0.07206092834472656, 0.07223705291748046, 0.07211007690429687, 0.07236300659179687, 0.07231590270996094, 0.07095603179931641, 0.0694814682006836, 0.06937395477294922, 0.06936780548095703, 0.07176601409912109, 0.07184486389160157, 0.0720343017578125, 0.07231897735595703, 0.07199129486083984, 0.07193299102783203, 0.0723834228515625, 0.07228825378417969, 0.07213260650634766, 0.07228108978271484, 0.07208448028564453, 0.07200153350830078, 0.07205990600585938, 0.07209983825683594, 0.07197491455078125, 0.07196057891845703, 0.0720547866821289, 0.073133056640625, 0.07183769226074219, 0.07257807922363281, 0.0721899185180664, 0.07195340728759765, 0.0722677764892578, 0.07214284515380859, 0.07201487731933594, 0.06975177764892578, 0.06966681671142579, 0.06874931335449219, 0.06921011352539062, 0.06935346984863282, 0.06987059020996093, 0.07226265716552735, 0.07209164428710937, 0.07197081756591797, 0.07196057891845703, 0.0713338851928711, 0.07061196899414063, 0.1456865234375, 0.0722279052734375, 0.07271212768554687, 0.07285043334960938, 0.0725432357788086, 0.07223500823974609, 0.0720865249633789, 0.07014915466308594, 0.06968726348876954, 0.06959820556640625, 0.06964019012451172, 0.0694681625366211, 0.07034880065917969, 0.07004876708984376, 0.06967193603515626, 0.06940160369873047, 0.07108914947509766, 0.07227597045898437, 0.07151821136474609, 0.06967501068115234, 0.06957772827148437, 0.0698071060180664, 0.07123967742919922, 0.07210086059570313, 0.07223603057861328, 0.07214284515380859, 0.07249510192871093, 0.07254835510253907, 0.07294361877441406, 0.07138508605957031, 0.07067545318603516, 0.07234969329833985, 0.07242546844482421, 0.07169741058349609, 0.07204966735839843, 0.06960543823242188, 0.06955718231201172, 0.07074406433105469, 0.07213772583007813, 0.07239683532714844, 0.07193702697753906, 0.07239676666259766, 0.0722012176513672, 0.07219200134277344, 0.07279411315917969, 0.07149260711669922, 0.07210393524169922, 0.07228211212158203, 0.07218380737304687, 0.07210598754882812, 0.07205171203613281, 0.07216435241699219, 0.07028326416015625, 0.06974668884277344, 0.07239577484130859, 0.07243673706054687, 0.07223094177246094, 0.07230358123779297, 0.07244390106201172, 0.07284838104248047, 0.07209062194824219, 0.072163330078125, 0.07201894378662109, 0.14699827575683594, 0.07176294708251953, 0.0707747802734375, 0.0720343017578125, 0.0722135009765625, 0.07226268768310547, 0.0720823974609375, 0.07234976196289063, 0.07255648040771484, 0.07198515319824218, 0.07197798156738282, 0.07256678771972656, 0.07222476959228516, 0.07208755493164062, 0.07251971435546875, 0.07147618865966797, 0.06935346984863282, 0.06952550506591797, 0.06944051361083985, 0.06955213165283203, 0.06944255828857422, 0.06934323120117188, 0.07061299133300782, 0.07203533172607422, 0.07238861083984376, 0.07203225708007813, 0.07213568115234376, 0.0720374755859375, 0.07249091339111328, 0.0709017562866211, 0.06946304321289062, 0.07154790496826172, 0.07224217224121093, 0.07243673706054687, 0.07215615844726563, 0.07217459106445312, 0.07221247863769531, 0.07201894378662109, 0.07098880004882813, 0.06991667175292969, 0.07119564819335937, 0.07190630340576172, 0.0721786880493164, 0.07199334716796875, 0.06973235321044922, 0.06972415924072266, 0.06972930908203125, 0.06958694458007812, 0.06966268920898437, 0.07221145629882812, 0.07257907104492188, 0.07218688201904297, 0.07213772583007813, 0.07218380737304687, 0.07208345794677734, 0.07205068969726562, 0.07378431701660157, 0.07249612426757812, 0.07221862030029297, 0.0720650863647461, 0.07227897644042969, 0.07233740997314453, 0.07240601348876953, 0.1438771514892578, 0.07196463775634766, 0.07234867095947266, 0.07229952239990234, 0.07298047637939453, 0.07293030548095703, 0.07180902099609375, 0.07185408020019532, 0.07234047698974609, 0.07194316864013672, 0.07218278503417969, 0.07320985412597657, 0.07245516967773437, 0.07180287933349609, 0.07000985717773438, 0.0729917449951172, 0.07227187347412109, 0.07225958251953125, 0.07234047698974609, 0.07209677124023438, 0.07210393524169922, 0.07231283569335938, 0.07199436950683594, 0.07164825439453125, 0.06944358062744141, 0.0694302749633789, 0.06980812835693359, 0.06952652740478515, 0.06984703826904297, 0.07122022247314454, 0.07212850952148438, 0.06986752319335937, 0.06965248107910156, 0.06977843475341797, 0.06978047943115234, 0.06932173156738282, 0.0697364501953125, 0.0695572509765625, 0.06944153594970703, 0.07092131042480469, 0.07262095642089844, 0.07213362884521485, 0.0723927001953125, 0.0719974365234375, 0.07194422149658203, 0.0721714859008789, 0.07187967681884766, 0.0721244125366211, 0.07195852661132812, 0.07079424285888672, 0.06962790679931641, 0.06972518157958985, 0.06952345275878906, 0.06946918487548828, 0.06943539428710938, 0.06939238739013671, 0.06983372497558593, 0.06957260894775391, 0.06938521575927735, 0.06923776245117187, 0.07017164611816407, 0.06932582092285157, 0.06944461059570313, 0.1413570556640625, 0.06928076934814453, 0.07196774291992188, 0.0723220443725586, 0.07226573181152343, 0.07199231719970703, 0.07206502532958985, 0.07222476959228516, 0.07213875579833984, 0.06940057373046875, 0.06947840118408204, 0.07286784362792968, 0.07213260650634766, 0.0724295654296875, 0.07200563049316407, 0.07263334655761719, 0.07273677062988282, 0.07225856018066407, 0.0693411865234375, 0.0695367660522461, 0.07020543670654297, 0.07206809234619141, 0.07203020477294922, 0.0722001953125, 0.07224012756347656, 0.07218688201904297, 0.07218694305419922, 0.07206291198730469, 0.07192678070068359, 0.0715857925415039, 0.07136051177978515, 0.0714076156616211, 0.07242240142822266, 0.07221453094482422, 0.07244499206542969, 0.07197689819335938, 0.07119974517822265, 0.07206297302246094, 0.0721981430053711, 0.0724316177368164, 0.07192473602294921, 0.07172608184814454, 0.06988082885742188, 0.06962483215332031, 0.07035497283935546, 0.06945276641845703, 0.06980198669433593, 0.07191449737548829, 0.07256678771972656, 0.07234457397460937, 0.07288422393798828, 0.07266918182373047, 0.07230668640136718, 0.07237324523925781, 0.0723978271484375, 0.07207526397705079, 0.07200057220458984, 0.07223903656005859, 0.07222172546386718, 0.07205680084228516, 0.0722012176513672, 0.07250841522216797, 0.07297948455810546, 0.14137648010253906, 0.06947328186035157, 0.069607421875, 0.06972518157958985, 0.07188377380371094, 0.07261798095703124, 0.07253196716308594, 0.07225138854980469, 0.0723763198852539, 0.07227289581298828, 0.0724869155883789, 0.0720404510498047, 0.07224832153320312, 0.0722790756225586, 0.07235478210449219, 0.07247462463378906, 0.07243059539794922, 0.07224217224121093, 0.07231999969482422, 0.07176908874511718, 0.07236198425292968, 0.07177011108398437, 0.06941900634765626, 0.0709969940185547, 0.07273983764648438, 0.07211827087402344, 0.07241318511962891, 0.07507456207275391, 0.07245209503173829, 0.07226681518554688, 0.07216531372070313, 0.07257190704345703, 0.07180902099609375, 0.07219200134277344, 0.07004978942871094, 0.06961151885986328, 0.06961663818359375, 0.06986041259765625, 0.06977324676513671, 0.06963404846191407, 0.06956646728515625, 0.06976614379882813, 0.06975692749023438, 0.06940774536132813, 0.07025663757324219, 0.06947020721435547, 0.07162265777587891, 0.07189810943603515, 0.07244185638427734, 0.07230054473876953, 0.07227497863769532, 0.07341053009033204, 0.07264460754394532, 0.07066521453857422, 0.06972415924072266, 0.0694620132446289, 0.06928281402587891, 0.06948761749267578, 0.0691251220703125, 0.06890598297119141, 0.06905343627929687, 0.06960025787353516, 0.06945689392089843, 0.14089112854003907, 0.0694128646850586, 0.069570556640625, 0.06914662170410156, 0.06950399780273438, 0.07003545379638672, 0.06975794982910156, 0.06944870758056641, 0.070181884765625, 0.06986444854736328, 0.069607421875, 0.0728453140258789, 0.07321395111083985, 0.07279718780517579, 0.07399935913085938, 0.07330099487304688, 0.07279718780517579, 0.07354777526855469, 0.07338188934326172, 0.073301025390625, 0.07308284759521484, 0.0729917449951172, 0.07306854248046875, 0.0727388153076172, 0.07293440246582031, 0.0727357406616211, 0.07297023773193359, 0.07296102142333985, 0.07303270721435547, 0.07281462097167969, 0.07316886138916015, 0.073059326171875, 0.0725074234008789, 0.07300093078613282, 0.07273785400390625, 0.07161849975585938, 0.06959820556640625, 0.06925823974609376, 0.06970883178710938, 0.06981629180908203, 0.06969036865234375, 0.06949683380126953, 0.06930738830566406, 0.06938419342041016, 0.06971298980712891, 0.0707419204711914, 0.07269376373291016, 0.07251353454589844, 0.07328562927246093, 0.06910873413085937, 0.06949581146240234, 0.06906163024902344, 0.07005286407470702, 0.07312281799316406, 0.07338393402099609, 0.07292108917236328, 0.07231078338623047, 0.07324467468261718, 0.07336345672607422, 0.07308595275878907, 0.07299993896484375, 0.07279001617431641, 0.07310745239257813, 0.14934938049316407, 0.07343206024169922, 0.07323033905029297, 0.07366553497314453, 0.07251865386962891, 0.07295795440673829, 0.07302963256835937, 0.073480224609375, 0.07338902282714843, 0.07340748596191406, 0.0733655014038086, 0.07314534759521485, 0.07351602935791016, 0.07292825317382813, 0.07283715057373047, 0.07243465423583985, 0.07391846466064453, 0.07298047637939453, 0.07312384033203125, 0.07325183868408203, 0.07271321868896484, 0.07309209442138671, 0.07304192352294922, 0.07337062072753907, 0.0729354248046875, 0.07303884887695312, 0.07325081634521484, 0.0731668472290039, 0.07339622497558594, 0.07318016052246094, 0.07310028839111328, 0.07337574768066406, 0.07324774169921874, 0.07300300598144531, 0.07360620880126953, 0.0735692138671875, 0.07393798065185547, 0.07294355010986328, 0.07357542419433594, 0.07405875396728516, 0.07427174377441406, 0.07364198303222656, 0.07310950469970703, 0.07334912109375, 0.0739277114868164, 0.07357746887207031, 0.07303472137451172, 0.07309414672851562, 0.07331635284423828, 0.07301529693603516, 0.0722135009765625, 0.07023104095458985, 0.07103794860839843, 0.06991667175292969, 0.0706355209350586, 0.0701685791015625, 0.06987264251708984, 0.0711445083618164, 0.0702627182006836, 0.0697343978881836, 0.06988800048828125, 0.07153561401367188, 0.07306034851074218]",tokens/s,13.767182754693582,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-6.7b,facebook/opt-6.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,2022.883328,5539.10272,0.0,4892.655616,4542.741504,s,10,5.694402282714844,0.5694402282714843,0.0022278964552512335,0.5688670349121093,0.5707117797851562,0.5731342102050782,0.5750721545410157,"[0.575556640625, 0.5690338134765625, 0.5678692626953125, 0.5687002563476562, 0.5676268920898437, 0.5674249267578125, 0.5684844970703125, 0.5696422119140625, 0.5701734619140625, 0.5698903198242188]",tokens/s,449.56430418883286,kWh,6.703966662839608e-06,3.6734817678128214e-06,3.077389807590622e-05,4.115134650655865e-05,tokens/kWh,6220938.601831633,MB,2022.883328,5539.10272,0.0,4892.655616,4726.280192,s,10,334.554328125,33.4554328125,0.012952548332308819,33.45466796875,33.46486484375,33.476350390625,33.485538828125,"[33.4878359375, 33.44866796875, 33.443125, 33.44916015625, 33.43737890625, 33.4588671875, 33.4546875, 33.4623125, 33.4546484375, 33.45764453125]",tokens/s,1.8831022259697452,kWh,0.00039510093122168826,0.00021654957452975094,0.0018051211817500937,0.0024167716875015325,tokens/kWh,26067.832690116305,,s,629,339.1397302856444,0.5391728621393395,0.06755475317116652,0.53097265625,0.5316083862304688,0.5317822387695312,1.09872478515625,"[0.5310126342773438, 0.53113037109375, 0.5317058715820312, 0.530935791015625, 0.5307023315429688, 0.531651611328125, 0.5313556518554687, 0.53151025390625, 0.5312081909179688, 0.5308344116210938, 0.5306572875976563, 0.531704833984375, 0.5313720092773437, 0.5314744262695312, 0.5315502319335937, 0.5319423828125, 0.5314641723632813, 0.53194140625, 0.5313074951171874, 0.53119384765625, 0.5315758056640625, 0.5314529418945313, 0.5310607299804687, 0.5313535766601563, 0.5317161254882813, 0.531689453125, 0.53146826171875, 0.531725341796875, 0.5313976440429687, 0.53174169921875, 0.5316290283203124, 0.5319515991210938, 0.5317816162109374, 0.5313218383789062, 0.5316331787109375, 0.5317826538085938, 0.5314375610351563, 0.5311324462890625, 0.5311549682617187, 0.5317427368164063, 0.5311119384765625, 0.5312276611328125, 0.5314365234375, 0.5316792602539062, 0.5316904907226563, 0.5320693969726562, 0.531862548828125, 0.5321605224609375, 0.5318092651367188, 0.5320345458984375, 0.5320376586914063, 0.531989501953125, 0.5317529296875, 0.5317816162109374, 0.532279296875, 0.531968994140625, 0.531651611328125, 0.5318492431640625, 0.5322158203125, 0.5313269653320313, 0.5311262817382812, 0.5317109985351562, 1.100517333984375, 0.531336181640625, 0.5311201171875, 0.5306900634765624, 0.5306705932617187, 0.5307289428710937, 0.5306470336914062, 0.5307658081054687, 0.530572265625, 0.5306316528320313, 0.5310679321289062, 0.5309183959960937, 0.5307955322265625, 0.5309910888671875, 0.5315430297851562, 0.531473388671875, 0.5305477294921875, 0.5311221923828126, 0.5307709350585937, 0.530966552734375, 0.5305128784179688, 0.5308323974609375, 0.5309839477539062, 0.5310494995117188, 0.5307473754882812, 0.5309757690429687, 0.5309235229492187, 0.5309337768554687, 0.53066650390625, 0.5311743774414063, 0.53091943359375, 0.5312245483398438, 0.5310628051757813, 0.5319270629882813, 0.5310371704101563, 0.5307791137695312, 0.5309276123046875, 0.530777099609375, 0.5308047485351562, 0.5308313598632812, 0.530946044921875, 0.5307647705078125, 0.5308211059570312, 0.5306654663085938, 0.530524169921875, 0.5307432861328125, 0.5308251953125, 0.53072998046875, 0.530977783203125, 0.5311867065429687, 0.5309869995117188, 0.5312542724609375, 0.5314273071289063, 0.5308969116210938, 0.5310075073242188, 0.5313853149414063, 0.5306746826171875, 0.53089892578125, 0.5307893676757812, 0.5309020385742188, 0.5307484130859375, 0.5309389038085938, 0.5306982421875, 1.0987735595703125, 0.5308344116210938, 0.530502685546875, 0.5315389404296875, 0.5309573364257812, 0.5307207641601562, 0.530703369140625, 0.531072021484375, 0.5305855712890625, 0.5308221435546875, 0.530577392578125, 0.530745361328125, 0.5306163330078125, 0.5306101684570312, 0.5306142578125, 0.5307074584960938, 0.5306920776367188, 0.530682861328125, 0.5304330444335937, 0.5306326904296875, 0.5307760620117188, 0.5308211059570312, 0.5306358032226562, 0.53072998046875, 0.5306920776367188, 0.5310648193359375, 0.5310341186523437, 0.5307627563476562, 0.530850830078125, 0.5307586669921875, 0.5305743408203125, 0.530819091796875, 0.5308047485351562, 0.5305784301757812, 0.5307218017578125, 0.5306890258789062, 0.5307053833007812, 0.5306900634765624, 0.5304514770507812, 0.5309265747070312, 0.5308579711914062, 0.5307044067382812, 0.5306583251953125, 0.530681884765625, 0.5314078979492187, 0.5308272705078125, 0.53083544921875, 0.5314437255859376, 0.5309910888671875, 0.5313760986328125, 0.5309757690429687, 0.5308323974609375, 0.5307709350585937, 0.5311876831054687, 0.5309788208007813, 0.5310750732421875, 0.5309757690429687, 0.5311590576171875, 0.530819091796875, 0.5316198120117187, 0.5310689086914062, 0.5308856201171875, 0.530956298828125, 1.099652099609375, 0.5306685180664062, 0.5309573364257812, 0.5311047973632812, 0.5308251953125, 0.530680908203125, 0.5306971435546874, 0.5309389038085938, 0.5305805053710938, 0.5308150024414062, 0.530714599609375, 0.5305753784179688, 0.530609130859375, 0.5306388549804687, 0.5310023803710937, 0.5311734008789063, 0.5306429443359375, 0.5309942016601562, 0.5309910888671875, 0.5308006591796876, 0.5307330322265625, 0.5312122802734375, 0.5308692626953125, 0.53094091796875, 0.5309368286132813, 0.5310699462890625, 0.531135498046875, 0.5310894165039063, 0.5310156860351563, 0.5310453491210938, 0.5309798583984375, 0.5312214965820312, 0.5310709838867187, 0.5311631469726562, 0.5312348022460938, 0.5308098754882813, 0.5313177490234375, 0.5314129638671875, 0.5311539306640625, 0.5311221923828126, 0.5308692626953125, 0.5309265747070312, 0.5310023803710937, 0.5306695556640625, 0.5307924194335938, 0.5308477172851562, 0.5312583618164063, 0.5308375244140625, 0.5306859741210938, 0.5308375244140625, 0.5310238647460938, 0.5310392456054688, 0.5313167114257813, 0.5309163818359375, 0.530735107421875, 0.5309808349609375, 0.5313843383789062, 0.53096240234375, 0.530924560546875, 0.5308661499023437, 0.5307893676757812, 0.5308641357421875, 0.5307678833007813, 1.098599365234375, 0.5306071166992188, 0.5305692138671875, 0.5305211181640626, 0.53089794921875, 0.5305497436523438, 0.5308221435546875, 0.5307658081054687, 0.5307996215820312, 0.530661376953125, 0.5306900634765624, 0.5311692504882812, 0.530609130859375, 0.5308006591796876, 0.5306644287109376, 0.5308375244140625, 0.5307730102539062, 0.5307525024414063, 0.5305477294921875, 0.5310003051757812, 0.53079345703125, 0.5306757202148438, 0.5304637451171875, 0.5307637939453125, 0.5306132202148437, 0.530956298828125, 0.530988037109375, 0.5308764038085938, 0.5306808471679687, 0.5308743896484375, 0.5307279663085938, 0.5307576293945313, 0.5310965576171875, 0.5306583251953125, 0.53056103515625, 0.5305681762695312, 0.530555908203125, 0.5308845825195313, 0.5308344116210938, 0.5308221435546875, 0.530787353515625, 0.5306757202148438, 0.5306941528320313, 0.5308262329101563, 0.5310812377929688, 0.530934814453125, 0.5305538330078124, 0.5308098754882813, 0.5308231811523437, 0.5310986328125, 0.5306429443359375, 0.530845703125, 0.5307422485351563, 0.5307197265625, 0.530492431640625, 0.5308323974609375, 0.5305548706054688, 0.53087744140625, 0.5306767578125, 0.5308753662109374, 0.5305855712890625, 0.5308129272460937, 0.5303736572265625, 1.098567626953125, 0.5309030151367188, 0.5317017822265625, 0.5314866943359375, 0.531140625, 0.5313515625, 0.5313720092773437, 0.530951171875, 0.5308917846679687, 0.530724853515625, 0.5307863159179688, 0.5310587158203125, 0.530777099609375, 0.530819091796875, 0.53087744140625, 0.5309696044921876, 0.5308897094726562, 0.5310904541015625, 0.531515380859375, 0.5310259399414062, 0.5307739868164062, 0.5312061157226563, 0.5318901977539062, 0.5310761108398437, 0.5309962158203125, 0.5314529418945313, 0.5309951782226563, 0.5310013427734375, 0.5313760986328125, 0.5313331298828124, 0.5312849731445313, 0.5316055297851563, 0.5308313598632812, 0.5314816284179688, 0.531324951171875, 0.5315655517578125, 0.5319567260742187, 0.5317283935546875, 0.5308856201171875, 0.5309910888671875, 0.5312266235351563, 0.53136279296875, 0.5310289916992188, 0.5309757690429687, 0.530524169921875, 0.5308047485351562, 0.5308436279296875, 0.5307371215820312, 0.5306552124023437, 0.5307576293945313, 0.530914306640625, 0.530629638671875, 0.530746337890625, 0.5308231811523437, 0.5308712768554688, 0.5312450561523437, 0.5309327392578125, 0.53103515625, 0.5309214477539063, 0.5310279541015624, 0.5311580200195313, 0.5310894165039063, 0.5313126220703125, 1.0993970947265626, 0.5308150024414062, 0.53054052734375, 0.53062451171875, 0.5311447143554687, 0.5309869995117188, 0.5307893676757812, 0.5311334228515625, 0.5309542236328125, 0.5308313598632812, 0.5310842895507812, 0.5308108520507813, 0.5306603393554687, 0.531177490234375, 0.5307023315429688, 0.5314396362304687, 0.5306491088867188, 0.5312952270507812, 0.5308108520507813, 0.530840576171875, 0.5308078002929687, 0.5307863159179688, 0.5308047485351562, 0.5311262817382812, 0.53103515625, 0.5310740356445313, 0.5309634399414063, 0.531051513671875, 0.5307965698242187, 0.5314877319335938, 0.530746337890625, 0.5310259399414062, 0.5311201171875, 0.5307340698242188, 0.5309931640625, 0.5309593505859375, 0.5321942749023437, 0.5316433715820312, 0.5312388916015625, 0.530872314453125, 0.5307698974609375, 0.5307975463867187, 0.53082421875, 0.5306982421875, 0.5306019897460937, 0.5307197265625, 0.5305917358398438, 0.5307258911132813, 0.53085693359375, 0.531398681640625, 0.53166796875, 0.5317969970703125, 0.531862548828125, 0.5320745239257813, 0.5316608276367187, 0.5313966064453125, 0.5307791137695312, 0.53075048828125, 0.5309102172851563, 0.5311057739257813, 0.5310842895507812, 0.5312010498046875, 0.5308037109375, 1.1007191162109375, 0.53106689453125, 0.5309931640625, 0.5313853149414063, 0.5308897094726562, 0.5309972534179688, 0.5310842895507812, 0.531324951171875, 0.5306767578125, 0.5316229248046875, 0.5309573364257812, 0.530756591796875, 0.5307310180664062, 0.53082421875, 0.5305927734375, 0.5316853637695312, 0.5315819702148438, 0.5312542724609375, 0.5311344604492187, 0.531040283203125, 0.5308897094726562, 0.5313392944335937, 0.5316823120117188, 0.53125634765625, 0.5309573364257812, 0.5317652587890624, 0.531435546875, 0.531162109375, 0.5312890625, 0.5313095703125, 0.5313320922851562, 0.5313013916015625, 0.5311661987304688, 0.531324951171875, 0.5308743896484375, 0.53102490234375, 0.5308999633789062, 0.53098291015625, 0.5310023803710937, 0.5311641845703124, 0.5315082397460937, 0.5309317016601562, 0.5308917846679687, 0.5309010009765625, 0.5306695556640625, 0.531409912109375, 0.5308856201171875, 0.531504150390625, 0.5308948364257813, 0.5309337768554687, 0.5309696044921876, 0.5308682250976563, 0.5308712768554688, 0.5310637817382813, 0.5310596923828125, 0.530924560546875, 0.5309398803710937, 0.5311754150390625, 0.5311918334960938, 0.5315389404296875, 0.5313822631835937, 0.5316004028320312, 0.5323612060546875, 1.101580322265625, 0.5309798583984375, 0.5314283447265625, 0.53097265625, 0.5306849365234375, 0.53102490234375, 0.5308344116210938, 0.5312901000976562, 0.5309429931640625, 0.5312133178710937, 0.53045556640625, 0.5307914428710937, 0.53096142578125, 0.5311590576171875, 0.531051513671875, 0.5311539306640625, 0.5308795166015625, 0.5314437255859376, 0.5311702880859375, 0.5311856689453125, 0.5311795043945312, 0.5310105590820312, 0.5306644287109376, 0.530904052734375, 0.5307525024414063, 0.530788330078125, 0.5305599975585937, 0.5308651733398437, 0.5308682250976563, 0.5307955322265625, 0.5306275634765625, 0.5308323974609375, 0.53082421875, 0.5314150390625, 0.5311631469726562, 0.53131982421875, 0.5312214965820312, 0.5309327392578125, 0.5311282958984375, 0.5312542724609375, 0.530845703125, 0.5310156860351563, 0.53068798828125, 0.5308897094726562, 0.5308231811523437, 0.5310628051757813, 0.531072021484375, 0.5317027587890625, 0.5310494995117188, 0.5310187377929687, 0.5309296875, 0.531330078125, 0.5311447143554687, 0.53096044921875, 0.5310719604492188, 0.531030029296875, 0.5310156860351563, 0.53125634765625, 0.5312440185546875, 0.5312337646484375, 0.5311641845703124, 0.5312235717773437, 0.530788330078125, 1.10097412109375, 0.5311385498046876, 0.5308712768554688, 0.5309481201171875, 0.5307709350585937, 0.5306480712890626, 0.5311713256835937, 0.5311160278320313, 0.5307095336914063, 0.5308487548828125, 0.53075048828125, 0.5309020385742188, 0.5306849365234375, 0.5309808349609375, 0.53065625, 0.5309685668945312, 0.5306695556640625, 0.5307422485351563, 0.5307944946289063, 0.53087744140625, 0.53075048828125, 0.5307781372070313, 0.5308733520507812, 0.5309255981445312, 0.5307822265625, 0.531041259765625, 0.5312808837890625, 0.5311181030273437, 0.531198974609375, 0.5309603881835937, 0.5314375610351563, 0.5310955810546875, 0.5312604370117188, 0.5309583129882812, 0.5311651611328125, 0.530998291015625, 0.53083544921875, 0.5311876831054687, 0.531009521484375, 0.531167236328125, 0.5306757202148438, 0.531293212890625, 0.5311754150390625, 0.53102490234375, 0.5311047973632812, 0.5315072021484375, 0.531167236328125, 0.5316218872070313, 0.5310310668945313, 0.531662841796875, 0.5310525512695312, 0.5312481079101562, 0.5310341186523437, 0.5309962158203125, 0.5309798583984375, 0.5312184448242188, 0.5310965576171875, 0.5310013427734375, 0.5311498413085938, 0.53165771484375, 0.5311181030273437, 0.531272705078125, 0.5313116455078125]",tokens/s,1.8546927529553001,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-72B,Qwen/Qwen-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 550, in from_pretrained + model_class = get_class_from_dynamic_module( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 498, in get_class_from_dynamic_module + final_module = get_cached_module_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 323, in get_cached_module_file + modules_needed = check_imports(resolved_module_file) + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 181, in check_imports + raise ImportError( +ImportError: This modeling file requires the following packages that were not found in your environment: transformers_stream_generator. Run `pip install transformers_stream_generator` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-6B,,cuda,0,42,,,,,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,,MB,1575.784448,5448.925184,0.0,4802.47808,4489.252352,s,10,5.114569244384766,0.5114569244384766,0.0014045752365005437,0.51150537109375,0.5130841125488281,0.5135362274169921,0.5138979193115234,"[0.512983642578125, 0.5139883422851562, 0.5091934204101562, 0.5101927795410156, 0.5104193725585937, 0.5101881103515625, 0.5112746887207031, 0.5121220092773437, 0.5117360534667968, 0.5124708251953125]",tokens/s,500.5309103617276,kWh,6.020330099595917e-06,3.2988774182740595e-06,2.7782869448500503e-05,3.710207696637048e-05,tokens/kWh,6899883.266158921,MB,1575.784448,5448.925184,0.0,4802.47808,4557.794816,s,10,301.333625,30.133362499999997,0.011965382718741807,30.1356015625,30.146014453125,30.1513353515625,30.1555920703125,"[30.12017578125, 30.120552734375, 30.1369375, 30.140724609375, 30.15665625, 30.137947265625, 30.123359375, 30.14483203125, 30.118173828125, 30.134265625]",tokens/s,2.0907059409649356,kWh,0.00035569889715148347,0.00019495335105611957,0.001580753417379499,0.002131405665587102,tokens/kWh,29557.958401431977,,s,629,305.4536612854004,0.48561790347440437,0.060780789003776696,0.4781598815917969,0.47936102294921873,0.4797579284667969,0.9888164648437501,"[0.47731610107421873, 0.4777103271484375, 0.4780533752441406, 0.4779223022460938, 0.47746868896484373, 0.47739285278320315, 0.4777676696777344, 0.47874050903320314, 0.47792022705078124, 0.4776212463378906, 0.47766937255859376, 0.4774625244140625, 0.4779018249511719, 0.47790591430664064, 0.47894528198242187, 0.47799398803710935, 0.47878759765625, 0.47878964233398436, 0.47852645874023436, 0.47817422485351563, 0.4778905334472656, 0.47780557250976563, 0.47809637451171877, 0.47767245483398435, 0.477765625, 0.4778649597167969, 0.4779622497558594, 0.47779531860351565, 0.47874456787109376, 0.47817214965820315, 0.4782950439453125, 0.47745944213867186, 0.47759051513671874, 0.4785008544921875, 0.47783013916015626, 0.4780267639160156, 0.4782438354492187, 0.4785858459472656, 0.4793190307617187, 0.47905487060546875, 0.47873635864257813, 0.47829400634765623, 0.4777697143554688, 0.4778240661621094, 0.47742047119140624, 0.4780851135253906, 0.4782008361816406, 0.47801651000976564, 0.47831243896484377, 0.47876199340820313, 0.4778547058105469, 0.47859405517578124, 0.4782591857910156, 0.47776461791992186, 0.47824075317382814, 0.4776304626464844, 0.4778383483886719, 0.4776642456054688, 0.47760385131835936, 0.478160888671875, 0.47876300048828124, 0.4783943786621094, 0.9927915649414063, 0.47756494140625, 0.4774000549316406, 0.47765298461914063, 0.47747378540039065, 0.4782704772949219, 0.477907958984375, 0.4783944091796875, 0.477813720703125, 0.47758950805664063, 0.47759051513671874, 0.4776069030761719, 0.47729867553710936, 0.4773724060058594, 0.47751473999023436, 0.47745742797851565, 0.4780113525390625, 0.4786903076171875, 0.47860427856445314, 0.4783482971191406, 0.47794790649414065, 0.47776461791992186, 0.4778291320800781, 0.4779427795410156, 0.4779376525878906, 0.47769189453125, 0.47908352661132814, 0.47773797607421875, 0.47901287841796875, 0.4777134094238281, 0.4779151306152344, 0.47777587890625, 0.47748709106445314, 0.47771136474609377, 0.4781527099609375, 0.47798886108398436, 0.47750042724609376, 0.47809127807617186, 0.47905892944335937, 0.4782591857910156, 0.4781598815917969, 0.4782981262207031, 0.4776468505859375, 0.47786392211914064, 0.4776560668945313, 0.478023681640625, 0.48116427612304685, 0.4793456726074219, 0.4777492370605469, 0.47873126220703127, 0.47907122802734375, 0.4783626098632813, 0.4779949951171875, 0.47790286254882813, 0.4776908874511719, 0.478065673828125, 0.4777738342285156, 0.477770751953125, 0.47832369995117185, 0.47939175415039065, 0.47875686645507814, 0.47773284912109376, 0.478023681640625, 0.988906494140625, 0.47783013916015626, 0.4776437683105469, 0.4778649597167969, 0.4783902587890625, 0.47753521728515624, 0.4795494384765625, 0.47817214965820315, 0.4789770202636719, 0.4776386413574219, 0.47787213134765627, 0.47758950805664063, 0.47774514770507814, 0.47762841796875, 0.47790386962890624, 0.479541259765625, 0.4791797790527344, 0.47914700317382813, 0.4779294738769531, 0.4777625732421875, 0.47816497802734376, 0.4782356567382812, 0.47807180786132814, 0.4797726745605469, 0.4777687072753906, 0.4803246154785156, 0.4800747375488281, 0.4793231506347656, 0.4796539001464844, 0.48039935302734377, 0.4796252136230469, 0.4790927429199219, 0.477655029296875, 0.4791510925292969, 0.4787394409179688, 0.47877017211914064, 0.47938558959960936, 0.47959552001953126, 0.47838516235351564, 0.47773593139648435, 0.478129150390625, 0.4781404113769531, 0.4776857604980469, 0.4779018249511719, 0.47747378540039065, 0.4776847229003906, 0.4781803588867187, 0.4799620971679687, 0.4786053161621094, 0.47794790649414065, 0.4777123718261719, 0.4776683654785156, 0.4774696960449219, 0.47712460327148437, 0.4774481811523438, 0.47752908325195315, 0.47732632446289064, 0.4782438354492187, 0.47871487426757814, 0.47837286376953125, 0.4777123718261719, 0.47754238891601564, 0.4779929504394531, 0.9885849609375, 0.47834625244140627, 0.47808615112304687, 0.47836468505859375, 0.47836468505859375, 0.4776908874511719, 0.47762738037109376, 0.4779254455566406, 0.477751220703125, 0.4784117736816406, 0.47790286254882813, 0.4781741943359375, 0.4787138671875, 0.47869439697265626, 0.47841998291015625, 0.4790876159667969, 0.47936920166015623, 0.4789770202636719, 0.4778157958984375, 0.47849676513671874, 0.4780349426269531, 0.4785776672363281, 0.4781475830078125, 0.4785581970214844, 0.4792135620117188, 0.47874969482421875, 0.4787026062011719, 0.47837899780273435, 0.47872308349609377, 0.47839334106445314, 0.4786411437988281, 0.4790169677734375, 0.47768063354492185, 0.4778465270996094, 0.4781793212890625, 0.47906610107421876, 0.47883367919921876, 0.4782213134765625, 0.4785172424316406, 0.47826739501953125, 0.478087158203125, 0.4777277526855469, 0.4778680419921875, 0.47746456909179685, 0.4802908020019531, 0.47867086791992186, 0.47896060180664063, 0.4792279052734375, 0.4781783142089844, 0.47810867309570315, 0.4786063232421875, 0.4785162353515625, 0.4785848388671875, 0.4778659973144531, 0.4776212463378906, 0.4777840576171875, 0.4788787231445312, 0.47973171997070313, 0.47971328735351565, 0.4792842102050781, 0.477655029296875, 0.47770315551757814, 0.4779049377441406, 0.9883463134765625, 0.4791357421875, 0.4782458801269531, 0.4794173583984375, 0.478166015625, 0.47800833129882814, 0.4775157775878906, 0.4780421142578125, 0.47852032470703126, 0.4791224365234375, 0.47801651000976564, 0.47799398803710935, 0.4782530517578125, 0.47967129516601564, 0.47834417724609374, 0.4779847717285156, 0.4790302734375, 0.47956378173828124, 0.4773304443359375, 0.4774420471191406, 0.4774143981933594, 0.4773447570800781, 0.47814349365234377, 0.47828582763671873, 0.479710205078125, 0.4788910217285156, 0.479056884765625, 0.47832369995117185, 0.4782438354492187, 0.4783964233398437, 0.479025146484375, 0.4786606140136719, 0.4782909545898438, 0.47821823120117185, 0.4802129821777344, 0.47790286254882813, 0.47835751342773436, 0.4782561340332031, 0.478635009765625, 0.47870156860351565, 0.4777001037597656, 0.4775802917480469, 0.47765914916992186, 0.479578125, 0.4784527282714844, 0.47922894287109374, 0.47923199462890625, 0.4788213806152344, 0.47869439697265626, 0.4787804260253906, 0.47939993286132815, 0.47935186767578125, 0.4787639770507813, 0.4786268310546875, 0.4793589782714844, 0.48041677856445314, 0.48034817504882815, 0.482777099609375, 0.4785745849609375, 0.478571533203125, 0.47958123779296874, 0.47865029907226564, 0.4787384338378906, 0.990581787109375, 0.4781793212890625, 0.4788521423339844, 0.47797039794921875, 0.47851007080078123, 0.47790286254882813, 0.4778321838378906, 0.47813427734375, 0.4789043273925781, 0.47835134887695313, 0.47826226806640626, 0.47895858764648436, 0.4790947875976562, 0.4783759460449219, 0.47830938720703126, 0.4781803588867187, 0.477991943359375, 0.47756903076171875, 0.47764480590820313, 0.47726080322265624, 0.47773797607421875, 0.47950848388671874, 0.47766015625, 0.47899853515625, 0.47830322265625, 0.4781537170410156, 0.47825204467773436, 0.4785592346191406, 0.479025146484375, 0.47944705200195314, 0.47835134887695313, 0.478445556640625, 0.47876199340820313, 0.47891659545898435, 0.4791101379394531, 0.47820391845703125, 0.4782581787109375, 0.47862374877929686, 0.4814622802734375, 0.47808819580078127, 0.47835751342773436, 0.47878964233398436, 0.4785551452636719, 0.4777082824707031, 0.4787271728515625, 0.4784322509765625, 0.47835546875, 0.4775679931640625, 0.4779346008300781, 0.47834521484375, 0.478160888671875, 0.47767962646484374, 0.47781170654296873, 0.4783585205078125, 0.4791285705566406, 0.4783697814941406, 0.47800216674804685, 0.47763250732421875, 0.4782294921875, 0.47773696899414064, 0.47801651000976564, 0.4774718017578125, 0.47802670288085936, 0.9896888427734375, 0.47829608154296877, 0.47802975463867187, 0.47756497192382813, 0.47750039672851563, 0.47811380004882814, 0.4781025390625, 0.4778936767578125, 0.47798370361328124, 0.47788851928710935, 0.4786268310546875, 0.4782899169921875, 0.4779632568359375, 0.47777484130859377, 0.47753521728515624, 0.4779007873535156, 0.47758544921875, 0.47753726196289065, 0.47787213134765627, 0.47847628784179685, 0.4782909545898438, 0.4778014831542969, 0.47860223388671874, 0.47986483764648435, 0.47783935546875, 0.47765298461914063, 0.4779980773925781, 0.47794073486328126, 0.4787517395019531, 0.4781506652832031, 0.4780155029296875, 0.47900875854492186, 0.47818548583984377, 0.47805029296875, 0.4782233581542969, 0.47802471923828127, 0.4779346008300781, 0.47827557373046875, 0.478497802734375, 0.47849063110351564, 0.47899237060546873, 0.4786196594238281, 0.47815167236328127, 0.4799846801757813, 0.47795709228515626, 0.47788851928710935, 0.4778270568847656, 0.4778547058105469, 0.478060546875, 0.47782608032226564, 0.4778434143066406, 0.477601806640625, 0.4788572082519531, 0.47828274536132814, 0.47832986450195314, 0.4779417724609375, 0.477633544921875, 0.477907958984375, 0.478065673828125, 0.4780482482910156, 0.47808819580078127, 0.47858380126953126, 0.47811892700195313, 0.9911572265625, 0.4780707702636719, 0.47788134765625, 0.4775536499023437, 0.4774696960449219, 0.47756185913085936, 0.47758642578125, 0.4777615356445313, 0.47794790649414065, 0.4787118225097656, 0.47779736328125, 0.47749325561523437, 0.4775475158691406, 0.4777697143554688, 0.4781363220214844, 0.4779141235351563, 0.4792012939453125, 0.4778874816894531, 0.47856845092773437, 0.47768267822265625, 0.4781311950683594, 0.4785070190429688, 0.47809127807617186, 0.4776365966796875, 0.4780707702636719, 0.47816497802734376, 0.47828582763671873, 0.47777279663085936, 0.4780728454589844, 0.47835751342773436, 0.47851211547851563, 0.47822235107421873, 0.4775055236816406, 0.4781465454101563, 0.47790386962890624, 0.4779018249511719, 0.478359619140625, 0.4794295654296875, 0.4784394226074219, 0.4795248718261719, 0.4780544128417969, 0.477949951171875, 0.48094821166992185, 0.4794306640625, 0.4784814147949219, 0.4792197265625, 0.4782981262207031, 0.47831655883789065, 0.47831243896484377, 0.4784998474121094, 0.47935488891601563, 0.47914599609375, 0.48005630493164064, 0.48013311767578126, 0.47997030639648436, 0.48005426025390624, 0.4799140625, 0.4794142150878906, 0.4800153503417969, 0.4793067626953125, 0.47961505126953125, 0.47790789794921873, 0.4778670043945312, 0.9900185546875, 0.4775372924804687, 0.47779840087890624, 0.477633544921875, 0.4782950439453125, 0.47761920166015626, 0.47800421142578126, 0.478803955078125, 0.4791029663085937, 0.4782643127441406, 0.47842098999023436, 0.4776908874511719, 0.4780185546875, 0.47770932006835937, 0.4777851257324219, 0.477681640625, 0.477812744140625, 0.47979006958007814, 0.47791717529296873, 0.4780451965332031, 0.4788193359375, 0.477955078125, 0.47790591430664064, 0.47803903198242187, 0.47824697875976563, 0.4786861572265625, 0.47816192626953125, 0.477655029296875, 0.47835134887695313, 0.478919677734375, 0.47831039428710936, 0.4777205810546875, 0.4773253173828125, 0.47773284912109376, 0.4778260498046875, 0.4776069030761719, 0.47773492431640624, 0.4777461853027344, 0.478587890625, 0.47767041015625, 0.4789801330566406, 0.47915618896484374, 0.4780257263183594, 0.47762841796875, 0.4778168334960938, 0.4776714172363281, 0.47783526611328125, 0.47757107543945315, 0.4779704284667969, 0.4786677856445313, 0.47908148193359373, 0.4783206481933594, 0.4781240234375, 0.47756494140625, 0.4773294067382812, 0.47757208251953126, 0.477348876953125, 0.47757720947265625, 0.47773492431640624, 0.47816192626953125, 0.47770932006835937, 0.4789381103515625, 0.47819366455078127, 0.9904609375, 0.47848550415039065, 0.478013427734375, 0.4775751647949219, 0.4774912109375, 0.4782847900390625, 0.4794931335449219, 0.4796549072265625, 0.47851318359375, 0.4785254211425781, 0.47859506225585935, 0.4788695068359375, 0.47816293334960935, 0.47827969360351563, 0.4785244140625, 0.4797358093261719, 0.4782438354492187, 0.4776570739746094, 0.4789125061035156, 0.47882650756835937, 0.4777082824707031, 0.47753216552734373, 0.47809637451171877, 0.47831655883789065, 0.4776488952636719, 0.4774471740722656, 0.4796170349121094, 0.47986483764648435, 0.4793456726074219, 0.479599609375, 0.4791654357910156, 0.4776488952636719, 0.4775475158691406, 0.47792538452148436, 0.4778240051269531, 0.4774819946289062, 0.4782612915039062, 0.4777738037109375, 0.4777543640136719, 0.47963134765625, 0.47828070068359374, 0.4778486022949219, 0.4777625427246094, 0.477601806640625, 0.4777185363769531, 0.47762432861328125, 0.47770932006835937, 0.4777062377929687, 0.47948800659179686, 0.4784066467285156, 0.47847015380859376, 0.4775106506347656, 0.4777420654296875, 0.47842098999023436, 0.47805233764648436, 0.47766015625, 0.47758746337890623, 0.47821926879882815, 0.4781087036132812, 0.4776806640625, 0.47913876342773437, 0.47851828002929686, 0.4783575744628906]",tokens/s,2.059232151132392,,,main,False,False,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen2-beta-72B,Qwen/Qwen2-beta-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 96.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-2.7B,,cuda,0,42,,,,,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,,MB,2236.862464,2932.342784,0.0,2285.89568,2082.706944,s,10,2.5091812438964842,0.25091812438964844,0.0016359490817563227,0.25024176025390626,0.25317278594970705,0.2535381980895996,0.25383052780151366,"[0.25309158325195313, 0.2539036102294922, 0.24987110900878906, 0.24888946533203124, 0.2497696075439453, 0.24914796447753906, 0.2504080047607422, 0.25179104614257813, 0.25223333740234377, 0.2500755157470703]",tokens/s,1020.253122897013,kWh,2.948670335578402e-06,1.6157431971317612e-06,1.3066561943753649e-05,1.763097547646381e-05,tokens/kWh,14519899.953450851,MB,2238.38208,2959.60576,0.0,2313.158656,2180.685312,s,10,143.8805810546875,14.38805810546875,0.011353783216855234,14.382828125,14.40145888671875,14.403668896484376,14.405436904296876,"[14.4009677734375, 14.40587890625, 14.39833203125, 14.3805361328125, 14.374423828125, 14.3830166015625, 14.400140625, 14.373748046875, 14.3808974609375, 14.3826396484375]",tokens/s,4.378631191102458,kWh,0.000169606630340582,9.295822156634463e-05,0.0007504743544440634,0.0010130392063509902,tokens/kWh,62189.10344736671,,s,629,145.90565167236326,0.2319644700673502,0.029875164751919897,0.22824755859375,0.22927770080566406,0.22946876831054688,0.47820606079101563,"[0.22992076110839843, 0.22815129089355468, 0.22794752502441407, 0.22817485046386718, 0.2277969970703125, 0.22800997924804686, 0.22908108520507814, 0.22838169860839844, 0.2285506591796875, 0.2286049346923828, 0.22830694580078126, 0.22801516723632811, 0.22815020751953125, 0.22767308044433593, 0.22776422119140624, 0.2276741180419922, 0.22829362487792967, 0.22767205810546876, 0.22779493713378907, 0.2277058563232422, 0.22765670776367186, 0.22813594055175782, 0.22820147705078125, 0.2283335723876953, 0.2275809326171875, 0.22812364196777343, 0.2283008575439453, 0.22861715698242188, 0.2281564178466797, 0.22812979125976562, 0.2289090576171875, 0.2288476104736328, 0.22834585571289062, 0.22857522583007814, 0.2290155487060547, 0.22790553283691406, 0.22831922912597657, 0.22821784973144532, 0.22858444213867188, 0.2286612548828125, 0.22839808654785157, 0.22902169799804686, 0.22916812133789063, 0.22946917724609375, 0.22901350402832032, 0.22913536071777343, 0.22909747314453124, 0.229138427734375, 0.22936679077148436, 0.22892851257324218, 0.22936268615722658, 0.22932582092285156, 0.22920909118652344, 0.22904730224609374, 0.2291425323486328, 0.22898892211914063, 0.22877183532714843, 0.2286202850341797, 0.229064697265625, 0.22876876831054688, 0.2292131805419922, 0.22902476501464844, 0.48329931640625, 0.2282854461669922, 0.228822021484375, 0.2279741516113281, 0.22834687805175782, 0.22893157958984375, 0.22888038635253907, 0.22889677429199218, 0.22806629943847656, 0.22842678833007812, 0.22791778564453125, 0.22817791748046876, 0.22819737243652344, 0.228274169921875, 0.22817791748046876, 0.2278656005859375, 0.22779391479492186, 0.22778982543945311, 0.22769561767578125, 0.22855783081054687, 0.2280273895263672, 0.22766490173339843, 0.2277734375, 0.2279331817626953, 0.22785125732421874, 0.2278594512939453, 0.227852294921875, 0.2281492462158203, 0.2276505584716797, 0.22739762878417968, 0.22808883666992188, 0.22794650268554686, 0.22817485046386718, 0.228094970703125, 0.22861415100097657, 0.22928793334960937, 0.22927769470214843, 0.22896333312988282, 0.22915072631835937, 0.22908006286621094, 0.22935040283203126, 0.22938829040527345, 0.22915583801269532, 0.22902578735351561, 0.22928282165527344, 0.22933914184570312, 0.23262924194335938, 0.22968319702148438, 0.22875648498535156, 0.22789529418945312, 0.2278666229248047, 0.2295767059326172, 0.22941900634765625, 0.22929306030273439, 0.22928282165527344, 0.22947532653808594, 0.2292490234375, 0.22923365783691407, 0.22946815490722655, 0.22925106811523438, 0.22942207336425782, 0.22925315856933592, 0.2293155517578125, 0.4786954345703125, 0.22789324951171874, 0.22778778076171874, 0.22853427124023437, 0.2276505584716797, 0.22818611145019532, 0.22808677673339844, 0.22816160583496095, 0.22942201232910156, 0.22774887084960938, 0.22915277099609374, 0.2292162628173828, 0.22927772521972656, 0.22797821044921876, 0.2286940155029297, 0.22868377685546876, 0.22756658935546875, 0.22776832580566406, 0.22760652160644532, 0.22787379455566406, 0.2284400634765625, 0.22826905822753907, 0.22876876831054688, 0.22959616088867188, 0.22788198852539063, 0.22754815673828124, 0.2290493469238281, 0.22905958557128905, 0.22827314758300782, 0.22785842895507813, 0.22869094848632812, 0.2282854461669922, 0.22885580444335937, 0.22893157958984375, 0.2289971160888672, 0.22875340270996095, 0.22852301025390626, 0.22836224365234375, 0.22938829040527345, 0.2279147491455078, 0.22865306091308593, 0.22844825744628905, 0.22864588928222657, 0.22837759399414062, 0.2290636749267578, 0.2289213409423828, 0.22904115295410157, 0.23109939575195312, 0.22880665588378907, 0.22881996154785156, 0.2284707794189453, 0.229232666015625, 0.22840829467773438, 0.2283520050048828, 0.23005081176757813, 0.22963821411132812, 0.22803654479980467, 0.22828851318359375, 0.22792909240722656, 0.22795878601074218, 0.22777548217773438, 0.22785331726074218, 0.22797004699707032, 0.47824075317382814, 0.22801408386230468, 0.22845132446289063, 0.22803660583496094, 0.22865408325195313, 0.22791679382324218, 0.22993101501464844, 0.22821376037597657, 0.22777548217773438, 0.2276874237060547, 0.22872781372070314, 0.22810418701171875, 0.22863462829589845, 0.2276290588378906, 0.22807244873046875, 0.2276546630859375, 0.22767514038085937, 0.22787481689453126, 0.22811648559570313, 0.22791270446777342, 0.22781336975097657, 0.22764236450195313, 0.2276822967529297, 0.22771916198730469, 0.22807347106933593, 0.22767718505859375, 0.22773248291015624, 0.2274396209716797, 0.2276177978515625, 0.2276126708984375, 0.22771405029296876, 0.2276884460449219, 0.22757273864746094, 0.22881689453125, 0.22877593994140624, 0.22816461181640624, 0.22797311401367187, 0.22859878540039064, 0.22825677490234375, 0.22843699645996093, 0.22875750732421876, 0.22935040283203126, 0.22820352172851563, 0.22832333374023436, 0.23120999145507812, 0.22912101745605468, 0.22907904052734376, 0.22970060729980468, 0.22873907470703125, 0.22823014831542968, 0.22959414672851564, 0.22853219604492186, 0.22784101867675782, 0.22805708312988282, 0.22899507141113282, 0.2288885803222656, 0.227852294921875, 0.2276669464111328, 0.2279219207763672, 0.22771916198730469, 0.22784921264648436, 0.2281492462158203, 0.22833561706542968, 0.47811685180664065, 0.22757478332519532, 0.2276433868408203, 0.227662841796875, 0.22771510314941407, 0.22770684814453124, 0.22767001342773438, 0.2279884796142578, 0.22770176696777344, 0.2278707275390625, 0.22784512329101564, 0.22757171630859374, 0.22767922973632812, 0.22853836059570312, 0.22817181396484376, 0.22842568969726562, 0.22801408386230468, 0.2303057861328125, 0.22843084716796874, 0.22846669006347656, 0.22829670715332032, 0.227631103515625, 0.22766796875, 0.22799462890625, 0.22806732177734376, 0.22791270446777342, 0.22781747436523436, 0.22777548217773438, 0.22781234741210937, 0.22849331665039063, 0.22844009399414061, 0.22815536499023437, 0.22805509948730468, 0.22854751586914063, 0.22936679077148436, 0.2276741180419922, 0.22772940063476563, 0.22759014892578125, 0.22804173278808593, 0.22774169921875, 0.22859373474121095, 0.22844819641113281, 0.22795266723632812, 0.22765052795410157, 0.22817893981933593, 0.22788096618652343, 0.2275768280029297, 0.227915771484375, 0.22789631652832032, 0.22774783325195314, 0.22783692932128907, 0.22787890625, 0.22841958618164063, 0.22780006408691406, 0.22799154663085938, 0.2284862060546875, 0.22839187622070312, 0.23194111633300782, 0.22911180114746094, 0.22951731872558595, 0.22850355529785157, 0.22849740600585938, 0.22812570190429687, 0.47752294921875, 0.2276556854248047, 0.22764544677734375, 0.22774169921875, 0.22828953552246095, 0.22794444274902342, 0.22778880310058594, 0.2285332489013672, 0.22908927917480468, 0.2287073211669922, 0.2291087341308594, 0.22870118713378906, 0.22815948486328125, 0.22893466186523437, 0.2281328582763672, 0.22798233032226561, 0.22857215881347656, 0.22788812255859375, 0.2276177978515625, 0.22820658874511718, 0.22790963745117188, 0.2278912353515625, 0.22779286193847656, 0.22789631652832032, 0.2286622772216797, 0.2280079345703125, 0.22793728637695312, 0.22807142639160155, 0.22804888916015625, 0.2280335388183594, 0.22812159729003906, 0.2278830108642578, 0.2276884460449219, 0.22792807006835938, 0.22778675842285157, 0.22810009765625, 0.22793215942382813, 0.22936473083496095, 0.23167283630371094, 0.22877081298828125, 0.22863258361816408, 0.22877593994140624, 0.22783282470703126, 0.2277181396484375, 0.22791270446777342, 0.22783183288574219, 0.22780720520019532, 0.22802841186523437, 0.2279536590576172, 0.2279619140625, 0.22795872497558595, 0.22792909240722656, 0.22810009765625, 0.22938214111328126, 0.22828031921386718, 0.228242431640625, 0.2287636413574219, 0.22942311096191406, 0.22908114624023437, 0.22857618713378905, 0.22855372619628905, 0.22922035217285155, 0.2285117492675781, 0.4789893188476563, 0.22813900756835936, 0.22828134155273438, 0.22873292541503906, 0.2284707794189453, 0.22896640014648437, 0.22858956909179687, 0.2285096893310547, 0.22846669006347656, 0.22824858093261718, 0.2286878662109375, 0.2290083770751953, 0.22912716674804687, 0.22828851318359375, 0.22853836059570312, 0.228384765625, 0.22901145935058595, 0.2284523468017578, 0.22929306030273439, 0.22884352111816406, 0.2288916473388672, 0.22799874877929688, 0.22763005065917968, 0.2294476776123047, 0.22789324951171874, 0.22849740600585938, 0.2280437774658203, 0.22870835876464843, 0.228890625, 0.228600830078125, 0.22893466186523437, 0.22901248168945312, 0.22887936401367187, 0.22905445861816406, 0.2291025848388672, 0.22861824035644532, 0.2289040069580078, 0.22949267578125, 0.22833255004882813, 0.22794650268554686, 0.2277928924560547, 0.2285117492675781, 0.2278656005859375, 0.22826905822753907, 0.22805914306640626, 0.22932992553710937, 0.2291025848388672, 0.2288046112060547, 0.22843597412109376, 0.22823526000976563, 0.2282977294921875, 0.22911488342285155, 0.22893772888183594, 0.2283100128173828, 0.22832127380371095, 0.22796287536621093, 0.22796389770507813, 0.22786151123046874, 0.22802024841308594, 0.22792703247070312, 0.2290247344970703, 0.2290872344970703, 0.2293729248046875, 0.48112127685546874, 0.22774476623535156, 0.22767205810546876, 0.22760140991210936, 0.22770790100097657, 0.228706298828125, 0.22839602661132813, 0.22828953552246095, 0.22751437377929687, 0.22827008056640624, 0.22855577087402343, 0.22760858154296876, 0.22853733825683595, 0.2290882568359375, 0.22853427124023437, 0.22787174987792969, 0.22959922790527343, 0.22843597412109376, 0.22793830871582033, 0.22853427124023437, 0.22780621337890625, 0.22782566833496093, 0.2280755157470703, 0.22808781433105468, 0.22815948486328125, 0.22787686157226564, 0.22864691162109374, 0.2278481903076172, 0.22766387939453125, 0.22787174987792969, 0.22767514038085937, 0.22795161437988282, 0.22794650268554686, 0.22880563354492187, 0.22775808715820312, 0.22767514038085937, 0.22767718505859375, 0.2277734375, 0.22772122192382813, 0.22761984252929687, 0.2289653778076172, 0.22932582092285156, 0.2285506591796875, 0.22796083068847656, 0.2276986846923828, 0.22787583923339844, 0.2282608642578125, 0.2280263671875, 0.22794137573242187, 0.2278461456298828, 0.22779391479492186, 0.22773248291015624, 0.2276259765625, 0.22793215942382813, 0.22887014770507813, 0.2283756103515625, 0.22759724426269531, 0.22863462829589845, 0.22802841186523437, 0.228279296875, 0.22952755737304686, 0.22824755859375, 0.2283223114013672, 0.4810905456542969, 0.22825372314453124, 0.22767202758789062, 0.22804582214355468, 0.22862745666503906, 0.227810302734375, 0.22774989318847655, 0.22795980834960938, 0.22772940063476563, 0.22866943359375, 0.22906982421875, 0.2284390411376953, 0.22901657104492187, 0.22875852966308594, 0.22855270385742188, 0.22873805236816405, 0.22831513977050782, 0.22964530944824219, 0.2282782745361328, 0.228890625, 0.22893772888183594, 0.22941389465332032, 0.22863871765136717, 0.22856192016601562, 0.2286868438720703, 0.22785536193847655, 0.22773554992675782, 0.22780723571777345, 0.22774887084960938, 0.22783795166015625, 0.2286028747558594, 0.2292316131591797, 0.22855474853515625, 0.2284277801513672, 0.22815129089355468, 0.22762086486816407, 0.22760345458984374, 0.2294599609375, 0.228251708984375, 0.22796998596191406, 0.22916300964355468, 0.22799667358398437, 0.22786355590820312, 0.2276986846923828, 0.22783282470703126, 0.22780621337890625, 0.22774681091308593, 0.22811955261230468, 0.22787890625, 0.22774374389648439, 0.2275000305175781, 0.22767001342773438, 0.2276884460449219, 0.22773554992675782, 0.2278778839111328, 0.22794650268554686, 0.22774476623535156, 0.22773452758789062, 0.22781234741210937, 0.22782975769042968, 0.22782054138183594, 0.22786151123046874, 0.22886604309082031, 0.4816783447265625, 0.22901145935058595, 0.22866943359375, 0.2290145263671875, 0.22938316345214843, 0.22859571838378906, 0.22785023498535156, 0.22841139221191406, 0.22864895629882812, 0.2287615966796875, 0.22899302673339844, 0.2279720916748047, 0.2287073211669922, 0.22850355529785157, 0.22853631591796875, 0.22866021728515626, 0.22911077880859376, 0.22943026733398436, 0.22803558349609376, 0.22808781433105468, 0.2292725830078125, 0.22914457702636717, 0.22839295959472655, 0.22797314453125, 0.22791267395019532, 0.22795161437988282, 0.22988902282714843, 0.2280447998046875, 0.22778163146972658, 0.2276444091796875, 0.2277928924560547, 0.2285015106201172, 0.22870835876464843, 0.22830181884765624, 0.22769766235351563, 0.22775193786621092, 0.22775091552734375, 0.22768333435058594, 0.22769664001464843, 0.22772633361816405, 0.22788607788085938, 0.22782054138183594, 0.22759837341308595, 0.22872572326660157, 0.2291568603515625, 0.22867762756347657, 0.228094970703125, 0.2280990753173828, 0.22783795166015625, 0.2277724151611328, 0.22763827514648438, 0.2278164520263672, 0.22774681091308593, 0.2276433868408203, 0.2277181396484375, 0.22805708312988282, 0.2277232666015625, 0.228068359375, 0.2278154296875, 0.2276864013671875, 0.22811546325683593, 0.22927462768554688, 0.2289459228515625]",tokens/s,4.311005041891343,,,main,False,False,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-66b,facebook/opt-66b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 162.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,internlm/internlm2-20b,internlm/internlm2-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,4144.504832,15760.621568,0.0,15114.174464,14045.205504,s,10,15.963228881835938,1.5963228881835938,0.0014829914244840832,1.5958744506835938,1.5984191040039064,1.5985000183105469,1.5985647497558595,"[1.595896728515625, 1.598401123046875, 1.5945810546875, 1.5952257080078125, 1.5950601806640625, 1.5945120849609375, 1.5958521728515624, 1.5975928955078125, 1.5975260009765626, 1.5985809326171876]",tokens/s,160.36855820021125,kWh,1.8825581471125285e-05,1.0316417645935872e-05,8.981001629239093e-05,0.00011895201540945208,tokens/kWh,2152128.310889114,MB,4144.504832,15760.621568,0.0,15114.174464,14169.857024,s,10,928.8058671874999,92.88058671875,0.004527086516632578,92.879125,92.88628203124999,92.886953515625,92.887490703125,"[92.8758671875, 92.88065625, 92.887625, 92.8848203125, 92.8846328125, 92.875515625, 92.87759375, 92.8768203125, 92.876203125, 92.8861328125]",tokens/s,0.6782902889144009,kWh,0.0010964920105205644,0.0006009745227833628,0.005288325397323612,0.006985791930627541,tokens/kWh,9018.304671198623,,s,629,941.6992608642577,1.4971371396888042,0.1896008283998211,1.4742733154296874,1.4748571533203125,1.4750364013671875,3.06951080078125,"[1.473944580078125, 1.47329638671875, 1.4738431396484375, 1.47358203125, 1.47386572265625, 1.4748426513671875, 1.47376123046875, 1.4739620361328125, 1.47388720703125, 1.47437158203125, 1.474060302734375, 1.4738883056640626, 1.4746326904296876, 1.473596435546875, 1.4739219970703126, 1.474044921875, 1.474198486328125, 1.4741043701171874, 1.4745589599609374, 1.4744586181640624, 1.4741165771484375, 1.4748856201171876, 1.474040771484375, 1.473933349609375, 1.474250732421875, 1.4747626953125, 1.4741483154296875, 1.4741063232421876, 1.47391796875, 1.47371826171875, 1.4743121337890626, 1.4736240234375, 1.474345947265625, 1.4741422119140626, 1.47420263671875, 1.473943603515625, 1.474134033203125, 1.474208740234375, 1.473671142578125, 1.474809814453125, 1.473850341796875, 1.47422314453125, 1.4740899658203126, 1.4746153564453126, 1.474150390625, 1.474164794921875, 1.474740234375, 1.4742220458984374, 1.474397216796875, 1.47470947265625, 1.4741094970703126, 1.474333740234375, 1.4749337158203124, 1.4750013427734374, 1.4743634033203126, 1.47471875, 1.4743265380859376, 1.4740859375, 1.4744422607421874, 1.4750106201171875, 1.4745364990234375, 1.4744473876953126, 3.072501708984375, 1.4742149658203125, 1.47382275390625, 1.4743377685546875, 1.4739732666015626, 1.4742138671875, 1.473607666015625, 1.473517578125, 1.4735635986328126, 1.4740213623046876, 1.47407568359375, 1.474303955078125, 1.474218017578125, 1.4742906494140624, 1.4739681396484374, 1.4741094970703126, 1.47432958984375, 1.47487744140625, 1.474449462890625, 1.474229248046875, 1.474135009765625, 1.474888671875, 1.4737879638671876, 1.4744893798828125, 1.4746142578125, 1.474745361328125, 1.4743009033203125, 1.4740203857421874, 1.4740869140625, 1.4739189453125, 1.474797607421875, 1.4739200439453124, 1.474556884765625, 1.4741739501953126, 1.4747054443359375, 1.4741309814453125, 1.4740633544921875, 1.4746910400390625, 1.4741361083984375, 1.4743233642578124, 1.474419677734375, 1.4740213623046876, 1.47390673828125, 1.4743306884765626, 1.4746715087890625, 1.474198486328125, 1.474620361328125, 1.47437255859375, 1.4746910400390625, 1.474514892578125, 1.4745426025390624, 1.47401220703125, 1.4740531005859374, 1.47437158203125, 1.475441650390625, 1.4746439208984374, 1.474841552734375, 1.47451904296875, 1.474598876953125, 1.474634765625, 1.4747381591796875, 1.4739384765625, 1.473933349609375, 3.0693232421875, 1.473892333984375, 1.474171875, 1.4738431396484375, 1.474007080078125, 1.4742384033203124, 1.47473095703125, 1.4749234619140625, 1.473754150390625, 1.4738544921875, 1.474039794921875, 1.47367529296875, 1.47413916015625, 1.47485693359375, 1.4745743408203125, 1.474783203125, 1.475303466796875, 1.474924560546875, 1.4741279296875, 1.4751273193359375, 1.4750792236328125, 1.474249755859375, 1.4747965087890624, 1.474628662109375, 1.4746798095703124, 1.47464599609375, 1.475072021484375, 1.4749490966796874, 1.474650146484375, 1.474839599609375, 1.4745999755859376, 1.47433984375, 1.474587646484375, 1.4739671630859374, 1.47449853515625, 1.47369580078125, 1.4738216552734376, 1.4743746337890624, 1.474156494140625, 1.47430712890625, 1.4742518310546875, 1.474651123046875, 1.4738739013671875, 1.474051025390625, 1.4744114990234376, 1.474198486328125, 1.474423828125, 1.4742803955078125, 1.4740162353515625, 1.473574951171875, 1.4743848876953125, 1.4738145751953124, 1.47405419921875, 1.4739793701171875, 1.475199951171875, 1.4745067138671875, 1.4744473876953126, 1.4743388671875, 1.474714599609375, 1.4745006103515625, 1.4744422607421874, 1.4752808837890625, 1.474572265625, 3.06950341796875, 1.4739578857421876, 1.4735145263671876, 1.4739886474609376, 1.4746470947265624, 1.4735728759765625, 1.4737674560546874, 1.4742681884765625, 1.4737213134765625, 1.47318994140625, 1.47401318359375, 1.47386474609375, 1.47394970703125, 1.4739599609375, 1.4740623779296875, 1.474145263671875, 1.4742425537109376, 1.4741483154296875, 1.4742333984375, 1.4738831787109374, 1.474155517578125, 1.4743111572265626, 1.474144287109375, 1.4746112060546874, 1.4750372314453124, 1.4743541259765625, 1.4747115478515624, 1.4738729248046876, 1.474272216796875, 1.47419140625, 1.4745753173828124, 1.4750064697265626, 1.47466650390625, 1.4748436279296875, 1.474193359375, 1.4741329345703125, 1.4744525146484375, 1.4744913330078124, 1.4748323974609374, 1.4744207763671875, 1.4753935546875, 1.4751805419921875, 1.474482177734375, 1.4740582275390626, 1.4748548583984376, 1.474802734375, 1.4744371337890625, 1.4744453125, 1.4744443359375, 1.4745057373046875, 1.4743941650390624, 1.4749481201171875, 1.47416162109375, 1.47437255859375, 1.4748753662109375, 1.474218994140625, 1.47468798828125, 1.4740306396484375, 1.4754232177734374, 1.474871337890625, 1.4750535888671874, 1.474193359375, 1.47433984375, 3.0700166015625, 1.474471923828125, 1.4742957763671876, 1.474418701171875, 1.4738094482421875, 1.473839111328125, 1.4741832275390625, 1.47508935546875, 1.474249755859375, 1.474166748046875, 1.4742117919921875, 1.47376953125, 1.474102294921875, 1.4737879638671876, 1.4748671875, 1.4747493896484376, 1.474883544921875, 1.4746419677734375, 1.4745159912109376, 1.4746552734375, 1.474293701171875, 1.4748323974609374, 1.4744105224609374, 1.474787353515625, 1.47443408203125, 1.4740316162109375, 1.4738883056640626, 1.47376953125, 1.47512939453125, 1.47419140625, 1.4745528564453125, 1.4741678466796875, 1.47439208984375, 1.4739793701171875, 1.4745281982421874, 1.47403466796875, 1.4748272705078125, 1.474185302734375, 1.4739229736328125, 1.4739056396484376, 1.473829833984375, 1.474093017578125, 1.4740869140625, 1.474112548828125, 1.4749757080078125, 1.4745753173828124, 1.4741094970703126, 1.4738267822265625, 1.4748529052734376, 1.4741513671875, 1.473798095703125, 1.47447607421875, 1.474387939453125, 1.4741156005859375, 1.4746163330078126, 1.474620361328125, 1.4744791259765626, 1.4744678955078125, 1.4747166748046876, 1.4750802001953125, 1.4745999755859376, 1.4739814453125, 1.4747626953125, 3.069513671875, 1.47412890625, 1.4743746337890624, 1.4740838623046875, 1.4746378173828125, 1.4743603515625, 1.4743265380859376, 1.4740521240234374, 1.474298828125, 1.474333740234375, 1.47361279296875, 1.4742005615234375, 1.4735267333984374, 1.473796142578125, 1.4742364501953125, 1.47390869140625, 1.473765380859375, 1.4739835205078125, 1.4745712890625, 1.4740101318359375, 1.4739835205078125, 1.4743961181640626, 1.474408447265625, 1.4742752685546876, 1.4744114990234376, 1.4744586181640624, 1.4740623779296875, 1.4737520751953126, 1.47413916015625, 1.4744801025390626, 1.4740203857421874, 1.474050048828125, 1.47420361328125, 1.4737581787109375, 1.4745753173828124, 1.474017333984375, 1.474734130859375, 1.47439208984375, 1.4745015869140625, 1.4748160400390624, 1.474017333984375, 1.473838134765625, 1.4741207275390624, 1.473882080078125, 1.4738841552734374, 1.47435107421875, 1.4739732666015626, 1.4739046630859376, 1.4738043212890626, 1.4740582275390626, 1.4742333984375, 1.4741370849609374, 1.4745313720703126, 1.47392822265625, 1.4746859130859375, 1.4744012451171875, 1.4736475830078124, 1.4745435791015624, 1.4739937744140625, 1.4749501953125, 1.4744842529296875, 1.474662353515625, 1.474905029296875, 3.070246826171875, 1.4750576171875, 1.4738739013671875, 1.4734581298828124, 1.4737919921875, 1.474293701171875, 1.4743428955078124, 1.47382275390625, 1.47458154296875, 1.47357080078125, 1.474060302734375, 1.4740726318359374, 1.4753177490234375, 1.473629150390625, 1.4743223876953124, 1.4743489990234375, 1.473955810546875, 1.4739876708984374, 1.4747186279296876, 1.4738134765625, 1.4743214111328125, 1.4742047119140624, 1.4742978515625, 1.4737418212890625, 1.4734571533203125, 1.473523681640625, 1.473606689453125, 1.473850341796875, 1.4743746337890624, 1.4742579345703124, 1.4741832275390625, 1.474040771484375, 1.4740684814453124, 1.4739732666015626, 1.4740592041015625, 1.474809814453125, 1.4747279052734374, 1.4743153076171875, 1.4744268798828124, 1.474154541015625, 1.4741319580078125, 1.473996826171875, 1.4746644287109374, 1.4744012451171875, 1.4747647705078124, 1.474788330078125, 1.4744166259765625, 1.474029541015625, 1.4741083984375, 1.4744698486328125, 1.47449853515625, 1.473933349609375, 1.475099609375, 1.4742476806640625, 1.474344970703125, 1.47441455078125, 1.47498388671875, 1.4740675048828125, 1.4742537841796874, 1.4744873046875, 1.474292724609375, 1.47422509765625, 1.4751826171875, 3.072003173828125, 1.47430908203125, 1.4742056884765624, 1.474298828125, 1.4740029296875, 1.47441357421875, 1.4743634033203126, 1.473666015625, 1.4739056396484376, 1.4738585205078125, 1.4737633056640624, 1.47390771484375, 1.474503662109375, 1.4742467041015626, 1.47406640625, 1.4739844970703124, 1.474124755859375, 1.4739844970703124, 1.47361279296875, 1.474107421875, 1.473881103515625, 1.47378076171875, 1.474008056640625, 1.4745169677734375, 1.474460693359375, 1.474466796875, 1.4740469970703125, 1.4737305908203124, 1.47407666015625, 1.4743223876953124, 1.4743223876953124, 1.4740029296875, 1.4749521484375, 1.474461669921875, 1.4740694580078124, 1.4741329345703125, 1.47427734375, 1.474609130859375, 1.4740633544921875, 1.474460693359375, 1.474523193359375, 1.4746634521484374, 1.4743521728515625, 1.4738759765625, 1.4742733154296874, 1.4747269287109375, 1.474681884765625, 1.474302001953125, 1.4744627685546876, 1.4744678955078125, 1.4744381103515625, 1.474345947265625, 1.4741944580078126, 1.4742548828125, 1.4745078125, 1.4744801025390626, 1.47441455078125, 1.47418115234375, 1.4743223876953124, 1.4743634033203126, 1.4744627685546876, 1.4740469970703125, 1.4741370849609374, 3.0712216796875, 1.4734315185546876, 1.4745855712890625, 1.47361279296875, 1.473840087890625, 1.4741380615234374, 1.4739364013671874, 1.4736414794921875, 1.473987548828125, 1.4741053466796874, 1.4739149169921875, 1.47364453125, 1.4739620361328125, 1.47382373046875, 1.473976318359375, 1.474334716796875, 1.4739261474609375, 1.473628173828125, 1.4737264404296875, 1.4739302978515625, 1.4734212646484375, 1.4736036376953126, 1.4741319580078125, 1.47426513671875, 1.474302978515625, 1.4739517822265624, 1.4746705322265625, 1.47380126953125, 1.4740101318359375, 1.4745528564453125, 1.4744248046875, 1.47426513671875, 1.4744791259765626, 1.4744033203125, 1.4740244140625, 1.474103271484375, 1.4746705322265625, 1.4742518310546875, 1.474093017578125, 1.474193359375, 1.4737838134765624, 1.473976318359375, 1.474071533203125, 1.47503515625, 1.4746429443359375, 1.4743326416015625, 1.4748948974609375, 1.473871826171875, 1.4741412353515626, 1.474756591796875, 1.474630615234375, 1.474577392578125, 1.474472900390625, 1.474883544921875, 1.4748580322265625, 1.47447705078125, 1.4747811279296874, 1.47460400390625, 1.4745589599609374, 1.4746265869140625, 1.4749808349609375, 1.4749000244140624, 1.474957275390625, 3.07293701171875, 1.4741534423828124, 1.473850341796875, 1.474186279296875, 1.4741903076171874, 1.47452001953125, 1.4740736083984376, 1.474038818359375, 1.474935791015625, 1.4743348388671875, 1.473924072265625, 1.4737950439453125, 1.474008056640625, 1.4743111572265626, 1.474093017578125, 1.4743223876953124, 1.4741851806640625, 1.4741422119140626, 1.4739793701171875, 1.474102294921875, 1.4738380126953126, 1.4738585205078125, 1.4744945068359374, 1.4743634033203126, 1.474292724609375, 1.4741781005859376, 1.4750125732421875, 1.4740316162109375, 1.4740899658203126, 1.473870849609375, 1.47460302734375, 1.4747423095703125, 1.4751129150390625, 1.4748599853515625, 1.474618408203125, 1.475168212890625, 1.4755379638671875, 1.4745087890625, 1.474472900390625, 1.474514892578125, 1.474840576171875, 1.47443603515625, 1.4743101806640626, 1.4746552734375, 1.4748037109375, 1.4745220947265625, 1.4747740478515625, 1.4746317138671876, 1.4742313232421875, 1.4737684326171876, 1.4743951416015626, 1.474081787109375, 1.4740859375, 1.47426611328125, 1.474807861328125, 1.47405615234375, 1.4741063232421876, 1.474302978515625, 1.4747801513671874, 1.4749112548828125, 1.4753310546875, 1.4741790771484375, 1.4742220458984374]",tokens/s,0.6679414821062154,,,,,,,, +4bit-awq-gemm-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-6.7b,EleutherAI/pythia-6.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,1368.223744,6256.328704,0.0,5609.8816,5292.617728,s,10,5.657606628417969,0.5657606628417968,0.0009498936514595362,0.5654808349609375,0.5660143920898437,0.5672829833984374,0.5682978564453125,"[0.5685515747070312, 0.5654931030273438, 0.565177734375, 0.5654685668945313, 0.5652743530273437, 0.5651702880859375, 0.5656128540039063, 0.5657227172851562, 0.5657324829101562, 0.5654029541015625]",tokens/s,452.48815764977473,kWh,6.682694372203615e-06,3.660675940348786e-06,3.1253065125888854e-05,4.159643543844126e-05,tokens/kWh,6154373.50103846,MB,1368.551424,6256.328704,0.0,5609.8816,5503.949312,s,10,330.81899609375,33.081899609375,0.006904428548072782,33.079876953125,33.091131640625,33.0930072265625,33.0945076953125,"[33.07798046875, 33.078234375, 33.07119921875, 33.0948828125, 33.09071484375, 33.08729296875, 33.08378515625, 33.07924609375, 33.0805078125, 33.07515234375]",tokens/s,1.9043646448327465,kWh,0.0003903993130889204,0.00021397087349180462,0.001808246989806311,0.0024126171763870358,tokens/kWh,26112.721328770578,,s,629,335.3824949340819,0.5331995149985406,0.06720644230645373,0.5250816040039062,0.52554873046875,0.52574228515625,1.0904419140625001,"[0.5248307495117187, 0.5246986083984375, 0.5245040893554688, 0.524674072265625, 0.52523828125, 0.5248563232421875, 0.5253027954101562, 0.5246597290039062, 0.5248173828125, 0.5247999877929688, 0.5255188598632813, 0.5246453247070313, 0.5245347900390624, 0.5246586303710937, 0.525032470703125, 0.5246986083984375, 0.5246525268554687, 0.5253273315429687, 0.5247651977539063, 0.5249915161132812, 0.5252413330078125, 0.5252423706054687, 0.5247785034179687, 0.5247273559570312, 0.5247057495117188, 0.5248440551757813, 0.5247232055664063, 0.5247467651367187, 0.5255465087890625, 0.5256212768554688, 0.5258025512695312, 0.5255167846679687, 0.5248409423828125, 0.5254369506835938, 0.5253928833007813, 0.5247928466796875, 0.5255188598632813, 0.5249791870117188, 0.5249310913085937, 0.5250969848632813, 0.5247672119140625, 0.5248972778320312, 0.5248655395507813, 0.5252689819335937, 0.5248040771484375, 0.525211669921875, 0.5255167846679687, 0.525169677734375, 0.5251573486328125, 0.524832763671875, 0.5248081665039063, 0.5248256225585938, 0.5252925415039063, 0.5251204833984375, 0.525022216796875, 0.5254482421875, 0.525276123046875, 0.5254348754882813, 0.5250242309570312, 0.5254471435546875, 0.5251328125, 0.5251553344726563, 1.0913822021484374, 0.5247088623046875, 0.52463818359375, 0.5248921508789063, 0.5250447387695313, 0.52459423828125, 0.5247610473632812, 0.524621826171875, 0.5250303955078125, 0.5247098999023437, 0.5248706665039062, 0.52496484375, 0.524769287109375, 0.525318115234375, 0.5248880615234375, 0.5248450317382812, 0.5251092529296875, 0.5249443969726563, 0.5248491821289063, 0.5246842651367187, 0.524643310546875, 0.5247713012695312, 0.5255167846679687, 0.5258618774414062, 0.5249320678710937, 0.5256673583984375, 0.5253673095703125, 0.525570068359375, 0.5251849975585937, 0.5251604614257812, 0.5255874633789063, 0.5250549926757813, 0.5249617309570312, 0.5247528686523437, 0.5250518798828125, 0.525127685546875, 0.5249607543945313, 0.5249474487304687, 0.525033447265625, 0.5248983154296875, 0.5250068359375, 0.52510205078125, 0.5252761840820312, 0.5249771728515625, 0.5250181274414063, 0.5249228515625, 0.5253406982421875, 0.5251481323242188, 0.5250140380859375, 0.5252628784179687, 0.525149169921875, 0.5251195068359376, 0.5251184692382812, 0.5249095458984375, 0.5248972778320312, 0.5250303955078125, 0.525391845703125, 0.524969970703125, 0.5247293701171875, 0.5249556274414062, 0.5251163940429687, 0.5255157470703125, 0.5251348266601562, 1.090335693359375, 0.524874755859375, 0.5245787963867188, 0.524506103515625, 0.5248809204101562, 0.5247160034179688, 0.5246280517578125, 0.5246934204101562, 0.5246658325195312, 0.5245573120117187, 0.5246976928710938, 0.5247764282226562, 0.524717041015625, 0.5248132934570312, 0.52461669921875, 0.52514306640625, 0.525487060546875, 0.5247119140625, 0.524663818359375, 0.524600341796875, 0.524516357421875, 0.524632080078125, 0.5244651489257812, 0.5252577514648438, 0.5246760864257812, 0.524959716796875, 0.5246556396484375, 0.5247354736328125, 0.5250734252929687, 0.5248143310546876, 0.5247682495117187, 0.5251287231445313, 0.5246996459960938, 0.5248102416992187, 0.5250416870117187, 0.5254154052734376, 0.5251204833984375, 0.524943359375, 0.5247672119140625, 0.5249054565429687, 0.5247723388671875, 0.5249924926757813, 0.5250969848632813, 0.5248839721679688, 0.524780517578125, 0.5248870239257812, 0.5249146728515625, 0.525117431640625, 0.5246310424804688, 0.5249474487304687, 0.5247897338867188, 0.52493310546875, 0.5250109252929688, 0.5253570556640625, 0.5257471923828125, 0.5256693725585937, 0.52505908203125, 0.5255249633789062, 0.5257697143554687, 0.5254031372070312, 0.5251993408203125, 0.5257103271484375, 0.5251307373046875, 1.090680908203125, 0.525180908203125, 0.524788818359375, 0.5252505493164062, 0.5254584350585938, 0.5254379272460937, 0.52552294921875, 0.5255587768554687, 0.525601806640625, 0.5258516235351562, 0.5258157958984375, 0.5254614868164063, 0.5251604614257812, 0.5255485229492187, 0.5253007202148438, 0.5253621826171875, 0.5256417236328125, 0.525749267578125, 0.5251840209960937, 0.5252259521484375, 0.5253324584960938, 0.5253058471679688, 0.52491162109375, 0.5250846557617187, 0.525254638671875, 0.525391845703125, 0.52520654296875, 0.5254430541992188, 0.5258055419921875, 0.525138916015625, 0.5251840209960937, 0.5249403076171875, 0.525707275390625, 0.5251296997070313, 0.52486962890625, 0.5257564086914063, 0.5252894897460938, 0.5254993896484375, 0.5254983520507812, 0.5253345336914063, 0.5253857421875, 0.5251768188476562, 0.52502734375, 0.5253447875976562, 0.5251942138671875, 0.52514306640625, 0.5253836669921875, 0.5265366821289063, 0.5247119140625, 0.5249392700195312, 0.5253396606445313, 0.5246771240234375, 0.5253150634765625, 0.5247682495117187, 0.5253621826171875, 0.5255966796875, 0.5253765258789063, 0.5253641967773437, 0.5249669189453126, 0.5252526245117187, 0.5254635009765625, 0.5250498657226562, 0.5251553344726563, 1.090404296875, 0.524969970703125, 0.5259376831054687, 0.5252003784179687, 0.5258106689453125, 0.5252955932617187, 0.5250447387695313, 0.525154296875, 0.5248737182617188, 0.5253058471679688, 0.5248440551757813, 0.524874755859375, 0.5246546020507813, 0.5248102416992187, 0.5245665283203125, 0.5248348388671875, 0.5247498168945313, 0.5249761352539063, 0.5246576538085937, 0.5247181396484375, 0.5246954956054688, 0.5256632080078125, 0.5248604125976563, 0.524747802734375, 0.5248604125976563, 0.5250211791992188, 0.5254573974609374, 0.5250242309570312, 0.52548095703125, 0.5251154174804687, 0.5251461181640625, 0.5252608032226562, 0.52522607421875, 0.5251276245117188, 0.5253990478515626, 0.5256478881835938, 0.525464599609375, 0.5254779052734375, 0.5252710571289062, 0.5255720825195312, 0.5254932250976563, 0.52564892578125, 0.5251912231445313, 0.5251798095703125, 0.5252771606445312, 0.5259796752929687, 0.5252730712890625, 0.5259386596679687, 0.5251870727539063, 0.525286376953125, 0.52527001953125, 0.5250816040039062, 0.525275146484375, 0.52520654296875, 0.5252454833984375, 0.5256109619140625, 0.5253765258789063, 0.52531201171875, 0.5252474975585938, 0.5265930786132812, 0.525812744140625, 0.5261884765625, 0.5251481323242188, 1.09045654296875, 0.5245020141601563, 0.5246781616210937, 0.5249658813476562, 0.5249915161132812, 0.5245040893554688, 0.52481640625, 0.5248297119140625, 0.5247836303710938, 0.524885986328125, 0.5248921508789063, 0.52484814453125, 0.5248286743164062, 0.524632080078125, 0.5257512817382812, 0.524990478515625, 0.5250908203125, 0.5249392700195312, 0.5247129516601563, 0.5252608032226562, 0.5249863891601563, 0.5254144287109375, 0.524600341796875, 0.5249567260742187, 0.5247467041015625, 0.5249075317382812, 0.5252260131835937, 0.5248870239257812, 0.5250662231445312, 0.5250048828125, 0.5253692626953125, 0.5249249267578125, 0.524853271484375, 0.5260421142578126, 0.5258383178710937, 0.5250867309570313, 0.5253775634765625, 0.524843017578125, 0.525390869140625, 0.5253294067382812, 0.5256530151367188, 0.5254686889648438, 0.525365234375, 0.525201416015625, 0.5253119506835937, 0.5253365478515625, 0.5253560180664063, 0.5255802612304687, 0.5252782592773437, 0.5253252563476563, 0.5253990478515626, 0.52508056640625, 0.5262468872070313, 0.5258137817382813, 0.525549560546875, 0.5252843627929688, 0.5254717407226562, 0.525453369140625, 0.5256119384765625, 0.5255628662109375, 0.5259458618164062, 0.5253795776367187, 0.52560693359375, 1.0914058837890626, 0.5248624877929687, 0.5247764282226562, 0.5250928344726562, 0.5249915161132812, 0.5249924926757813, 0.524921875, 0.5255986938476562, 0.525233154296875, 0.5249976196289062, 0.525106201171875, 0.525254638671875, 0.5254993896484375, 0.52495361328125, 0.5254133911132812, 0.5256365966796875, 0.5252393188476563, 0.5251287231445313, 0.5250734252929687, 0.5250263061523438, 0.5252188110351562, 0.5251512451171875, 0.5254318237304687, 0.5252034301757813, 0.5248993530273437, 0.5250560302734375, 0.5257349243164062, 0.5247313842773438, 0.5247518920898437, 0.5246392211914063, 0.5248256225585938, 0.5250816040039062, 0.5247979736328126, 0.5251195068359376, 0.5250089111328125, 0.525053955078125, 0.5249034423828125, 0.5249392700195312, 0.525000732421875, 0.5251604614257812, 0.5249515380859375, 0.5250344848632813, 0.5250201416015625, 0.524843017578125, 0.5250570068359375, 0.525391845703125, 0.5252781982421875, 0.5252321166992188, 0.5249515380859375, 0.5251287231445313, 0.5253140258789063, 0.524959716796875, 0.5254113159179687, 0.5252976684570313, 0.5251635131835938, 0.525433837890625, 0.5251604614257812, 0.5252894897460938, 0.5252474975585938, 0.5253939208984375, 0.5253867797851562, 0.5253079223632813, 0.525212646484375, 1.0918123779296875, 0.5249392700195312, 0.5248511962890625, 0.5250714111328125, 0.525445068359375, 0.5252464599609376, 0.5249423217773438, 0.5251881103515625, 0.5252290649414062, 0.5253109741210937, 0.5252495727539063, 0.524802001953125, 0.52481640625, 0.5247365112304687, 0.5250303955078125, 0.5251235961914062, 0.5250078735351562, 0.5248880615234375, 0.5249197998046875, 0.5250938720703126, 0.525497314453125, 0.5250078735351562, 0.5247528686523437, 0.5247733764648438, 0.524821533203125, 0.524938232421875, 0.5251942138671875, 0.5249832763671874, 0.5250949096679688, 0.5249238891601562, 0.524906494140625, 0.5249658813476562, 0.5247979736328126, 0.5248522338867188, 0.52481640625, 0.52491162109375, 0.5247928466796875, 0.5249320678710937, 0.5248573608398438, 0.5247938842773437, 0.5249791870117188, 0.52478466796875, 0.5247897338867188, 0.5247600708007812, 0.52487890625, 0.5249884033203125, 0.5252925415039063, 0.5250416870117187, 0.5249464111328125, 0.5251000366210937, 0.5250816040039062, 0.5249525756835938, 0.5252413940429688, 0.5255003662109375, 0.5252474975585938, 0.5254031372070312, 0.5251502075195312, 0.5256693725585937, 0.5256007690429687, 0.5253478393554688, 0.5253816528320312, 0.525201416015625, 0.5253990478515626, 1.0916168212890625, 0.52502734375, 0.5249310913085937, 0.524822509765625, 0.5248081665039063, 0.5246361694335937, 0.5247744140625, 0.5250693359375, 0.5248952026367187, 0.5249024047851563, 0.5249915161132812, 0.5248173828125, 0.525412353515625, 0.5252208862304687, 0.5250160522460937, 0.524969970703125, 0.525085693359375, 0.5248706665039062, 0.5251307373046875, 0.5251522827148437, 0.5256038208007813, 0.5248409423828125, 0.5249832763671874, 0.5250211791992188, 0.524906494140625, 0.524864501953125, 0.52510107421875, 0.5247754516601563, 0.5251942138671875, 0.5250908203125, 0.5250857543945312, 0.5254880981445312, 0.5257154541015625, 0.5253683471679688, 0.5251204833984375, 0.5250826416015625, 0.525117431640625, 0.525169677734375, 0.5254266967773438, 0.525581298828125, 0.5249771728515625, 0.5250396118164062, 0.5251522827148437, 0.5250263061523438, 0.5251163940429687, 0.5250775146484375, 0.5251696166992188, 0.5251348266601562, 0.5248726806640625, 0.5248040771484375, 0.525322265625, 0.5251829833984375, 0.5251051635742188, 0.5248706665039062, 0.5250303955078125, 0.5250109252929688, 0.5251963500976562, 0.5250313720703125, 0.5255403442382812, 0.5248737182617188, 0.5250242309570312, 0.5249843139648438, 0.5250836181640625, 1.0909224853515624, 0.5246965942382813, 0.5248829345703125, 0.5249362182617188, 0.5247160034179688, 0.5250303955078125, 0.5250089111328125, 0.5251287231445313, 0.5249238891601562, 0.5249658813476562, 0.5249515380859375, 0.5246607055664062, 0.52463720703125, 0.525317138671875, 0.5247047729492188, 0.5247109375, 0.5245911254882812, 0.5245593872070312, 0.5245286254882813, 0.5245368041992188, 0.5244630737304687, 0.5245419311523437, 0.5247344360351562, 0.5246781616210937, 0.5250242309570312, 0.525285400390625, 0.5248081665039063, 0.5248297119140625, 0.5248153686523438, 0.5249310913085937, 0.5255239868164062, 0.5253621826171875, 0.5255106811523438, 0.5247969360351562, 0.52493115234375, 0.5251900634765625, 0.525365234375, 0.52534375, 0.52518603515625, 0.5249095458984375, 0.524705810546875, 0.5250477905273437, 0.5248071899414063, 0.5250426635742188, 0.5253519287109375, 0.52510205078125, 0.5249894409179687, 0.5251450805664063, 0.52478466796875, 0.5255065307617187, 0.5247191162109375, 0.5256325073242187, 0.5251266479492187, 0.5251030883789063, 0.5249208374023437, 0.5255485229492187, 0.525053955078125, 0.5254256591796875, 0.525159423828125, 0.525106201171875, 0.5251512451171875, 0.5250416870117187, 0.5253683471679688]",tokens/s,1.8754705731544732,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-1.4b,EleutherAI/pythia-1.4b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 1031, in forward + outputs = self.gpt_neox( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 922, in forward + outputs = layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 687, in forward + attention_layer_outputs = self.attention( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 327, in forward + qkv = self.query_key_value(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Salesforce/codegen-6B-nl,Salesforce/codegen-6B-nl,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1469, in _autoset_attn_implementation + cls._check_and_enable_flash_attn_2( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1560, in _check_and_enable_flash_attn_2 + raise ValueError( +ValueError: CodeGenForCausalLM does not support Flash Attention 2.0 yet. Please request to add support where the model is hosted, on its model hub page: https://huggingface.co//tmp/tmpodtbpat2/no_weights_model/discussions/new or in the Transformers GitHub repo: https://github.com/huggingface/transformers/issues/new + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-70m,EleutherAI/pythia-70m,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 1031, in forward + outputs = self.gpt_neox( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 922, in forward + outputs = layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 687, in forward + attention_layer_outputs = self.attention( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 327, in forward + qkv = self.query_key_value(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-110B,Qwen/Qwen1.5-110B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 32.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciCoder-1b,Deci/DeciCoder-1b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 68, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1164, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 968, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 713, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) +TypeError: DeciCoderAttention.forward() got an unexpected keyword argument 'cache_position' + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/recurrentgemma-2b,google/recurrentgemma-2b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/recurrentgemma-2b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aaa7b-26f026ea38819f413564c63e;fe8b807e-5250-4567-a8b9-8ed4e2fe38a5) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/recurrentgemma-2b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/recurrentgemma-2b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,internlm/internlm-20b,internlm/internlm-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm-20b/80729bcf52fbc4553d965926b27304ac5e156d98/modeling_internlm.py"", line 976, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm-20b/80729bcf52fbc4553d965926b27304ac5e156d98/modeling_internlm.py"", line 866, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm-20b/80729bcf52fbc4553d965926b27304ac5e156d98/modeling_internlm.py"", line 583, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm-20b/80729bcf52fbc4553d965926b27304ac5e156d98/modeling_internlm.py"", line 411, in forward + query_states = self.q_proj(hidden_states).view(bsz, q_len, self.num_heads, self.head_dim).transpose(1, 2) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-2.7b,facebook/opt-2.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 1117, in forward + outputs = self.model.decoder( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 883, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 524, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 291, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-125m,facebook/opt-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 1117, in forward + outputs = self.model.decoder( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 883, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 524, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 291, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/gemma-7b,google/gemma-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/gemma-7b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aaa29-787e900806192c8213850fc6;4f3bbb2a-de72-43cb-a202-0733e6f310a8) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/gemma-7b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/gemma-7b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-7B,Qwen/Qwen-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 550, in from_pretrained + model_class = get_class_from_dynamic_module( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 498, in get_class_from_dynamic_module + final_module = get_cached_module_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 323, in get_cached_module_file + modules_needed = check_imports(resolved_module_file) + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 181, in check_imports + raise ImportError( +ImportError: This modeling file requires the following packages that were not found in your environment: transformers_stream_generator. Run `pip install transformers_stream_generator` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-72B,Qwen/Qwen1.5-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 96.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciLM-7B,Deci/DeciLM-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 68, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/Deci/DeciLM-7B/c3c9f4226801dc0433f32aebffe0aac68ee2f051/transformers_v4_35_2__modeling_llama.py"", line 1034, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/Deci/DeciLM-7B/c3c9f4226801dc0433f32aebffe0aac68ee2f051/modeling_decilm.py"", line 274, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/Deci/DeciLM-7B/c3c9f4226801dc0433f32aebffe0aac68ee2f051/transformers_v4_35_2__modeling_llama.py"", line 672, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/Deci/DeciLM-7B/c3c9f4226801dc0433f32aebffe0aac68ee2f051/modeling_decilm.py"", line 84, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-j-6b,EleutherAI/gpt-j-6b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 68, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gptj/modeling_gptj.py"", line 1124, in forward + transformer_outputs = self.transformer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gptj/modeling_gptj.py"", line 950, in forward + outputs = block( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gptj/modeling_gptj.py"", line 578, in forward + attn_outputs = self.attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gptj/modeling_gptj.py"", line 317, in forward + query = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-0.5B,Qwen/Qwen1.5-0.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1149, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1034, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 748, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 344, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-160m,EleutherAI/pythia-160m,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 1031, in forward + outputs = self.gpt_neox( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 922, in forward + outputs = layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 687, in forward + attention_layer_outputs = self.attention( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 327, in forward + qkv = self.query_key_value(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Salesforce/codegen-16B-nl,Salesforce/codegen-16B-nl,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1469, in _autoset_attn_implementation + cls._check_and_enable_flash_attn_2( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1560, in _check_and_enable_flash_attn_2 + raise ValueError( +ValueError: CodeGenForCausalLM does not support Flash Attention 2.0 yet. Please request to add support where the model is hosted, on its model hub page: https://huggingface.co//tmp/tmpcdzp_90u/no_weights_model/discussions/new or in the Transformers GitHub repo: https://github.com/huggingface/transformers/issues/new + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-MoE-A2.7B,Qwen/Qwen1.5-MoE-A2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3640, in from_pretrained + hf_quantizer.preprocess_model( + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 182, in preprocess_model + return self._process_model_before_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 85, in _process_model_before_weight_loading + model, has_been_replaced = replace_with_awq_linear( + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 178, in replace_with_awq_linear + _, has_been_replaced = replace_with_awq_linear( + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 178, in replace_with_awq_linear + _, has_been_replaced = replace_with_awq_linear( + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 178, in replace_with_awq_linear + _, has_been_replaced = replace_with_awq_linear( + [Previous line repeated 1 more time] + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 165, in replace_with_awq_linear + model._modules[name] = target_cls( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 46, in __init__ + assert out_features % (32 // self.w_bit) == 0 +AssertionError + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-1.3b,EleutherAI/pythia-1.3b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 1031, in forward + outputs = self.gpt_neox( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 922, in forward + outputs = layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 687, in forward + attention_layer_outputs = self.attention( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 327, in forward + qkv = self.query_key_value(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-2.7b,EleutherAI/pythia-2.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 1031, in forward + outputs = self.gpt_neox( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 922, in forward + outputs = layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 687, in forward + attention_layer_outputs = self.attention( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 327, in forward + qkv = self.query_key_value(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-14B,Qwen/Qwen1.5-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1149, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1034, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 748, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 344, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen2-beta-14B,Qwen/Qwen2-beta-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1149, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1034, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 748, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 344, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-30b,facebook/opt-30b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 1117, in forward + outputs = self.model.decoder( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 883, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 524, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 291, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-1.3B,EleutherAI/gpt-neo-1.3B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 68, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 971, in forward + transformer_outputs = self.transformer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 839, in forward + outputs = block( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 566, in forward + attn_outputs = self.attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 518, in forward + return self.attention( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 319, in forward + query = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,databricks/dbrx-base,databricks/dbrx-base,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/databricks/dbrx-base/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aa545-705ec7dc4cd12f24336f4d9d;5c1cee1f-944f-427e-a8bb-3306cbc75d56) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/databricks/dbrx-base/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like databricks/dbrx-base is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,huggyllama/llama-7b,huggyllama/llama-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1164, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 968, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 713, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 416, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neox-20b,EleutherAI/gpt-neox-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 68, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 1031, in forward + outputs = self.gpt_neox( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 922, in forward + outputs = layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 687, in forward + attention_layer_outputs = self.attention( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 327, in forward + qkv = self.query_key_value(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-4B,Qwen/Qwen1.5-4B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1149, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1034, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 748, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 344, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-1.8B,Qwen/Qwen1.5-1.8B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1149, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1034, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 748, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 344, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-14B,Qwen/Qwen-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 550, in from_pretrained + model_class = get_class_from_dynamic_module( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 498, in get_class_from_dynamic_module + final_module = get_cached_module_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 323, in get_cached_module_file + modules_needed = check_imports(resolved_module_file) + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 181, in check_imports + raise ImportError( +ImportError: This modeling file requires the following packages that were not found in your environment: transformers_stream_generator. Run `pip install transformers_stream_generator` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-7.5B,facebook/xglm-7.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1469, in _autoset_attn_implementation + cls._check_and_enable_flash_attn_2( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1560, in _check_and_enable_flash_attn_2 + raise ValueError( +ValueError: XGLMForCausalLM does not support Flash Attention 2.0 yet. Please request to add support where the model is hosted, on its model hub page: https://huggingface.co//tmp/tmpmd0d_7j8/no_weights_model/discussions/new or in the Transformers GitHub repo: https://github.com/huggingface/transformers/issues/new + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-12b,EleutherAI/pythia-12b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 1031, in forward + outputs = self.gpt_neox( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 922, in forward + outputs = layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 687, in forward + attention_layer_outputs = self.attention( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 327, in forward + qkv = self.query_key_value(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-7B,Qwen/Qwen1.5-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1149, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1034, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 748, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 344, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-34B,01-ai/Yi-34B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 68, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1164, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 968, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 713, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 416, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/gemma-2b,google/gemma-2b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/gemma-2b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aa9d2-6cea5b9e6e7329f8696f71af;249e196c-75ac-467f-9a04-33b9b349438e) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/gemma-2b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/gemma-2b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/polyglot-ko-12.8b,EleutherAI/polyglot-ko-12.8b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 68, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 1031, in forward + outputs = self.gpt_neox( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 922, in forward + outputs = layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 687, in forward + attention_layer_outputs = self.attention( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 327, in forward + qkv = self.query_key_value(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-125m,EleutherAI/gpt-neo-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 68, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 971, in forward + transformer_outputs = self.transformer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 839, in forward + outputs = block( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 566, in forward + attn_outputs = self.attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 518, in forward + return self.attention( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 319, in forward + query = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-13b,facebook/opt-13b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 1117, in forward + outputs = self.model.decoder( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 883, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 524, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 291, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/recurrentgemma-7b,google/recurrentgemma-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 404 Client Error: Not Found for url: https://huggingface.co/google/recurrentgemma-7b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1823, in _raise_on_head_call_error + raise head_call_error + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 352, in hf_raise_for_status + raise RepositoryNotFoundError(message, response) from e +huggingface_hub.utils._errors.RepositoryNotFoundError: 404 Client Error. (Request ID: Root=1-664aaace-4d4258fb7c9f944a6686853f;8c2985fb-9dcf-4258-a94c-a87ee9f4c7d4) + +Repository Not Found for url: https://huggingface.co/google/recurrentgemma-7b/resolve/main/config.json. +Please make sure you specified the correct `repo_id` and `repo_type`. +If you are trying to access a private or gated repo, make sure you are authenticated. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 422, in cached_file + raise EnvironmentError( +OSError: google/recurrentgemma-7b is not a local folder and is not a valid model identifier listed on 'https://huggingface.co/models' +If this is a private repository, make sure to pass a token having permission to this repo either by logging in with `huggingface-cli login` or by passing `token=` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-410m,EleutherAI/pythia-410m,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 1031, in forward + outputs = self.gpt_neox( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 922, in forward + outputs = layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 687, in forward + attention_layer_outputs = self.attention( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 327, in forward + qkv = self.query_key_value(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-32B,Qwen/Qwen1.5-32B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1149, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1034, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 748, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 344, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-4.5B,facebook/xglm-4.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1469, in _autoset_attn_implementation + cls._check_and_enable_flash_attn_2( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1560, in _check_and_enable_flash_attn_2 + raise ValueError( +ValueError: XGLMForCausalLM does not support Flash Attention 2.0 yet. Please request to add support where the model is hosted, on its model hub page: https://huggingface.co//tmp/tmpwm_obutw/no_weights_model/discussions/new or in the Transformers GitHub repo: https://github.com/huggingface/transformers/issues/new + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,TencentARC/Mistral_Pro_8B_v0.1,TencentARC/Mistral_Pro_8B_v0.1,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/mistral/modeling_mistral.py"", line 1139, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/mistral/modeling_mistral.py"", line 1024, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/mistral/modeling_mistral.py"", line 738, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/mistral/modeling_mistral.py"", line 344, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-564M,facebook/xglm-564M,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1469, in _autoset_attn_implementation + cls._check_and_enable_flash_attn_2( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1560, in _check_and_enable_flash_attn_2 + raise ValueError( +ValueError: XGLMForCausalLM does not support Flash Attention 2.0 yet. Please request to add support where the model is hosted, on its model hub page: https://huggingface.co//tmp/tmpe7602us0/no_weights_model/discussions/new or in the Transformers GitHub repo: https://github.com/huggingface/transformers/issues/new + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-350m,facebook/opt-350m,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 1117, in forward + outputs = self.model.decoder( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 835, in forward + inputs_embeds = self.project_in(inputs_embeds) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-6.7b,facebook/opt-6.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 1117, in forward + outputs = self.model.decoder( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 883, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 524, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 291, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-72B,Qwen/Qwen-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 550, in from_pretrained + model_class = get_class_from_dynamic_module( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 498, in get_class_from_dynamic_module + final_module = get_cached_module_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 323, in get_cached_module_file + modules_needed = check_imports(resolved_module_file) + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 181, in check_imports + raise ImportError( +ImportError: This modeling file requires the following packages that were not found in your environment: transformers_stream_generator. Run `pip install transformers_stream_generator` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-6B,01-ai/Yi-6B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 68, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1164, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 968, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 713, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 416, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen2-beta-72B,Qwen/Qwen2-beta-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 96.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-2.7B,EleutherAI/gpt-neo-2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 68, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 971, in forward + transformer_outputs = self.transformer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 839, in forward + outputs = block( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 566, in forward + attn_outputs = self.attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 518, in forward + return self.attention( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 319, in forward + query = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-66b,facebook/opt-66b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 162.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,internlm/internlm2-20b,internlm/internlm2-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm2-20b/f363ea8a116b3ea829c7a068ca24bc9d3e668083/modeling_internlm2.py"", line 1047, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm2-20b/f363ea8a116b3ea829c7a068ca24bc9d3e668083/modeling_internlm2.py"", line 932, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm2-20b/f363ea8a116b3ea829c7a068ca24bc9d3e668083/modeling_internlm2.py"", line 639, in forward + hidden_states, self_attn_weights, present_key_value = self.attention( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm2-20b/f363ea8a116b3ea829c7a068ca24bc9d3e668083/modeling_internlm2.py"", line 446, in forward + qkv_states = self.wqkv(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-6.7b,EleutherAI/pythia-6.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 1031, in forward + outputs = self.gpt_neox( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 922, in forward + outputs = layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 687, in forward + attention_layer_outputs = self.attention( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 327, in forward + qkv = self.query_key_value(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-1.4b,EleutherAI/pythia-1.4b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Salesforce/codegen-6B-nl,Salesforce/codegen-6B-nl,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1469, in _autoset_attn_implementation + cls._check_and_enable_flash_attn_2( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1560, in _check_and_enable_flash_attn_2 + raise ValueError( +ValueError: CodeGenForCausalLM does not support Flash Attention 2.0 yet. Please request to add support where the model is hosted, on its model hub page: https://huggingface.co//tmp/tmpek2o7wtm/no_weights_model/discussions/new or in the Transformers GitHub repo: https://github.com/huggingface/transformers/issues/new + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-70m,EleutherAI/pythia-70m,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-110B,Qwen/Qwen1.5-110B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 32.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciCoder-1b,Deci/DeciCoder-1b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 609, in resolve_trust_remote_code + answer = input( +EOFError: EOF when reading a line + +During handling of the above exception, another exception occurred: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 228, in load_model_with_no_weights + self.create_no_weights_model() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 210, in create_no_weights_model + meta_model = self.automodel_class.from_config(self.pretrained_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 418, in from_config + trust_remote_code = resolve_trust_remote_code( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 622, in resolve_trust_remote_code + raise ValueError( +ValueError: The repository for Deci/DeciCoder-1b contains custom code which must be executed to correctly load the model. You can inspect the repository content at https://hf.co/Deci/DeciCoder-1b. +Please pass the argument `trust_remote_code=True` to allow custom code to be run. + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/recurrentgemma-2b,google/recurrentgemma-2b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/recurrentgemma-2b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aaa89-09800c3f4c244d442c6f81a5;e15a8c85-5c56-4b39-8c50-d268f7196ce2) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/recurrentgemma-2b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/recurrentgemma-2b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,internlm/internlm-20b,internlm/internlm-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 609, in resolve_trust_remote_code + answer = input( +EOFError: EOF when reading a line + +During handling of the above exception, another exception occurred: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 228, in load_model_with_no_weights + self.create_no_weights_model() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 210, in create_no_weights_model + meta_model = self.automodel_class.from_config(self.pretrained_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 418, in from_config + trust_remote_code = resolve_trust_remote_code( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 622, in resolve_trust_remote_code + raise ValueError( +ValueError: The repository for internlm/internlm-20b contains custom code which must be executed to correctly load the model. You can inspect the repository content at https://hf.co/internlm/internlm-20b. +Please pass the argument `trust_remote_code=True` to allow custom code to be run. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-2.7b,facebook/opt-2.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-125m,facebook/opt-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/gemma-7b,google/gemma-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/gemma-7b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aaa37-6311646a5c47c7bb5c64afcd;74ae4787-636d-4952-ab78-1b5535e8a752) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/gemma-7b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/gemma-7b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-7B,Qwen/Qwen-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 609, in resolve_trust_remote_code + answer = input( +EOFError: EOF when reading a line + +During handling of the above exception, another exception occurred: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 228, in load_model_with_no_weights + self.create_no_weights_model() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 210, in create_no_weights_model + meta_model = self.automodel_class.from_config(self.pretrained_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 418, in from_config + trust_remote_code = resolve_trust_remote_code( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 622, in resolve_trust_remote_code + raise ValueError( +ValueError: The repository for Qwen/Qwen-7B contains custom code which must be executed to correctly load the model. You can inspect the repository content at https://hf.co/Qwen/Qwen-7B. +Please pass the argument `trust_remote_code=True` to allow custom code to be run. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-72B,Qwen/Qwen1.5-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 96.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciLM-7B,Deci/DeciLM-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 609, in resolve_trust_remote_code + answer = input( +EOFError: EOF when reading a line + +During handling of the above exception, another exception occurred: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 228, in load_model_with_no_weights + self.create_no_weights_model() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 210, in create_no_weights_model + meta_model = self.automodel_class.from_config(self.pretrained_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 418, in from_config + trust_remote_code = resolve_trust_remote_code( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 622, in resolve_trust_remote_code + raise ValueError( +ValueError: The repository for Deci/DeciLM-7B contains custom code which must be executed to correctly load the model. You can inspect the repository content at https://hf.co/Deci/DeciLM-7B. +Please pass the argument `trust_remote_code=True` to allow custom code to be run. + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-j-6b,EleutherAI/gpt-j-6b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-0.5B,Qwen/Qwen1.5-0.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-160m,EleutherAI/pythia-160m,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Salesforce/codegen-16B-nl,Salesforce/codegen-16B-nl,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1469, in _autoset_attn_implementation + cls._check_and_enable_flash_attn_2( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1560, in _check_and_enable_flash_attn_2 + raise ValueError( +ValueError: CodeGenForCausalLM does not support Flash Attention 2.0 yet. Please request to add support where the model is hosted, on its model hub page: https://huggingface.co//tmp/tmppfgzvkn7/no_weights_model/discussions/new or in the Transformers GitHub repo: https://github.com/huggingface/transformers/issues/new + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-MoE-A2.7B,Qwen/Qwen1.5-MoE-A2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-1.3b,EleutherAI/pythia-1.3b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-2.7b,EleutherAI/pythia-2.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-14B,Qwen/Qwen1.5-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen2-beta-14B,Qwen/Qwen2-beta-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-30b,facebook/opt-30b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-1.3B,EleutherAI/gpt-neo-1.3B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,databricks/dbrx-base,databricks/dbrx-base,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/databricks/dbrx-base/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aa553-0c59137e0a74ecdf71618267;6116998d-12de-4f25-a1ad-ace342cb13c6) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/databricks/dbrx-base/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like databricks/dbrx-base is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,huggyllama/llama-7b,huggyllama/llama-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neox-20b,EleutherAI/gpt-neox-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-4B,Qwen/Qwen1.5-4B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-1.8B,Qwen/Qwen1.5-1.8B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-14B,Qwen/Qwen-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 609, in resolve_trust_remote_code + answer = input( +EOFError: EOF when reading a line + +During handling of the above exception, another exception occurred: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 228, in load_model_with_no_weights + self.create_no_weights_model() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 210, in create_no_weights_model + meta_model = self.automodel_class.from_config(self.pretrained_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 418, in from_config + trust_remote_code = resolve_trust_remote_code( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 622, in resolve_trust_remote_code + raise ValueError( +ValueError: The repository for Qwen/Qwen-14B contains custom code which must be executed to correctly load the model. You can inspect the repository content at https://hf.co/Qwen/Qwen-14B. +Please pass the argument `trust_remote_code=True` to allow custom code to be run. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-7.5B,facebook/xglm-7.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1469, in _autoset_attn_implementation + cls._check_and_enable_flash_attn_2( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1560, in _check_and_enable_flash_attn_2 + raise ValueError( +ValueError: XGLMForCausalLM does not support Flash Attention 2.0 yet. Please request to add support where the model is hosted, on its model hub page: https://huggingface.co//tmp/tmp2q60bee7/no_weights_model/discussions/new or in the Transformers GitHub repo: https://github.com/huggingface/transformers/issues/new + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-12b,EleutherAI/pythia-12b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-7B,Qwen/Qwen1.5-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-34B,01-ai/Yi-34B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/gemma-2b,google/gemma-2b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/gemma-2b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aa9e0-7cb1c6095de40d4d74f70121;53231438-7253-4643-8b86-7c0a4575b07d) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/gemma-2b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/gemma-2b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/polyglot-ko-12.8b,EleutherAI/polyglot-ko-12.8b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-125m,EleutherAI/gpt-neo-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-13b,facebook/opt-13b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/recurrentgemma-7b,google/recurrentgemma-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 404 Client Error: Not Found for url: https://huggingface.co/google/recurrentgemma-7b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1823, in _raise_on_head_call_error + raise head_call_error + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 352, in hf_raise_for_status + raise RepositoryNotFoundError(message, response) from e +huggingface_hub.utils._errors.RepositoryNotFoundError: 404 Client Error. (Request ID: Root=1-664aaadb-0269a15f0fb061c510273b88;4bed6e2f-8ace-40f2-a799-ba33dce191eb) + +Repository Not Found for url: https://huggingface.co/google/recurrentgemma-7b/resolve/main/config.json. +Please make sure you specified the correct `repo_id` and `repo_type`. +If you are trying to access a private or gated repo, make sure you are authenticated. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 422, in cached_file + raise EnvironmentError( +OSError: google/recurrentgemma-7b is not a local folder and is not a valid model identifier listed on 'https://huggingface.co/models' +If this is a private repository, make sure to pass a token having permission to this repo either by logging in with `huggingface-cli login` or by passing `token=` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-410m,EleutherAI/pythia-410m,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-32B,Qwen/Qwen1.5-32B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-4.5B,facebook/xglm-4.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1469, in _autoset_attn_implementation + cls._check_and_enable_flash_attn_2( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1560, in _check_and_enable_flash_attn_2 + raise ValueError( +ValueError: XGLMForCausalLM does not support Flash Attention 2.0 yet. Please request to add support where the model is hosted, on its model hub page: https://huggingface.co//tmp/tmpbvzim6x7/no_weights_model/discussions/new or in the Transformers GitHub repo: https://github.com/huggingface/transformers/issues/new + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,TencentARC/Mistral_Pro_8B_v0.1,TencentARC/Mistral_Pro_8B_v0.1,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-564M,facebook/xglm-564M,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1469, in _autoset_attn_implementation + cls._check_and_enable_flash_attn_2( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1560, in _check_and_enable_flash_attn_2 + raise ValueError( +ValueError: XGLMForCausalLM does not support Flash Attention 2.0 yet. Please request to add support where the model is hosted, on its model hub page: https://huggingface.co//tmp/tmpz6s3yav_/no_weights_model/discussions/new or in the Transformers GitHub repo: https://github.com/huggingface/transformers/issues/new + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-350m,facebook/opt-350m,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-6.7b,facebook/opt-6.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-72B,Qwen/Qwen-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 609, in resolve_trust_remote_code + answer = input( +EOFError: EOF when reading a line + +During handling of the above exception, another exception occurred: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 228, in load_model_with_no_weights + self.create_no_weights_model() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 210, in create_no_weights_model + meta_model = self.automodel_class.from_config(self.pretrained_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 418, in from_config + trust_remote_code = resolve_trust_remote_code( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 622, in resolve_trust_remote_code + raise ValueError( +ValueError: The repository for Qwen/Qwen-72B contains custom code which must be executed to correctly load the model. You can inspect the repository content at https://hf.co/Qwen/Qwen-72B. +Please pass the argument `trust_remote_code=True` to allow custom code to be run. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-6B,01-ai/Yi-6B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen2-beta-72B,Qwen/Qwen2-beta-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 96.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-2.7B,EleutherAI/gpt-neo-2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-66b,facebook/opt-66b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 162.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,internlm/internlm2-20b,internlm/internlm2-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 609, in resolve_trust_remote_code + answer = input( +EOFError: EOF when reading a line + +During handling of the above exception, another exception occurred: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 228, in load_model_with_no_weights + self.create_no_weights_model() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 210, in create_no_weights_model + meta_model = self.automodel_class.from_config(self.pretrained_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 418, in from_config + trust_remote_code = resolve_trust_remote_code( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 622, in resolve_trust_remote_code + raise ValueError( +ValueError: The repository for internlm/internlm2-20b contains custom code which must be executed to correctly load the model. You can inspect the repository content at https://hf.co/internlm/internlm2-20b. +Please pass the argument `trust_remote_code=True` to allow custom code to be run. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-6.7b,EleutherAI/pythia-6.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-1.4b,EleutherAI/pythia-1.4b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Salesforce/codegen-6B-nl,Salesforce/codegen-6B-nl,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-70m,EleutherAI/pythia-70m,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-110B,Qwen/Qwen1.5-110B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 32.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciCoder-1b,Deci/DeciCoder-1b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 558, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/recurrentgemma-2b,google/recurrentgemma-2b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/recurrentgemma-2b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aaa4b-2763b8c34bf14a5178850de7;021cff04-caef-48f5-aafd-5c2a0aebe482) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/recurrentgemma-2b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/recurrentgemma-2b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,internlm/internlm-20b,internlm/internlm-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 558, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-2.7b,facebook/opt-2.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-125m,facebook/opt-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/gemma-7b,google/gemma-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/gemma-7b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aa9f7-0ecdb001106b067c7d916e76;5d387675-6b80-4a7c-b00d-fa0710b78b97) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/gemma-7b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/gemma-7b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,huggyllama/llama-13b,huggyllama/llama-13b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-7B,Qwen/Qwen-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 550, in from_pretrained + model_class = get_class_from_dynamic_module( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 498, in get_class_from_dynamic_module + final_module = get_cached_module_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 323, in get_cached_module_file + modules_needed = check_imports(resolved_module_file) + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 181, in check_imports + raise ImportError( +ImportError: This modeling file requires the following packages that were not found in your environment: transformers_stream_generator. Run `pip install transformers_stream_generator` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-72B,Qwen/Qwen1.5-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 96.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciLM-7B,Deci/DeciLM-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 558, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-j-6b,EleutherAI/gpt-j-6b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-0.5B,Qwen/Qwen1.5-0.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-160m,EleutherAI/pythia-160m,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Salesforce/codegen-16B-nl,Salesforce/codegen-16B-nl,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-MoE-A2.7B,Qwen/Qwen1.5-MoE-A2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-1.3b,EleutherAI/pythia-1.3b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-2.7b,EleutherAI/pythia-2.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-14B,Qwen/Qwen1.5-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen2-beta-14B,Qwen/Qwen2-beta-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-30b,facebook/opt-30b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-1.3B,EleutherAI/gpt-neo-1.3B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,databricks/dbrx-base,databricks/dbrx-base,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/databricks/dbrx-base/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aa515-4cd9ee526e393e0b2c53f940;1c287a92-4eed-4b60-b1ed-c821656f6375) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/databricks/dbrx-base/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like databricks/dbrx-base is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,huggyllama/llama-7b,huggyllama/llama-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neox-20b,EleutherAI/gpt-neox-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-4B,Qwen/Qwen1.5-4B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-1.8B,Qwen/Qwen1.5-1.8B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-14B,Qwen/Qwen-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 550, in from_pretrained + model_class = get_class_from_dynamic_module( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 498, in get_class_from_dynamic_module + final_module = get_cached_module_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 323, in get_cached_module_file + modules_needed = check_imports(resolved_module_file) + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 181, in check_imports + raise ImportError( +ImportError: This modeling file requires the following packages that were not found in your environment: transformers_stream_generator. Run `pip install transformers_stream_generator` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-7.5B,facebook/xglm-7.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-12b,EleutherAI/pythia-12b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-7B,Qwen/Qwen1.5-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-34B,01-ai/Yi-34B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/gemma-2b,google/gemma-2b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/gemma-2b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aa99c-1760a7a77cd8ee007be0f51e;b3893274-4098-4455-b131-b1af67f7075e) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/gemma-2b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/gemma-2b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/polyglot-ko-12.8b,EleutherAI/polyglot-ko-12.8b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-125m,EleutherAI/gpt-neo-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-13b,facebook/opt-13b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/recurrentgemma-7b,google/recurrentgemma-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 404 Client Error: Not Found for url: https://huggingface.co/google/recurrentgemma-7b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1823, in _raise_on_head_call_error + raise head_call_error + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 352, in hf_raise_for_status + raise RepositoryNotFoundError(message, response) from e +huggingface_hub.utils._errors.RepositoryNotFoundError: 404 Client Error. (Request ID: Root=1-664aaa9d-18c73fe40319df131108d89a;65834509-e2ea-4a4a-8802-05955ae3eca9) + +Repository Not Found for url: https://huggingface.co/google/recurrentgemma-7b/resolve/main/config.json. +Please make sure you specified the correct `repo_id` and `repo_type`. +If you are trying to access a private or gated repo, make sure you are authenticated. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 422, in cached_file + raise EnvironmentError( +OSError: google/recurrentgemma-7b is not a local folder and is not a valid model identifier listed on 'https://huggingface.co/models' +If this is a private repository, make sure to pass a token having permission to this repo either by logging in with `huggingface-cli login` or by passing `token=` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-410m,EleutherAI/pythia-410m,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-32B,Qwen/Qwen1.5-32B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-4.5B,facebook/xglm-4.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,TencentARC/Mistral_Pro_8B_v0.1,TencentARC/Mistral_Pro_8B_v0.1,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-564M,facebook/xglm-564M,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-350m,facebook/opt-350m,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-6.7b,facebook/opt-6.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-72B,Qwen/Qwen-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 550, in from_pretrained + model_class = get_class_from_dynamic_module( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 498, in get_class_from_dynamic_module + final_module = get_cached_module_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 323, in get_cached_module_file + modules_needed = check_imports(resolved_module_file) + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 181, in check_imports + raise ImportError( +ImportError: This modeling file requires the following packages that were not found in your environment: transformers_stream_generator. Run `pip install transformers_stream_generator` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-6B,01-ai/Yi-6B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen2-beta-72B,Qwen/Qwen2-beta-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 96.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-2.7B,EleutherAI/gpt-neo-2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-66b,facebook/opt-66b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 162.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,internlm/internlm2-20b,internlm/internlm2-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 558, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-exllama-v1-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-6.7b,EleutherAI/pythia-6.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 461, in post_init_awq_exllama_modules + model = exllama_post_init(model) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 144, in exllama_post_init + submodule.post_init() + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllama.py"", line 77, in post_init + self.q4 = exl_ext.make_q4( +NameError: name 'exl_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,1,64,1 +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-1.4b,EleutherAI/pythia-1.4b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 1031, in forward + outputs = self.gpt_neox( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 922, in forward + outputs = layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 687, in forward + attention_layer_outputs = self.attention( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 172, in forward + qkv = self.query_key_value(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Salesforce/codegen-6B-nl,Salesforce/codegen-6B-nl,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/codegen/modeling_codegen.py"", line 667, in forward + transformer_outputs = self.transformer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/codegen/modeling_codegen.py"", line 536, in forward + outputs = block( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/codegen/modeling_codegen.py"", line 272, in forward + attn_outputs = self.attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/codegen/modeling_codegen.py"", line 163, in forward + qkv = self.qkv_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-70m,EleutherAI/pythia-70m,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 1031, in forward + outputs = self.gpt_neox( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 922, in forward + outputs = layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 687, in forward + attention_layer_outputs = self.attention( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 172, in forward + qkv = self.query_key_value(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-110B,Qwen/Qwen1.5-110B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 32.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciCoder-1b,Deci/DeciCoder-1b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 68, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1164, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 968, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 713, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) +TypeError: DeciCoderAttention.forward() got an unexpected keyword argument 'cache_position' + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/recurrentgemma-2b,google/recurrentgemma-2b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/recurrentgemma-2b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aaa44-5f302e853db5bb8f135149b5;bf92de86-fc4a-494b-9935-69d15aad2191) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/recurrentgemma-2b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/recurrentgemma-2b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,internlm/internlm-20b,internlm/internlm-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm-20b/80729bcf52fbc4553d965926b27304ac5e156d98/modeling_internlm.py"", line 976, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm-20b/80729bcf52fbc4553d965926b27304ac5e156d98/modeling_internlm.py"", line 866, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm-20b/80729bcf52fbc4553d965926b27304ac5e156d98/modeling_internlm.py"", line 583, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm-20b/80729bcf52fbc4553d965926b27304ac5e156d98/modeling_internlm.py"", line 339, in forward + query_states = self.q_proj(hidden_states).view(bsz, q_len, self.num_heads, self.head_dim).transpose(1, 2) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-2.7b,facebook/opt-2.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 1117, in forward + outputs = self.model.decoder( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 883, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 524, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 154, in forward + query_states = self.q_proj(hidden_states) * self.scaling + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-125m,facebook/opt-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 1117, in forward + outputs = self.model.decoder( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 883, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 524, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 154, in forward + query_states = self.q_proj(hidden_states) * self.scaling + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/gemma-7b,google/gemma-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/gemma-7b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aa9ef-446c812a6bb84e200264b4e9;60b8f8a0-ab75-49fe-a5fa-0a14c6516974) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/gemma-7b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/gemma-7b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,huggyllama/llama-13b,huggyllama/llama-13b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1164, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 968, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 713, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 327, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-7B,Qwen/Qwen-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 550, in from_pretrained + model_class = get_class_from_dynamic_module( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 498, in get_class_from_dynamic_module + final_module = get_cached_module_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 323, in get_cached_module_file + modules_needed = check_imports(resolved_module_file) + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 181, in check_imports + raise ImportError( +ImportError: This modeling file requires the following packages that were not found in your environment: transformers_stream_generator. Run `pip install transformers_stream_generator` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-72B,Qwen/Qwen1.5-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 96.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciLM-7B,Deci/DeciLM-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 68, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/Deci/DeciLM-7B/c3c9f4226801dc0433f32aebffe0aac68ee2f051/transformers_v4_35_2__modeling_llama.py"", line 1034, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/Deci/DeciLM-7B/c3c9f4226801dc0433f32aebffe0aac68ee2f051/modeling_decilm.py"", line 274, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/Deci/DeciLM-7B/c3c9f4226801dc0433f32aebffe0aac68ee2f051/transformers_v4_35_2__modeling_llama.py"", line 672, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/Deci/DeciLM-7B/c3c9f4226801dc0433f32aebffe0aac68ee2f051/modeling_decilm.py"", line 84, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-j-6b,EleutherAI/gpt-j-6b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 68, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gptj/modeling_gptj.py"", line 1124, in forward + transformer_outputs = self.transformer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gptj/modeling_gptj.py"", line 950, in forward + outputs = block( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gptj/modeling_gptj.py"", line 578, in forward + attn_outputs = self.attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gptj/modeling_gptj.py"", line 224, in forward + query = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-0.5B,Qwen/Qwen1.5-0.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1149, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1034, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 748, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 249, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-160m,EleutherAI/pythia-160m,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 1031, in forward + outputs = self.gpt_neox( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 922, in forward + outputs = layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 687, in forward + attention_layer_outputs = self.attention( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 172, in forward + qkv = self.query_key_value(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Salesforce/codegen-16B-nl,Salesforce/codegen-16B-nl,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/codegen/modeling_codegen.py"", line 667, in forward + transformer_outputs = self.transformer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/codegen/modeling_codegen.py"", line 536, in forward + outputs = block( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/codegen/modeling_codegen.py"", line 272, in forward + attn_outputs = self.attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/codegen/modeling_codegen.py"", line 163, in forward + qkv = self.qkv_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-MoE-A2.7B,Qwen/Qwen1.5-MoE-A2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3640, in from_pretrained + hf_quantizer.preprocess_model( + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 182, in preprocess_model + return self._process_model_before_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 85, in _process_model_before_weight_loading + model, has_been_replaced = replace_with_awq_linear( + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 178, in replace_with_awq_linear + _, has_been_replaced = replace_with_awq_linear( + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 178, in replace_with_awq_linear + _, has_been_replaced = replace_with_awq_linear( + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 178, in replace_with_awq_linear + _, has_been_replaced = replace_with_awq_linear( + [Previous line repeated 1 more time] + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 165, in replace_with_awq_linear + model._modules[name] = target_cls( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 46, in __init__ + assert out_features % (32 // self.w_bit) == 0 +AssertionError + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-1.3b,EleutherAI/pythia-1.3b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 68, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 1031, in forward + outputs = self.gpt_neox( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 922, in forward + outputs = layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 687, in forward + attention_layer_outputs = self.attention( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 172, in forward + qkv = self.query_key_value(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-2.7b,EleutherAI/pythia-2.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 1031, in forward + outputs = self.gpt_neox( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 922, in forward + outputs = layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 687, in forward + attention_layer_outputs = self.attention( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 172, in forward + qkv = self.query_key_value(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-14B,Qwen/Qwen1.5-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1149, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1034, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 748, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 249, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen2-beta-14B,Qwen/Qwen2-beta-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1149, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1034, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 748, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 249, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-30b,facebook/opt-30b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 1117, in forward + outputs = self.model.decoder( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 883, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 524, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 154, in forward + query_states = self.q_proj(hidden_states) * self.scaling + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-1.3B,EleutherAI/gpt-neo-1.3B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 68, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 971, in forward + transformer_outputs = self.transformer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 839, in forward + outputs = block( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 566, in forward + attn_outputs = self.attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 518, in forward + return self.attention( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 260, in forward + query = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,databricks/dbrx-base,databricks/dbrx-base,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/databricks/dbrx-base/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aa50f-438d9f71258d9d1c056536e1;fab07f65-199e-4628-99dc-2a71d05c15a8) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/databricks/dbrx-base/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like databricks/dbrx-base is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,huggyllama/llama-7b,huggyllama/llama-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1164, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 968, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 713, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 327, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neox-20b,EleutherAI/gpt-neox-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 68, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 1031, in forward + outputs = self.gpt_neox( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 922, in forward + outputs = layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 687, in forward + attention_layer_outputs = self.attention( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 172, in forward + qkv = self.query_key_value(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-4B,Qwen/Qwen1.5-4B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1149, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1034, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 748, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 249, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-1.8B,Qwen/Qwen1.5-1.8B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1149, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1034, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 748, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 249, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-14B,Qwen/Qwen-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 550, in from_pretrained + model_class = get_class_from_dynamic_module( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 498, in get_class_from_dynamic_module + final_module = get_cached_module_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 323, in get_cached_module_file + modules_needed = check_imports(resolved_module_file) + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 181, in check_imports + raise ImportError( +ImportError: This modeling file requires the following packages that were not found in your environment: transformers_stream_generator. Run `pip install transformers_stream_generator` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-7.5B,facebook/xglm-7.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/xglm/modeling_xglm.py"", line 761, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/xglm/modeling_xglm.py"", line 647, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/xglm/modeling_xglm.py"", line 414, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/xglm/modeling_xglm.py"", line 244, in forward + query_states = self.q_proj(hidden_states) * self.scaling + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-12b,EleutherAI/pythia-12b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 1031, in forward + outputs = self.gpt_neox( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 922, in forward + outputs = layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 687, in forward + attention_layer_outputs = self.attention( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 172, in forward + qkv = self.query_key_value(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-7B,Qwen/Qwen1.5-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1149, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1034, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 748, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 249, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-34B,01-ai/Yi-34B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 68, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1164, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 968, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 713, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 327, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/gemma-2b,google/gemma-2b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/gemma-2b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aa993-7bd2fc9d24fc39b05f415c50;6b5b352e-ab93-4509-8ee1-d7bf66379068) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/gemma-2b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/gemma-2b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/polyglot-ko-12.8b,EleutherAI/polyglot-ko-12.8b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 68, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 1031, in forward + outputs = self.gpt_neox( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 922, in forward + outputs = layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 687, in forward + attention_layer_outputs = self.attention( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 172, in forward + qkv = self.query_key_value(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-125m,EleutherAI/gpt-neo-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 68, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 971, in forward + transformer_outputs = self.transformer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 839, in forward + outputs = block( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 566, in forward + attn_outputs = self.attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 518, in forward + return self.attention( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 260, in forward + query = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-13b,facebook/opt-13b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 1117, in forward + outputs = self.model.decoder( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 883, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 524, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 154, in forward + query_states = self.q_proj(hidden_states) * self.scaling + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/recurrentgemma-7b,google/recurrentgemma-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 404 Client Error: Not Found for url: https://huggingface.co/google/recurrentgemma-7b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1823, in _raise_on_head_call_error + raise head_call_error + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 352, in hf_raise_for_status + raise RepositoryNotFoundError(message, response) from e +huggingface_hub.utils._errors.RepositoryNotFoundError: 404 Client Error. (Request ID: Root=1-664aaa97-73d091fc4f369e11602676a9;cabb7c40-abc4-4bdb-8451-8159f0672eeb) + +Repository Not Found for url: https://huggingface.co/google/recurrentgemma-7b/resolve/main/config.json. +Please make sure you specified the correct `repo_id` and `repo_type`. +If you are trying to access a private or gated repo, make sure you are authenticated. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 422, in cached_file + raise EnvironmentError( +OSError: google/recurrentgemma-7b is not a local folder and is not a valid model identifier listed on 'https://huggingface.co/models' +If this is a private repository, make sure to pass a token having permission to this repo either by logging in with `huggingface-cli login` or by passing `token=` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-410m,EleutherAI/pythia-410m,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 1031, in forward + outputs = self.gpt_neox( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 922, in forward + outputs = layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 687, in forward + attention_layer_outputs = self.attention( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 172, in forward + qkv = self.query_key_value(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-32B,Qwen/Qwen1.5-32B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1149, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1034, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 748, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 249, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-4.5B,facebook/xglm-4.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/xglm/modeling_xglm.py"", line 761, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/xglm/modeling_xglm.py"", line 647, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/xglm/modeling_xglm.py"", line 414, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/xglm/modeling_xglm.py"", line 244, in forward + query_states = self.q_proj(hidden_states) * self.scaling + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,TencentARC/Mistral_Pro_8B_v0.1,TencentARC/Mistral_Pro_8B_v0.1,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/mistral/modeling_mistral.py"", line 1139, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/mistral/modeling_mistral.py"", line 1024, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/mistral/modeling_mistral.py"", line 738, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/mistral/modeling_mistral.py"", line 251, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-564M,facebook/xglm-564M,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/xglm/modeling_xglm.py"", line 761, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/xglm/modeling_xglm.py"", line 647, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/xglm/modeling_xglm.py"", line 414, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/xglm/modeling_xglm.py"", line 244, in forward + query_states = self.q_proj(hidden_states) * self.scaling + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-350m,facebook/opt-350m,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 1117, in forward + outputs = self.model.decoder( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 835, in forward + inputs_embeds = self.project_in(inputs_embeds) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-6.7b,facebook/opt-6.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 1117, in forward + outputs = self.model.decoder( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 883, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 524, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/opt/modeling_opt.py"", line 154, in forward + query_states = self.q_proj(hidden_states) * self.scaling + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-72B,Qwen/Qwen-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 550, in from_pretrained + model_class = get_class_from_dynamic_module( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 498, in get_class_from_dynamic_module + final_module = get_cached_module_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 323, in get_cached_module_file + modules_needed = check_imports(resolved_module_file) + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 181, in check_imports + raise ImportError( +ImportError: This modeling file requires the following packages that were not found in your environment: transformers_stream_generator. Run `pip install transformers_stream_generator` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-6B,01-ai/Yi-6B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 68, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1164, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 968, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 713, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 327, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen2-beta-72B,Qwen/Qwen2-beta-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 96.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-2.7B,EleutherAI/gpt-neo-2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 68, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 971, in forward + transformer_outputs = self.transformer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 839, in forward + outputs = block( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 566, in forward + attn_outputs = self.attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 518, in forward + return self.attention( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 260, in forward + query = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-66b,facebook/opt-66b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 162.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,internlm/internlm2-20b,internlm/internlm2-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm2-20b/f363ea8a116b3ea829c7a068ca24bc9d3e668083/modeling_internlm2.py"", line 1047, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm2-20b/f363ea8a116b3ea829c7a068ca24bc9d3e668083/modeling_internlm2.py"", line 932, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm2-20b/f363ea8a116b3ea829c7a068ca24bc9d3e668083/modeling_internlm2.py"", line 639, in forward + hidden_states, self_attn_weights, present_key_value = self.attention( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm2-20b/f363ea8a116b3ea829c7a068ca24bc9d3e668083/modeling_internlm2.py"", line 344, in forward + qkv_states = self.wqkv(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-6.7b,EleutherAI/pythia-6.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 1031, in forward + outputs = self.gpt_neox( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 922, in forward + outputs = layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 687, in forward + attention_layer_outputs = self.attention( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neox/modeling_gpt_neox.py"", line 172, in forward + qkv = self.query_key_value(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-1.4b,EleutherAI/pythia-1.4b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,,,MB,1625.034752,2254.962688,0.0,1608.515584,1463.6928,s,10,1.2100343170166015,0.12100343170166014,0.001022141281647672,0.12073452758789063,0.12208388671874999,0.1227883918762207,0.12335199600219726,"[0.1234928970336914, 0.12192733001708984, 0.12002118682861328, 0.12029216003417968, 0.12038220977783202, 0.1198985595703125, 0.1206871337890625, 0.12078192138671875, 0.12138086700439453, 0.12117005157470703]",tokens/s,2115.6424772413106,kWh,1.4169729851856437e-06,7.764314618920512e-07,6.3009111253815796e-06,8.494315572459274e-06,tokens/kWh,30137801.90013389,MB,1625.034752,2254.962688,0.0,1608.515584,1560.974848,s,10,70.266875,7.0266874999999995,0.015802312461115554,7.021189208984375,7.04873154296875,7.0518982421875,7.0544316015624995,"[7.05506494140625, 7.04802783203125, 7.0200771484375, 7.04010400390625, 7.03416455078125, 7.01334814453125, 7.0082919921875, 7.02230126953125, 7.0167861328125, 7.008708984375]",tokens/s,8.965817819562917,kWh,8.271868514320838e-05,4.533574606695635e-05,0.00036366391262401533,0.0004917183438341799,tokens/kWh,128122.12680282927,,s,629,71.24009573364258,0.11325929369418533,0.01441184935436087,0.111351806640625,0.11215667266845704,0.11239075622558593,0.2323562939453125,"[0.11357389068603516, 0.11349612426757813, 0.11200812530517579, 0.11255910491943359, 0.1121167984008789, 0.11198560333251953, 0.11219558715820313, 0.11200819396972657, 0.11196825408935547, 0.11275468444824219, 0.11222630310058594, 0.11180441284179687, 0.11134464263916016, 0.11147980499267578, 0.11213721466064454, 0.11178495788574219, 0.11166719818115234, 0.11174092864990234, 0.11117977905273438, 0.1110456314086914, 0.11137638092041016, 0.11204710388183593, 0.11143679809570313, 0.11129449462890625, 0.11121353912353515, 0.1112074203491211, 0.11121561431884766, 0.11156582641601563, 0.11124736022949219, 0.1111910400390625, 0.1112965087890625, 0.11185664367675781, 0.11158732604980469, 0.1118187484741211, 0.11172659301757812, 0.11204096221923829, 0.11242803192138671, 0.11206655883789063, 0.11170611572265625, 0.1124925765991211, 0.11240444946289062, 0.11224883270263672, 0.11220172882080078, 0.11224473571777344, 0.11245168304443359, 0.11209001922607421, 0.11190579223632813, 0.11188838195800781, 0.11227750396728516, 0.11232051086425782, 0.11256114959716797, 0.11237478637695313, 0.11200921630859376, 0.11215462493896484, 0.1119549789428711, 0.11200611114501953, 0.11207987213134765, 0.11194371032714844, 0.11197846221923828, 0.11205836486816406, 0.11220787048339843, 0.11210034942626954, 0.2345594940185547, 0.1117286376953125, 0.11186585235595703, 0.11204914855957031, 0.11177677154541016, 0.1115832290649414, 0.11116851043701172, 0.1114071044921875, 0.11131391906738282, 0.11137741088867187, 0.11209728240966797, 0.11197542572021485, 0.11190681457519532, 0.11151052856445312, 0.11113471984863281, 0.11245875549316406, 0.11208191680908203, 0.11218540954589844, 0.11216889953613281, 0.11212287902832031, 0.11191302490234376, 0.11190470123291016, 0.11167948913574219, 0.11192729949951172, 0.11203174591064453, 0.11200819396972657, 0.11200409698486329, 0.11204198455810546, 0.11203584289550782, 0.11207577514648437, 0.11203174591064453, 0.11206861114501954, 0.11190476989746094, 0.11196518707275391, 0.11201945495605468, 0.11220070648193359, 0.1121239013671875, 0.11213619232177735, 0.11214745330810547, 0.11210034942626954, 0.11117158508300781, 0.11123609924316406, 0.11129138946533203, 0.11250688171386719, 0.11131903839111328, 0.11113369750976562, 0.11244338989257813, 0.11134259033203125, 0.11169075012207032, 0.11205427551269531, 0.11214745330810547, 0.11191705322265624, 0.11207884979248046, 0.11197849273681641, 0.11211673736572265, 0.11211161804199218, 0.11191500854492188, 0.11222118377685547, 0.11239730834960937, 0.11132415771484375, 0.11126681518554687, 0.1112985610961914, 0.11182182312011718, 0.23266201782226562, 0.11202662658691406, 0.11120025634765625, 0.11109683227539062, 0.11149005126953125, 0.11119513702392578, 0.11112754821777343, 0.11123916625976563, 0.11105177307128906, 0.1111900177001953, 0.11129036712646484, 0.11116851043701172, 0.11114803314208985, 0.11196927642822266, 0.11116134643554687, 0.11125965118408203, 0.11103539276123046, 0.11127398681640625, 0.11154434967041016, 0.11195798492431641, 0.11191193389892579, 0.1111377944946289, 0.1111562271118164, 0.11188735961914062, 0.11112754821777343, 0.11104972839355469, 0.11122278594970703, 0.11126271820068359, 0.11140402984619141, 0.11229183959960938, 0.11202559661865234, 0.11156588745117188, 0.1111817626953125, 0.11196927642822266, 0.11145932769775391, 0.11130470275878906, 0.11111219024658203, 0.11142041778564453, 0.11112140655517579, 0.11115110778808594, 0.11208601379394531, 0.11188428497314454, 0.11182284545898437, 0.11141017913818359, 0.11125043487548827, 0.11123404693603516, 0.11116646575927734, 0.11186994934082031, 0.11114189147949219, 0.11129446411132812, 0.11111833953857422, 0.1112442855834961, 0.11120845031738281, 0.11120435333251953, 0.11193344116210938, 0.11145116424560547, 0.11117052459716797, 0.11132723236083984, 0.11132109069824218, 0.1112074203491211, 0.11136307525634766, 0.11232870483398437, 0.11170201873779297, 0.2324981689453125, 0.11132723236083984, 0.11152384185791016, 0.11143577575683594, 0.11334963226318359, 0.11229798126220703, 0.11186176300048828, 0.11187916564941407, 0.11197337341308594, 0.11203993225097657, 0.11198976135253906, 0.11205836486816406, 0.11139788818359375, 0.11115929412841796, 0.11112754821777343, 0.11136921691894532, 0.11119513702392578, 0.11117056274414062, 0.11119206237792968, 0.11117465972900391, 0.11126067352294922, 0.11122380828857421, 0.11115827178955077, 0.11134873962402343, 0.1123768310546875, 0.11169586944580077, 0.11113676452636718, 0.11131903839111328, 0.11172249603271485, 0.11171635437011719, 0.11152793884277344, 0.11154022216796874, 0.11121971130371094, 0.11125247955322265, 0.11109174346923828, 0.11109065246582031, 0.111283203125, 0.11159961700439452, 0.11181670379638672, 0.11218841552734375, 0.11207065582275391, 0.11193856048583985, 0.11212083435058594, 0.11199385833740234, 0.11282943725585938, 0.11240140533447265, 0.11203072357177735, 0.11216486358642579, 0.11200102233886719, 0.11202969360351563, 0.112, 0.11213005065917969, 0.11189555358886719, 0.11197647857666015, 0.11237065887451173, 0.11224269104003906, 0.11207270050048829, 0.11202355194091797, 0.11200204467773438, 0.11213520050048828, 0.1117081298828125, 0.11131187438964844, 0.11138969421386719, 0.23215206909179686, 0.11141426849365234, 0.11262156677246093, 0.11177779388427735, 0.11186585235595703, 0.11191295623779297, 0.11129036712646484, 0.11202355194091797, 0.11225702667236329, 0.11216793823242187, 0.11205120086669922, 0.11202252960205078, 0.11183513641357422, 0.11183513641357422, 0.1118361587524414, 0.11228672027587891, 0.11196723175048828, 0.11265023803710937, 0.11201439666748046, 0.11204192352294921, 0.11198668670654296, 0.11204812622070312, 0.1128499526977539, 0.11216585540771484, 0.11173174285888672, 0.11194057464599609, 0.11185971069335937, 0.11190889739990234, 0.11204195404052734, 0.11242700958251953, 0.1118361587524414, 0.11156070709228516, 0.11167436981201172, 0.11125555419921875, 0.11129446411132812, 0.11137535858154297, 0.1112125473022461, 0.11132723236083984, 0.11112652587890624, 0.11121356964111329, 0.11136511993408203, 0.11127808380126954, 0.11119308471679687, 0.11122278594970703, 0.11130265808105469, 0.11126374053955078, 0.11110604858398437, 0.11118386840820313, 0.11117158508300781, 0.11144499206542968, 0.11132316589355469, 0.11124323272705078, 0.11118899536132812, 0.11127705383300782, 0.11124531555175782, 0.11131084442138672, 0.11114189147949219, 0.11251712036132812, 0.11123916625976563, 0.1111562271118164, 0.11102105712890625, 0.111172607421875, 0.11132518768310547, 0.23168409729003905, 0.11126067352294922, 0.11120333099365234, 0.11120333099365234, 0.11132112121582032, 0.11132003021240235, 0.1112279052734375, 0.11117772674560547, 0.11110399627685547, 0.1111541748046875, 0.1111695327758789, 0.11118796539306641, 0.11124940490722657, 0.11112960052490234, 0.11125452423095702, 0.11125759887695312, 0.11122585296630859, 0.11124224090576172, 0.11116646575927734, 0.11114598083496094, 0.11131494140625, 0.11128832244873046, 0.11116134643554687, 0.11131391906738282, 0.11123404693603516, 0.11126681518554687, 0.11123916625976563, 0.1112985610961914, 0.11178495788574219, 0.11137741088867187, 0.11124121856689453, 0.11152281951904297, 0.11121766662597657, 0.11117362976074219, 0.1112125473022461, 0.11121868896484376, 0.111098876953125, 0.1111900177001953, 0.11227442932128906, 0.11120025634765625, 0.11103846740722656, 0.11157708740234375, 0.11146240234375, 0.11182284545898437, 0.11149517059326172, 0.111499267578125, 0.11131289672851563, 0.11149209594726563, 0.11197853088378906, 0.11136713409423828, 0.11115724945068359, 0.11126067352294922, 0.11121151733398438, 0.11126067352294922, 0.11109478759765624, 0.11183001708984375, 0.11150643157958984, 0.11150745391845703, 0.11111833953857422, 0.11116544342041015, 0.11122994995117187, 0.11125452423095702, 0.1113088607788086, 0.23280429077148437, 0.11111014556884766, 0.11133030700683594, 0.11111119842529296, 0.11112957000732422, 0.11102105712890625, 0.11110809326171875, 0.11106201934814453, 0.11103334045410156, 0.11108870697021485, 0.11110905456542969, 0.11238092803955078, 0.11154227447509765, 0.11123506927490234, 0.11109273529052735, 0.11120333099365234, 0.11103129577636718, 0.11100672149658203, 0.11094937896728516, 0.1110487060546875, 0.11100672149658203, 0.1109534683227539, 0.11100569915771484, 0.11172767639160157, 0.11108550262451172, 0.11122688293457031, 0.11107020568847656, 0.11119721221923828, 0.11113673400878907, 0.11125043487548827, 0.1111562271118164, 0.11116236877441406, 0.1111551971435547, 0.11128012847900391, 0.11118694305419922, 0.11127091217041016, 0.11107123565673828, 0.11122994995117187, 0.11122380828857421, 0.1111551971435547, 0.11100466918945312, 0.11141836547851562, 0.11136511993408203, 0.11123814392089844, 0.11123302459716797, 0.11117670440673828, 0.11112960052490234, 0.11146444702148438, 0.11135078430175781, 0.11130470275878906, 0.11135897827148437, 0.11127808380126954, 0.11142451477050781, 0.11136716461181641, 0.1120901107788086, 0.11140300750732422, 0.11124326324462891, 0.11130985260009765, 0.111257568359375, 0.1111551971435547, 0.1111756820678711, 0.11127295684814453, 0.11109580993652343, 0.23243571472167968, 0.11107839965820313, 0.11138361358642578, 0.11138758087158203, 0.111172607421875, 0.11116441345214843, 0.11112754821777343, 0.11145728302001953, 0.11120230102539062, 0.11218739318847656, 0.11158425903320313, 0.11142348480224609, 0.11124326324462891, 0.11134566497802735, 0.11134770965576171, 0.11127603149414063, 0.11175628662109376, 0.11245772552490234, 0.11182899475097656, 0.11139379119873047, 0.11116031646728515, 0.11131187438964844, 0.11135897827148437, 0.11138253021240234, 0.11129446411132812, 0.11140914916992188, 0.1112995834350586, 0.11167334747314453, 0.11146444702148438, 0.11127398681640625, 0.1114286117553711, 0.11160883331298828, 0.111388671875, 0.11126783752441406, 0.111283203125, 0.11141529846191406, 0.11138355255126953, 0.11169382476806641, 0.111425537109375, 0.11148297882080078, 0.11134454345703125, 0.11160371398925781, 0.11167334747314453, 0.11133235168457031, 0.11129036712646484, 0.11148089599609375, 0.11127085113525391, 0.11113471984863281, 0.11119821166992187, 0.11146444702148438, 0.1114419174194336, 0.11144499206542968, 0.11137741088867187, 0.11137126159667969, 0.11174604797363281, 0.11149619293212891, 0.11148902130126953, 0.1114419174194336, 0.1115525131225586, 0.11180646514892578, 0.11210854339599609, 0.11200819396972657, 0.11173580932617187, 0.2335253143310547, 0.11135481262207031, 0.11140300750732422, 0.11150540924072265, 0.11128627014160156, 0.11110707092285156, 0.11116851043701172, 0.11114393615722656, 0.1112995834350586, 0.11127808380126954, 0.11109273529052735, 0.11120953369140625, 0.1113087387084961, 0.11121971130371094, 0.11118284606933594, 0.11124736022949219, 0.1122508773803711, 0.11217715454101562, 0.1113733139038086, 0.11129344177246094, 0.11147058868408204, 0.11150335693359376, 0.111351806640625, 0.11135897827148437, 0.11148492431640625, 0.11133952331542969, 0.11130675506591797, 0.11139481353759766, 0.11144703674316406, 0.11154124450683593, 0.11132825469970703, 0.11111219024658203, 0.11122073364257813, 0.11118386840820313, 0.11125971221923828, 0.1125447006225586, 0.11146546936035157, 0.11134361267089844, 0.11123097229003906, 0.11145728302001953, 0.11127091217041016, 0.1112074203491211, 0.1114224624633789, 0.11159859466552734, 0.11146854400634766, 0.11118284606933594, 0.11170816040039062, 0.11130777740478516, 0.11138662719726562, 0.11138253021240234, 0.11114701080322266, 0.11128729248046874, 0.11132109069824218, 0.11136102294921875, 0.11118284606933594, 0.11117881774902344, 0.11109677124023437, 0.11118796539306641, 0.11113267517089843, 0.11119411468505859, 0.111246337890625, 0.1112279052734375, 0.11181465911865235, 0.23307980346679688, 0.11142041778564453, 0.1113538589477539, 0.11121561431884766, 0.11131903839111328, 0.11121561431884766, 0.11105689239501954, 0.111246337890625, 0.11114291381835938, 0.1110835189819336, 0.1110241928100586, 0.11124934387207032, 0.11121561431884766, 0.11143475341796875, 0.11115007781982422, 0.11116031646728515, 0.11110912322998047, 0.11113881683349609, 0.11112754821777343, 0.11146137237548828, 0.11116646575927734, 0.1111173095703125, 0.11114291381835938, 0.11114495849609375, 0.11101900482177735, 0.11122176361083984, 0.11194265747070313, 0.11108147430419922, 0.11138662719726562, 0.11130368041992188, 0.11116748809814453, 0.11115110778808594, 0.11116649627685547, 0.11112751770019531, 0.1111695327758789, 0.11108454132080078, 0.11120127868652344, 0.11118796539306641, 0.11125555419921875, 0.11124018859863281, 0.11112140655517579, 0.11123712158203125, 0.11195391845703125, 0.1116231689453125, 0.111388671875, 0.11123404693603516, 0.11115110778808594, 0.11121459197998047, 0.1111562271118164, 0.11116748809814453, 0.11117772674560547, 0.11126374053955078, 0.1111910400390625, 0.11120333099365234, 0.11108454132080078, 0.11108761596679688, 0.1111203842163086, 0.1111357421875, 0.11103231811523437, 0.11127603149414063, 0.11121868896484376, 0.11217817687988281, 0.11138457489013671]",tokens/s,8.829297511779728,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Salesforce/codegen-6B-nl,Salesforce/codegen-6B-nl,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1469, in _autoset_attn_implementation + cls._check_and_enable_flash_attn_2( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1560, in _check_and_enable_flash_attn_2 + raise ValueError( +ValueError: CodeGenForCausalLM does not support Flash Attention 2.0 yet. Please request to add support where the model is hosted, on its model hub page: https://huggingface.co//tmp/tmpoc0pnttw/no_weights_model/discussions/new or in the Transformers GitHub repo: https://github.com/huggingface/transformers/issues/new + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-70m,EleutherAI/pythia-70m,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,1279.725568,872.93952,0.0,226.492416,184.397824,s,12,0.17399129676818847,0.014499274730682373,0.0004405777377392672,0.014355583667755127,0.01475712022781372,0.015286664152145383,0.01576770728111267,"[0.01588796806335449, 0.014419008255004882, 0.014310144424438476, 0.01430396842956543, 0.014377471923828124, 0.014377087593078612, 0.01433407974243164, 0.014250847816467285, 0.014390144348144531, 0.01479468822479248, 0.014266719818115234, 0.014279168128967285]",tokens/s,17656.055544507362,kWh,1.7155437431226257e-07,9.399407219055296e-08,3.2474264005518264e-07,5.902910865579981e-07,tokens/kWh,433684339.52262825,MB,1280.053248,872.93952,0.0,226.492416,197.932544,s,12,10.420531616210937,0.8683776346842448,0.009197306451406238,0.8653596801757812,0.8716784362792969,0.8835345275878906,0.8948784387207032,"[0.8977144165039063, 0.866240966796875, 0.8642362670898438, 0.8648695068359376, 0.8719328002929687, 0.8635298461914063, 0.862948974609375, 0.866379638671875, 0.86938916015625, 0.865849853515625, 0.8638052368164062, 0.8636349487304688]",tokens/s,72.54908174012076,kWh,1.0699516429705703e-05,5.862862695252688e-06,1.9223555826850592e-05,3.578593495180898e-05,tokens/kWh,1760468.186309475,,s,755,10.563054601669307,0.013990800796912995,0.001781085498877292,0.013702143669128418,0.014027775764465332,0.014339276885986327,0.028389335021972658,"[0.015265791893005372, 0.014698495864868164, 0.014765055656433105, 0.014334976196289062, 0.014128128051757812, 0.014334976196289062, 0.014543871879577636, 0.014234623908996581, 0.0144650239944458, 0.01454694366455078, 0.014418944358825684, 0.01439027214050293, 0.01425100803375244, 0.014261247634887696, 0.014215167999267577, 0.014323712348937988, 0.014708736419677734, 0.014633983612060546, 0.01434931182861328, 0.01467801570892334, 0.014370816230773926, 0.014229503631591797, 0.014525440216064453, 0.014231552124023437, 0.014211071968078613, 0.014244864463806153, 0.014368767738342286, 0.014796799659729003, 0.014531583786010742, 0.014297087669372559, 0.014436384201049805, 0.014063584327697753, 0.014220288276672363, 0.013916159629821777, 0.014133248329162598, 0.014468095779418945, 0.014696479797363281, 0.014910431861877441, 0.014256128311157227, 0.014102527618408203, 0.01407590389251709, 0.014150655746459961, 0.014071807861328126, 0.013922304153442382, 0.013725728034973144, 0.01371132755279541, 0.013733887672424316, 0.013731840133666993, 0.014045184135437011, 0.014156800270080566, 0.014639103889465332, 0.014141440391540527, 0.013728863716125488, 0.013714336395263671, 0.013676544189453126, 0.013729791641235351, 0.013732864379882812, 0.013691904067993164, 0.013711359977722168, 0.013743103981018067, 0.013743103981018067, 0.013660223960876464, 0.0284968318939209, 0.013691904067993164, 0.01368883228302002, 0.013686783790588379, 0.013670399665832519, 0.01399295997619629, 0.013728768348693847, 0.013667327880859375, 0.013691904067993164, 0.01365503978729248, 0.013710335731506347, 0.013666303634643554, 0.013699071884155273, 0.013669376373291015, 0.013689855575561523, 0.013734911918640137, 0.013744128227233888, 0.01374617576599121, 0.01366323184967041, 0.013693951606750488, 0.013799424171447755, 0.013740032196044923, 0.013728768348693847, 0.013709312438964843, 0.013709312438964843, 0.01379532814025879, 0.013731840133666993, 0.013713408470153808, 0.013707263946533203, 0.013776896476745605, 0.01368883228302002, 0.013706239700317382, 0.013716480255126954, 0.013789183616638183, 0.014757887840270996, 0.013946911811828614, 0.013715423583984376, 0.013734911918640137, 0.013693951606750488, 0.01369600009918213, 0.013839360237121581, 0.013735936164855958, 0.013712384223937989, 0.013700096130371094, 0.013719552040100098, 0.013707263946533203, 0.013677568435668945, 0.013913087844848633, 0.013707263946533203, 0.013698047637939453, 0.013667360305786132, 0.013723615646362305, 0.013713408470153808, 0.013699071884155273, 0.013731840133666993, 0.013686783790588379, 0.013784064292907714, 0.01374828815460205, 0.013744064331054687, 0.013783040046691895, 0.013717503547668456, 0.013740032196044923, 0.013710335731506347, 0.028412927627563478, 0.013706239700317382, 0.013686783790588379, 0.013731840133666993, 0.013669376373291015, 0.013661184310913087, 0.013645824432373046, 0.013658111572265624, 0.013669376373291015, 0.013628416061401367, 0.013694975852966309, 0.013636608123779297, 0.013705216407775878, 0.013678591728210449, 0.013858816146850587, 0.01386086368560791, 0.01368883228302002, 0.013693951606750488, 0.013660160064697266, 0.01374518394470215, 0.013757408142089844, 0.013689855575561523, 0.013717503547668456, 0.01369600009918213, 0.013725695610046386, 0.013678591728210449, 0.013732895851135254, 0.013722592353820801, 0.0136878080368042, 0.013707263946533203, 0.013698047637939453, 0.013684736251831055, 0.013669376373291015, 0.013684736251831055, 0.01375641632080078, 0.013708288192749024, 0.013765631675720215, 0.013703167915344238, 0.013721599578857421, 0.013702143669128418, 0.013694975852966309, 0.013763584136962891, 0.01369600009918213, 0.013714431762695312, 0.013794303894042969, 0.013682687759399414, 0.013697024345397948, 0.013739007949829102, 0.013708288192749024, 0.013678591728210449, 0.013742079734802246, 0.013710335731506347, 0.013716480255126954, 0.013678591728210449, 0.013706239700317382, 0.013776896476745605, 0.013827072143554688, 0.013702143669128418, 0.013737983703613281, 0.013683712005615235, 0.013726719856262207, 0.013710335731506347, 0.013788160324096679, 0.028406784057617186, 0.013711359977722168, 0.013690879821777344, 0.013694975852966309, 0.013714431762695312, 0.013697024345397948, 0.013701120376586913, 0.01367142391204834, 0.013676544189453126, 0.013768704414367675, 0.01366528034210205, 0.013711423873901367, 0.013736895561218262, 0.013741056442260742, 0.01368064022064209, 0.013705216407775878, 0.013719552040100098, 0.013714431762695312, 0.013728768348693847, 0.013700096130371094, 0.01368883228302002, 0.013689855575561523, 0.013683712005615235, 0.013835264205932616, 0.013717503547668456, 0.013843520164489746, 0.013794239997863769, 0.013710335731506347, 0.013718527793884277, 0.013725695610046386, 0.01368064022064209, 0.013674495697021484, 0.013712384223937989, 0.013718527793884277, 0.013752320289611816, 0.013718527793884277, 0.013736960411071777, 0.01367347240447998, 0.013757439613342285, 0.01370419216156006, 0.013725695610046386, 0.013716480255126954, 0.013707263946533203, 0.013740032196044923, 0.013692928314208984, 0.013726719856262207, 0.013685759544372558, 0.013721664428710938, 0.013703104019165038, 0.01367142391204834, 0.013768704414367675, 0.01369600009918213, 0.013686783790588379, 0.013691904067993164, 0.01368883228302002, 0.013734911918640137, 0.013686783790588379, 0.01375334358215332, 0.013711359977722168, 0.013979647636413574, 0.013771776199340821, 0.013735936164855958, 0.013831232070922852, 0.028631999969482423, 0.013729824066162109, 0.013675488471984864, 0.01368172836303711, 0.013717439651489258, 0.013657088279724122, 0.01366528034210205, 0.013697024345397948, 0.013677568435668945, 0.01365503978729248, 0.013760512351989745, 0.013660160064697266, 0.01366220760345459, 0.013717503547668456, 0.013717503547668456, 0.013676544189453126, 0.013710335731506347, 0.013657088279724122, 0.013947903633117676, 0.013619199752807617, 0.013640704154968262, 0.01366528034210205, 0.013693951606750488, 0.01365401554107666, 0.013627391815185547, 0.013701120376586913, 0.013742079734802246, 0.013648896217346192, 0.013675552368164062, 0.013744095802307128, 0.01440665626525879, 0.01619865608215332, 0.016926719665527345, 0.014181376457214356, 0.01386188793182373, 0.013735936164855958, 0.01376153564453125, 0.013664256095886231, 0.013693951606750488, 0.013689855575561523, 0.013661184310913087, 0.013678591728210449, 0.013649920463562011, 0.014139391899108887, 0.013820927619934082, 0.013752320289611816, 0.013699071884155273, 0.014027775764465332, 0.01368064022064209, 0.013682687759399414, 0.013675519943237305, 0.013695039749145509, 0.01367750358581543, 0.01376460838317871, 0.014016511917114258, 0.0136878080368042, 0.013678591728210449, 0.01369600009918213, 0.013848575592041015, 0.013717503547668456, 0.0136878080368042, 0.013729791641235351, 0.013838335990905762, 0.028457984924316407, 0.013737983703613281, 0.013685759544372558, 0.013691904067993164, 0.013822976112365723, 0.013697024345397948, 0.013604864120483399, 0.013649920463562011, 0.013724672317504882, 0.013774847984313965, 0.013691904067993164, 0.01366528034210205, 0.013736960411071777, 0.013702143669128418, 0.0136878080368042, 0.013701120376586913, 0.013822976112365723, 0.013692928314208984, 0.013674495697021484, 0.013799424171447755, 0.013674495697021484, 0.01368166446685791, 0.013725695610046386, 0.013814784049987794, 0.013684736251831055, 0.01368883228302002, 0.013848608016967773, 0.013685728073120117, 0.013669376373291015, 0.013674495697021484, 0.01367961597442627, 0.01367244815826416, 0.013739007949829102, 0.01367961597442627, 0.013712384223937989, 0.013697024345397948, 0.013697024345397948, 0.013740032196044923, 0.01376460838317871, 0.013657088279724122, 0.013855744361877441, 0.013742079734802246, 0.013683712005615235, 0.013691904067993164, 0.013676544189453126, 0.013682687759399414, 0.013661184310913087, 0.013661184310913087, 0.013647904396057129, 0.013673439979553222, 0.013668352127075196, 0.013643775939941406, 0.013797375679016113, 0.013658111572265624, 0.013664256095886231, 0.013652031898498534, 0.01367033576965332, 0.013643775939941406, 0.013650943756103515, 0.013666303634643554, 0.013641728401184081, 0.01367142391204834, 0.013726719856262207, 0.028403711318969727, 0.013668352127075196, 0.013674495697021484, 0.01363046360015869, 0.013641728401184081, 0.013643775939941406, 0.01365503978729248, 0.01366220760345459, 0.01366220760345459, 0.01368166446685791, 0.013640704154968262, 0.013685759544372558, 0.013637632369995116, 0.013725695610046386, 0.013702143669128418, 0.013674495697021484, 0.013714431762695312, 0.013770751953125, 0.013796352386474609, 0.013739007949829102, 0.013711359977722168, 0.01367244815826416, 0.013684736251831055, 0.013736960411071777, 0.01367244815826416, 0.013675519943237305, 0.01368992042541504, 0.01368569564819336, 0.01368166446685791, 0.01367961597442627, 0.0136878080368042, 0.01367961597442627, 0.013676544189453126, 0.013953023910522461, 0.013818880081176758, 0.01379532814025879, 0.013685759544372558, 0.01367347240447998, 0.013747200012207032, 0.013718527793884277, 0.013677568435668945, 0.013668352127075196, 0.013685759544372558, 0.01365503978729248, 0.013649920463562011, 0.013637632369995116, 0.013633536338806153, 0.013768704414367675, 0.01365401554107666, 0.013760543823242188, 0.013644767761230468, 0.013698047637939453, 0.013644800186157227, 0.013639679908752441, 0.013675519943237305, 0.013670399665832519, 0.013651968002319336, 0.013715456008911133, 0.01376972770690918, 0.013659135818481445, 0.01366528034210205, 0.013699071884155273, 0.013642751693725585, 0.02831974411010742, 0.013700096130371094, 0.013639679908752441, 0.013622271537780761, 0.013687840461730957, 0.013710304260253907, 0.013718527793884277, 0.01368166446685791, 0.01366323184967041, 0.013705216407775878, 0.01368064022064209, 0.013686783790588379, 0.013666303634643554, 0.013668352127075196, 0.013682687759399414, 0.013727744102478028, 0.013724672317504882, 0.013721599578857421, 0.013724672317504882, 0.014102527618408203, 0.01380352020263672, 0.013740032196044923, 0.013737983703613281, 0.013920255661010742, 0.013740032196044923, 0.013717503547668456, 0.013843456268310546, 0.01380352020263672, 0.013684736251831055, 0.013757439613342285, 0.013758463859558106, 0.013974528312683105, 0.013858816146850587, 0.01417420768737793, 0.013920255661010742, 0.013936639785766602, 0.01377280044555664, 0.013703167915344238, 0.01368883228302002, 0.013677568435668945, 0.013684736251831055, 0.013694975852966309, 0.013688863754272461, 0.013667296409606934, 0.013786111831665039, 0.013828096389770507, 0.013668352127075196, 0.013664256095886231, 0.013733887672424316, 0.013702143669128418, 0.013656064033508301, 0.013676544189453126, 0.01377894401550293, 0.013708288192749024, 0.0136878080368042, 0.013683712005615235, 0.013735936164855958, 0.01395404815673828, 0.013817855834960938, 0.013650943756103515, 0.013703167915344238, 0.013696063995361328, 0.013751232147216796, 0.028461055755615236, 0.013700096130371094, 0.013645824432373046, 0.014017536163330077, 0.013718527793884277, 0.015047679901123047, 0.014234623908996581, 0.014027775764465332, 0.01408512020111084, 0.014048255920410157, 0.014173184394836426, 0.013982720375061035, 0.014036992073059081, 0.013888511657714844, 0.013721599578857421, 0.01374617576599121, 0.013685759544372558, 0.013708288192749024, 0.013652992248535157, 0.013644800186157227, 0.013687871932983398, 0.013667263984680177, 0.01366220760345459, 0.01367347240447998, 0.01367961597442627, 0.013639679908752441, 0.01368166446685791, 0.013686783790588379, 0.013645824432373046, 0.013686783790588379, 0.01366220760345459, 0.013650943756103515, 0.013632512092590332, 0.013693951606750488, 0.013690879821777344, 0.013845503807067871, 0.013686783790588379, 0.013660160064697266, 0.013715456008911133, 0.01387724781036377, 0.014060544013977052, 0.01405951976776123, 0.013849599838256836, 0.013645824432373046, 0.013682687759399414, 0.013699071884155273, 0.013699071884155273, 0.013707263946533203, 0.013682687759399414, 0.013713408470153808, 0.013652992248535157, 0.013708288192749024, 0.01368166446685791, 0.01368064022064209, 0.013719552040100098, 0.013873151779174805, 0.01370419216156006, 0.013788224220275879, 0.013749183654785157, 0.013675519943237305, 0.01366329574584961, 0.01380140781402588, 0.014252032279968262, 0.029146112442016602, 0.013622271537780761, 0.013830143928527832, 0.013846528053283692, 0.013924351692199707, 0.013684736251831055, 0.013667327880859375, 0.01367347240447998, 0.01363865566253662, 0.01366323184967041, 0.013697024345397948, 0.013644800186157227, 0.013669376373291015, 0.01386086368560791, 0.013948927879333496, 0.013683712005615235, 0.013719552040100098, 0.013699071884155273, 0.013691904067993164, 0.013702143669128418, 0.013697024345397948, 0.013853728294372558, 0.013726688385009766, 0.013720576286315917, 0.013678591728210449, 0.01369600009918213, 0.013735936164855958, 0.013697024345397948, 0.013715456008911133, 0.013755392074584961, 0.013735936164855958, 0.013721599578857421, 0.013707263946533203, 0.013883392333984374, 0.013845503807067871, 0.013723648071289063, 0.013712384223937989, 0.013705216407775878, 0.01368883228302002, 0.013685759544372558, 0.013721599578857421, 0.013728768348693847, 0.013697024345397948, 0.013751296043395997, 0.013793279647827148, 0.013677568435668945, 0.01369600009918213, 0.013702143669128418, 0.013920255661010742, 0.013789183616638183, 0.013728768348693847, 0.013712384223937989, 0.01368166446685791, 0.013733920097351074, 0.013666272163391113, 0.01387724781036377, 0.013720576286315917, 0.013702143669128418, 0.013740032196044923, 0.013721599578857421, 0.013722623825073242, 0.013712384223937989, 0.01376153564453125, 0.02837708854675293, 0.01365401554107666, 0.01364684772491455, 0.013660160064697266, 0.013664256095886231, 0.013692928314208984, 0.013957119941711426, 0.01405951976776123, 0.013983743667602539, 0.01367961597442627, 0.013649920463562011, 0.013858880043029785, 0.013722559928894044, 0.013641728401184081, 0.013668352127075196, 0.013664256095886231, 0.013723648071289063, 0.01365503978729248, 0.013669376373291015, 0.01367142391204834, 0.01370419216156006, 0.01368166446685791, 0.01367142391204834, 0.013675519943237305, 0.013628416061401367, 0.013627391815185547, 0.013661184310913087, 0.013818880081176758, 0.013710335731506347, 0.013647871971130371, 0.013670399665832519, 0.01368064022064209, 0.01367347240447998, 0.013683712005615235, 0.01368883228302002, 0.013701120376586913, 0.013677568435668945, 0.013657088279724122, 0.013692928314208984, 0.013644800186157227, 0.013676544189453126, 0.013677568435668945, 0.01380352020263672, 0.013749247550964355, 0.013674495697021484, 0.013695039749145509, 0.013658047676086426, 0.013690879821777344, 0.01367244815826416, 0.013657088279724122, 0.013640704154968262, 0.013657088279724122, 0.013793279647827148, 0.013695008277893066, 0.013856736183166504, 0.013647871971130371, 0.013661215782165528, 0.013750240325927735, 0.013722623825073242, 0.013768704414367675, 0.013710335731506347, 0.013717503547668456, 0.01366220760345459, 0.028322816848754883, 0.013648896217346192, 0.013686783790588379, 0.01369600009918213, 0.013699071884155273, 0.0136878080368042, 0.013691904067993164, 0.013702143669128418, 0.013697024345397948, 0.013683712005615235, 0.013700096130371094, 0.013693951606750488, 0.013722623825073242, 0.013705216407775878, 0.013691935539245605, 0.013717472076416016, 0.013724672317504882, 0.013700096130371094, 0.013664256095886231, 0.013639679908752441, 0.013757439613342285, 0.013656064033508301, 0.013667327880859375, 0.01366431999206543, 0.013849535942077636, 0.013924351692199707, 0.013633567810058594, 0.013722592353820801, 0.013668352127075196, 0.013906944274902343, 0.013789183616638183, 0.013662240028381348, 0.013687775611877442, 0.013670399665832519, 0.01367347240447998, 0.01366528034210205, 0.013657088279724122, 0.013661184310913087, 0.013678591728210449, 0.013684736251831055, 0.013658143997192384, 0.013722592353820801, 0.01366528034210205, 0.013637632369995116, 0.01367347240447998, 0.01365503978729248, 0.01367347240447998, 0.013744128227233888, 0.01370419216156006, 0.013693951606750488, 0.013711359977722168, 0.01368064022064209, 0.01368064022064209, 0.013691904067993164, 0.01369600009918213, 0.013724672317504882, 0.013697024345397948, 0.01367142391204834, 0.013682687759399414, 0.01365503978729248, 0.013683712005615235, 0.013806591987609864, 0.013955072402954101]",tokens/s,71.4755369986145,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-110B,Qwen/Qwen1.5-110B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 32.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciCoder-1b,Deci/DeciCoder-1b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 68, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1164, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 968, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 713, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) +TypeError: DeciCoderAttention.forward() got an unexpected keyword argument 'cache_position' + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/recurrentgemma-2b,google/recurrentgemma-2b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/recurrentgemma-2b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aaa75-6f2c86ba78417c5c3491bdce;17ad7322-4be7-4fd8-9b87-0741881cc338) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/recurrentgemma-2b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/recurrentgemma-2b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,internlm/internlm-20b,internlm/internlm-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,4499.935232,14621.868032,0.0,13975.420928,13365.937152,s,10,16.410234252929687,1.6410234252929687,0.0022480805464608366,1.6403901367187501,1.6418083374023436,1.6447224060058594,1.6470536608886719,"[1.647636474609375, 1.63996875, 1.6411607666015624, 1.640802978515625, 1.64035791015625, 1.6397760009765625, 1.6397884521484376, 1.6398060302734374, 1.6405145263671874, 1.64042236328125]",tokens/s,156.00021063337155,kWh,1.9377978377872043e-05,1.0617021217130968e-05,9.108710064739678e-05,0.00012108210024239978,tokens/kWh,2114267.9181109504,MB,4499.935232,14621.868032,0.0,13975.420928,13814.413824,s,10,975.3211015625,97.53211015625,0.011202982505412533,97.52955078125,97.54984375,97.55100390625,97.55193203125,"[97.540578125, 97.52715625, 97.522640625, 97.5219296875, 97.5304609375, 97.51653125, 97.5521640625, 97.5495859375, 97.5314140625, 97.528640625]",tokens/s,0.6459411151780855,kWh,0.0011513337257504463,0.0006310327948099438,0.005346427777138807,0.007128794297699197,tokens/kWh,8837.399056434146,,s,629,988.5350541992196,1.5715978604121124,0.19496209506894618,1.548072998046875,1.5490094970703125,1.549420166015625,3.189316650390625,"[1.547826171875, 1.5488624267578126, 1.5484969482421875, 1.5483443603515625, 1.5481630859375, 1.5493089599609375, 1.548452880859375, 1.5482459716796875, 1.5481619873046875, 1.5497840576171875, 1.54772998046875, 1.5493929443359375, 1.547630615234375, 1.5480760498046875, 1.547937744140625, 1.5482552490234376, 1.5470531005859376, 1.5489443359375, 1.5480074462890625, 1.548041259765625, 1.5472659912109374, 1.549042724609375, 1.5482132568359375, 1.5479869384765625, 1.54851123046875, 1.5473919677734376, 1.5474892578125, 1.547378662109375, 1.548716064453125, 1.548441650390625, 1.5491409912109375, 1.5472435302734375, 1.5476378173828125, 1.5480125732421874, 1.5484580078125, 1.5489659423828126, 1.5496949462890626, 1.5483074951171876, 1.5473623046875, 1.547552734375, 1.547135009765625, 1.548559326171875, 1.5482398681640626, 1.5498638916015626, 1.54827880859375, 1.5483668212890624, 1.5489197998046875, 1.548549072265625, 1.54787939453125, 1.547683837890625, 1.549091796875, 1.5491134033203124, 1.548537841796875, 1.548310546875, 1.547894775390625, 1.54776171875, 1.5491932373046875, 1.5472711181640626, 1.547672607421875, 1.5473602294921875, 1.5482440185546875, 1.548180419921875, 3.190286376953125, 1.5480146484375, 1.5486883544921874, 1.5482091064453125, 1.54775244140625, 1.5473919677734376, 1.547894775390625, 1.5473858642578124, 1.548252197265625, 1.5479337158203126, 1.54754052734375, 1.5482122802734375, 1.54810986328125, 1.5476787109375, 1.5477442626953124, 1.54800634765625, 1.5478446044921874, 1.5483709716796874, 1.5485460205078125, 1.5486474609375, 1.5467080078125, 1.546982421875, 1.5480340576171876, 1.5470633544921875, 1.5475240478515624, 1.5482020263671874, 1.5471483154296874, 1.5476029052734375, 1.548863525390625, 1.5478087158203124, 1.5482347412109374, 1.54766845703125, 1.54899658203125, 1.5479388427734375, 1.547747314453125, 1.547916259765625, 1.5490928955078125, 1.548327880859375, 1.548775390625, 1.5474073486328126, 1.547236328125, 1.5476397705078124, 1.5474749755859376, 1.5477001953125, 1.5486075439453124, 1.54838427734375, 1.5483780517578125, 1.5483616943359375, 1.5482193603515626, 1.548074951171875, 1.547906005859375, 1.549001708984375, 1.5483565673828126, 1.5475179443359375, 1.5475548095703124, 1.5477432861328124, 1.54893310546875, 1.54842724609375, 1.5487344970703125, 1.5480279541015625, 1.5475323486328125, 1.5480032958984375, 1.5488101806640624, 3.191152587890625, 1.5476182861328125, 1.548537841796875, 1.5480074462890625, 1.54785791015625, 1.548368896484375, 1.5487181396484375, 1.548142578125, 1.5480279541015625, 1.547535400390625, 1.5482972412109375, 1.54859521484375, 1.5482921142578125, 1.5488890380859375, 1.5477381591796875, 1.5481619873046875, 1.54773193359375, 1.54867919921875, 1.547737060546875, 1.5486033935546875, 1.5482552490234376, 1.5477698974609375, 1.5484517822265624, 1.5468912353515625, 1.5473387451171876, 1.5476397705078124, 1.5484375, 1.547588623046875, 1.5468543701171875, 1.54747802734375, 1.547462646484375, 1.5476142578125, 1.5486832275390625, 1.5478487548828126, 1.54682373046875, 1.547431884765625, 1.54697216796875, 1.5476439208984376, 1.547420654296875, 1.547546630859375, 1.5470274658203125, 1.5474892578125, 1.547757568359375, 1.547947021484375, 1.5482930908203125, 1.549538330078125, 1.547505615234375, 1.548253173828125, 1.549297607421875, 1.547484130859375, 1.548220458984375, 1.548078125, 1.547826171875, 1.54688916015625, 1.5470919189453125, 1.5475538330078125, 1.54765625, 1.5476009521484375, 1.5491522216796876, 1.5480196533203125, 1.5481773681640625, 1.5486719970703124, 1.54946044921875, 3.19001904296875, 1.5467274169921874, 1.5485091552734376, 1.54712060546875, 1.5472425537109376, 1.548200927734375, 1.5474647216796875, 1.5484794921875, 1.5470428466796875, 1.5474134521484375, 1.5471728515625, 1.54741455078125, 1.547334716796875, 1.5476378173828125, 1.5483443603515625, 1.5476234130859374, 1.5473817138671875, 1.548241943359375, 1.548495849609375, 1.5476910400390624, 1.5492137451171875, 1.5486300048828125, 1.547925537109375, 1.5481497802734374, 1.5477288818359376, 1.547779052734375, 1.5494564208984376, 1.5475230712890624, 1.547504638671875, 1.5472803955078125, 1.5474217529296874, 1.54842724609375, 1.547715576171875, 1.54804736328125, 1.54720458984375, 1.5474442138671876, 1.5468114013671874, 1.548011474609375, 1.5478446044921874, 1.5482470703125, 1.548304443359375, 1.5482255859375, 1.549054931640625, 1.548583984375, 1.5483463134765625, 1.5490682373046876, 1.5485091552734376, 1.5480699462890626, 1.5479337158203126, 1.5475302734375, 1.5479091796875, 1.548273681640625, 1.5492208251953126, 1.547388916015625, 1.54747802734375, 1.5476285400390626, 1.54806787109375, 1.548291015625, 1.5489659423828126, 1.5477606201171874, 1.5477279052734374, 1.5477841796875, 1.5487047119140624, 3.189357666015625, 1.54832177734375, 1.54777294921875, 1.5475640869140626, 1.548190673828125, 1.54800439453125, 1.54889111328125, 1.548938232421875, 1.547832275390625, 1.5472568359375, 1.5471380615234376, 1.5473438720703125, 1.5473499755859375, 1.548291015625, 1.5484302978515625, 1.547525146484375, 1.54766748046875, 1.54760400390625, 1.54752001953125, 1.547726806640625, 1.54796435546875, 1.5488388671875, 1.5483873291015624, 1.54832177734375, 1.54785791015625, 1.548337158203125, 1.548981201171875, 1.5488040771484375, 1.5480023193359376, 1.547864013671875, 1.5482235107421876, 1.54768994140625, 1.5485205078125, 1.5483873291015624, 1.5479071044921875, 1.54815380859375, 1.5473756103515626, 1.5484302978515625, 1.5488572998046874, 1.54754248046875, 1.548105712890625, 1.5480648193359374, 1.548222412109375, 1.5480238037109375, 1.547357177734375, 1.5477821044921876, 1.5485142822265625, 1.54811181640625, 1.5482706298828126, 1.547442138671875, 1.547783203125, 1.548359619140625, 1.54857373046875, 1.5488572998046874, 1.5484447021484375, 1.5478978271484376, 1.5483546142578124, 1.5478446044921874, 1.54789990234375, 1.548568603515625, 1.5481129150390625, 1.5489515380859376, 1.5481129150390625, 3.18936376953125, 1.5473162841796875, 1.548347412109375, 1.5468809814453126, 1.5473695068359374, 1.5474237060546876, 1.5484447021484375, 1.548760009765625, 1.5489996337890626, 1.547483154296875, 1.547210693359375, 1.5470223388671875, 1.5469271240234375, 1.5475926513671876, 1.54830029296875, 1.5476080322265624, 1.547652099609375, 1.5482685546875, 1.547872314453125, 1.5475855712890625, 1.5479234619140625, 1.5485009765625, 1.54821630859375, 1.54749853515625, 1.54756298828125, 1.5474810791015625, 1.5483934326171875, 1.5483914794921876, 1.5475958251953126, 1.5481610107421875, 1.5478609619140624, 1.5482193603515626, 1.5480648193359374, 1.5481988525390624, 1.5481087646484375, 1.5474114990234376, 1.54781494140625, 1.5477237548828124, 1.5477821044921876, 1.5474083251953126, 1.5495556640625, 1.5478026123046875, 1.54785693359375, 1.547255859375, 1.54735302734375, 1.5477073974609374, 1.547969482421875, 1.5477288818359376, 1.5476070556640624, 1.5480606689453125, 1.548168212890625, 1.5479029541015625, 1.5480186767578126, 1.5485255126953126, 1.5468739013671875, 1.546857421875, 1.54753125, 1.5483514404296874, 1.54785888671875, 1.548099609375, 1.54840478515625, 1.54806884765625, 1.5480648193359374, 3.188536376953125, 1.547326416015625, 1.5486146240234375, 1.5495721435546874, 1.5473060302734376, 1.5487816162109376, 1.549453369140625, 1.5489608154296874, 1.5496785888671876, 1.5480616455078124, 1.54798486328125, 1.5475947265625, 1.549253662109375, 1.5481968994140625, 1.548053466796875, 1.549401123046875, 1.5481220703125, 1.547989990234375, 1.548347412109375, 1.5480863037109376, 1.548441650390625, 1.5491942138671875, 1.5479951171875, 1.5477596435546874, 1.5479132080078124, 1.5479019775390626, 1.5476695556640625, 1.5486505126953125, 1.548642333984375, 1.5483924560546876, 1.548464111328125, 1.5485399169921874, 1.5489935302734374, 1.5492823486328124, 1.549126708984375, 1.5488082275390624, 1.5485020751953125, 1.5481436767578125, 1.5480247802734375, 1.5473438720703125, 1.5484078369140626, 1.5497154541015625, 1.548389404296875, 1.548622802734375, 1.5483453369140625, 1.5479080810546875, 1.548316650390625, 1.548980224609375, 1.5480084228515625, 1.5479080810546875, 1.5483177490234374, 1.54800634765625, 1.5485972900390625, 1.5479234619140625, 1.5486607666015626, 1.548347412109375, 1.5482081298828125, 1.547925537109375, 1.548674072265625, 1.548291015625, 1.55034423828125, 1.54886865234375, 1.5481036376953126, 3.189357666015625, 1.5474586181640626, 1.5491717529296876, 1.5481129150390625, 1.548072998046875, 1.5476490478515625, 1.5485419921875, 1.5480074462890625, 1.54784765625, 1.5483238525390626, 1.547341796875, 1.547925537109375, 1.547810791015625, 1.5478056640625, 1.5477412109375, 1.54859521484375, 1.5481068115234375, 1.548200927734375, 1.547937744140625, 1.547672607421875, 1.547968505859375, 1.5481712646484376, 1.5496058349609374, 1.54743603515625, 1.5484375, 1.548396484375, 1.5478035888671875, 1.5489986572265626, 1.54951171875, 1.547946044921875, 1.5478077392578125, 1.5477493896484376, 1.5488133544921876, 1.5480872802734376, 1.5481773681640625, 1.54931103515625, 1.5479542236328125, 1.5490611572265625, 1.547883544921875, 1.547809814453125, 1.5479930419921875, 1.549116455078125, 1.5483084716796875, 1.54863720703125, 1.54834130859375, 1.548610595703125, 1.5488941650390624, 1.5497103271484376, 1.5490672607421876, 1.5479500732421876, 1.5488173828125, 1.5487139892578126, 1.5484384765625, 1.5487242431640624, 1.5489197998046875, 1.5499970703125, 1.5479736328125, 1.549170654296875, 1.548205078125, 1.5482030029296876, 1.54952294921875, 1.548652587890625, 1.5488470458984376, 3.190570068359375, 1.5475732421875, 1.54899560546875, 1.5490723876953125, 1.548262451171875, 1.5477442626953124, 1.5488890380859375, 1.5484302978515625, 1.5486300048828125, 1.548895263671875, 1.5487139892578126, 1.54817333984375, 1.548970947265625, 1.5481988525390624, 1.5474852294921875, 1.548536865234375, 1.5487886962890625, 1.548294189453125, 1.54817333984375, 1.547431884765625, 1.54737353515625, 1.5477801513671876, 1.5483740234375, 1.547725830078125, 1.547284423828125, 1.5479593505859375, 1.5474329833984375, 1.5482255859375, 1.54836376953125, 1.54859521484375, 1.5475343017578125, 1.5487181396484375, 1.548205078125, 1.5478056640625, 1.5484302978515625, 1.54993359375, 1.5491451416015625, 1.548178466796875, 1.5477718505859375, 1.5477421875, 1.5473060302734376, 1.548078125, 1.548304443359375, 1.5474052734375, 1.5479808349609374, 1.547030517578125, 1.547672607421875, 1.5480975341796874, 1.5481231689453125, 1.547429931640625, 1.5478343505859375, 1.5476173095703125, 1.5473643798828125, 1.547039794921875, 1.5472691650390624, 1.5484384765625, 1.5475506591796875, 1.548432373046875, 1.54777294921875, 1.5470867919921876, 1.54768896484375, 1.5496007080078125, 1.548421142578125, 3.189211181640625, 1.5468963623046874, 1.547869140625, 1.54876416015625, 1.5484989013671875, 1.5492301025390625, 1.5475916748046874, 1.546978271484375, 1.5469158935546874, 1.54728759765625, 1.5486658935546875, 1.5481077880859375, 1.5487529296875, 1.54709912109375, 1.547663330078125, 1.5471912841796875, 1.54923828125, 1.54927001953125, 1.54796435546875, 1.5475865478515625, 1.5476448974609376, 1.547410400390625, 1.548801025390625, 1.5483770751953125, 1.547400146484375, 1.5478026123046875, 1.54826953125, 1.547557861328125, 1.547953125, 1.549897705078125, 1.5482357177734376, 1.5481138916015624, 1.547904052734375, 1.5478404541015625, 1.5473480224609375, 1.54821630859375, 1.5495045166015624, 1.547509765625, 1.5476746826171874, 1.546893310546875, 1.54714013671875, 1.5480177001953126, 1.548304443359375, 1.547969482421875, 1.547404296875, 1.5485450439453126, 1.547778076171875, 1.5482039794921876, 1.549432861328125, 1.54890234375, 1.5484302978515625, 1.5484652099609375, 1.547947998046875, 1.5474166259765625, 1.547483154296875, 1.5496939697265626, 1.5482757568359375, 1.547537353515625, 1.5478927001953124, 1.54752001953125, 1.547925537109375, 1.5486146240234375, 1.5490406494140625]",tokens/s,0.6362950887053097,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-2.7b,facebook/opt-2.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,2109.853696,2844.2624,0.0,2197.815296,1927.351296,s,10,2.3800795135498047,0.23800795135498043,0.0010254107368052306,0.23830428314208985,0.23930529022216795,0.23934926986694335,0.23938445358276367,"[0.2383564453125, 0.23939324951171875, 0.23619731140136718, 0.23845465087890624, 0.23723452758789063, 0.23663900756835937, 0.23750480651855468, 0.2382521209716797, 0.2387518768310547, 0.23929551696777343]",tokens/s,1075.594317511624,kWh,2.793223439723023e-06,1.5305452026781454e-06,1.2162554949675677e-05,1.6486323592076845e-05,tokens/kWh,15528022.276781643,MB,2109.853696,2844.2624,0.0,2197.815296,2031.97184,s,10,139.4725029296875,13.94725029296875,0.016539028053013265,13.941937500000002,13.97250703125,13.977812109375,13.982056171875,"[13.953560546875, 13.945388671875, 13.94201171875, 13.9831171875, 13.9309609375, 13.9364580078125, 13.92993359375, 13.94186328125, 13.971328125, 13.937880859375]",tokens/s,4.517019389245513,kWh,0.000164597820511391,9.021296607492203e-05,0.0007178143901425277,0.0009726251767288406,tokens/kWh,64773.153633430826,,s,629,141.39150625610358,0.2247877682926924,0.028297149074530483,0.2211778564453125,0.22228438720703125,0.22277959289550783,0.45843198486328124,"[0.22260429382324218, 0.22169497680664063, 0.2212351989746094, 0.22115533447265626, 0.22176870727539064, 0.22137344360351563, 0.22118502807617188, 0.22141644287109374, 0.22102528381347655, 0.22195199584960937, 0.2219018249511719, 0.22150553894042968, 0.22119935607910157, 0.2213509063720703, 0.22125465393066407, 0.22120550537109376, 0.22150962829589843, 0.2213017578125, 0.22201548767089843, 0.2213632049560547, 0.22165196228027345, 0.22151373291015625, 0.22148095703125, 0.22154649353027345, 0.22147174072265624, 0.22180044555664064, 0.22122802734375, 0.22110105895996093, 0.22116249084472656, 0.221444091796875, 0.22163456726074218, 0.22287872314453125, 0.2216417236328125, 0.222202880859375, 0.2216407012939453, 0.22104165649414062, 0.22102323913574218, 0.22114508056640625, 0.22116044616699218, 0.22107955932617188, 0.2210508728027344, 0.22112460327148437, 0.22108979797363282, 0.2215557098388672, 0.22159564208984375, 0.22146662902832032, 0.2216785888671875, 0.22263194274902343, 0.22135398864746095, 0.22125056457519532, 0.22187930297851563, 0.22126797485351563, 0.2213396453857422, 0.22115225219726561, 0.22118502807617188, 0.22127410888671875, 0.22127104187011717, 0.22109901428222656, 0.22111744689941407, 0.22118911743164063, 0.2214871063232422, 0.22157005310058595, 0.46144613647460936, 0.22118707275390626, 0.2218403778076172, 0.2209945526123047, 0.22095564270019533, 0.22111538696289063, 0.22113792419433595, 0.22112973022460938, 0.22129766845703125, 0.22095974731445311, 0.22107244873046875, 0.2212720031738281, 0.22100991821289062, 0.22110617065429689, 0.22098739624023436, 0.22138470458984374, 0.22130073547363283, 0.22108773803710938, 0.22093721008300782, 0.2209566650390625, 0.2209310760498047, 0.2208204803466797, 0.22089932250976563, 0.22095155334472658, 0.22101708984375, 0.22186495971679687, 0.22131610107421876, 0.22127513122558592, 0.22142361450195314, 0.22218751525878908, 0.2213519287109375, 0.2209669189453125, 0.22220083618164063, 0.2210744323730469, 0.22159461975097655, 0.22113792419433595, 0.22117990112304686, 0.2222161865234375, 0.22108262634277343, 0.2210293731689453, 0.22102117919921874, 0.22107955932617188, 0.22112562561035157, 0.22216499328613282, 0.22140518188476563, 0.22183526611328125, 0.22283879089355468, 0.22166937255859376, 0.2215045166015625, 0.22186904907226562, 0.22122700500488282, 0.2214256591796875, 0.22139903259277344, 0.2215004119873047, 0.22137753295898438, 0.22115737915039063, 0.22106419372558594, 0.22163967895507813, 0.22111334228515625, 0.22137753295898438, 0.2212833251953125, 0.2219683837890625, 0.22127923583984374, 0.45813043212890625, 0.22103347778320312, 0.2213949432373047, 0.22112973022460938, 0.22099250793457031, 0.22118502807617188, 0.2218219451904297, 0.22100376892089843, 0.221154296875, 0.22091162109375, 0.22253260803222658, 0.22119935607910157, 0.2210529327392578, 0.22278041076660157, 0.22130482482910158, 0.22150553894042968, 0.2213519287109375, 0.22121983337402343, 0.2217584686279297, 0.22111436462402342, 0.22102015686035156, 0.22090444946289062, 0.22078054809570313, 0.2211778564453125, 0.22104576110839844, 0.2211420135498047, 0.22199909973144533, 0.22127308654785155, 0.22181170654296875, 0.2217164764404297, 0.2210918426513672, 0.2213201904296875, 0.22124339294433593, 0.2210150451660156, 0.22111231994628905, 0.22106008911132813, 0.22112051391601562, 0.22103245544433595, 0.22106419372558594, 0.2211778564453125, 0.22102323913574218, 0.22111846923828124, 0.22116146850585938, 0.2212884521484375, 0.22095872497558594, 0.22103347778320312, 0.221048828125, 0.2216294403076172, 0.22115122985839844, 0.22118400573730468, 0.221127685546875, 0.22102117919921874, 0.22113177490234376, 0.22128025817871094, 0.2211031036376953, 0.22143180847167968, 0.22129254150390626, 0.22108773803710938, 0.22222642517089844, 0.22198886108398438, 0.2217205810546875, 0.2217574462890625, 0.22159461975097655, 0.4614256591796875, 0.22283775329589844, 0.22295756530761718, 0.22274969482421875, 0.2225858612060547, 0.22346035766601563, 0.22301695251464843, 0.22341127014160156, 0.22270252990722655, 0.22250291442871092, 0.22281216430664064, 0.22286746215820313, 0.22281011962890626, 0.22280703735351562, 0.22268620300292968, 0.22271487426757813, 0.22270976257324218, 0.22246092224121095, 0.222814208984375, 0.22279373168945313, 0.22285516357421875, 0.222635009765625, 0.2227783660888672, 0.2227271728515625, 0.22253773498535157, 0.22261862182617187, 0.223025146484375, 0.22222540283203124, 0.22242611694335937, 0.22123930358886718, 0.22103450012207032, 0.2209105987548828, 0.22106008911132813, 0.22086451721191405, 0.2209740753173828, 0.22110208129882813, 0.2218260498046875, 0.2210365447998047, 0.2210713653564453, 0.22134988403320313, 0.22149017333984375, 0.22106419372558594, 0.22114309692382814, 0.22111225891113281, 0.2225971221923828, 0.22118400573730468, 0.22139187622070314, 0.22111538696289063, 0.22112153625488282, 0.22092594909667967, 0.22134783935546876, 0.221127685546875, 0.22192127990722657, 0.2215854034423828, 0.22127308654785155, 0.22121881103515625, 0.2214686737060547, 0.2213396453857422, 0.22123008728027344, 0.22115327453613282, 0.22119218444824218, 0.22103347778320312, 0.22114303588867187, 0.45854925537109376, 0.22104678344726564, 0.22133247375488282, 0.22091571044921876, 0.22102630615234375, 0.2210508728027344, 0.22108364868164063, 0.22100274658203126, 0.22103141784667968, 0.22081741333007812, 0.22105906677246093, 0.22115737915039063, 0.22105702209472655, 0.22100991821289062, 0.22107749938964844, 0.22084402465820313, 0.22094540405273438, 0.22115327453613282, 0.22110719299316406, 0.220980224609375, 0.22154444885253907, 0.22112460327148437, 0.2211031036376953, 0.22097817993164062, 0.2210498504638672, 0.22134375, 0.2211266632080078, 0.22089112854003906, 0.2209566650390625, 0.22103756713867187, 0.2209976348876953, 0.22094540405273438, 0.22101708984375, 0.2214256591796875, 0.22106623840332032, 0.22102528381347655, 0.2213017578125, 0.2209976348876953, 0.221233154296875, 0.22100274658203126, 0.22102220153808594, 0.22098329162597657, 0.2211584014892578, 0.2211092529296875, 0.22123417663574219, 0.22111744689941407, 0.2211235809326172, 0.22127001953125, 0.22116761779785157, 0.22121881103515625, 0.2211031036376953, 0.2210498504638672, 0.221085693359375, 0.22113587951660157, 0.22125978088378906, 0.2211041259765625, 0.221154296875, 0.22228172302246094, 0.22109901428222656, 0.22121267700195313, 0.22124339294433593, 0.2211584014892578, 0.2214686737060547, 0.457933837890625, 0.2211962890625, 0.22194586181640624, 0.22120755004882814, 0.22100889587402345, 0.2212843475341797, 0.22189056396484375, 0.2212884521484375, 0.2211727294921875, 0.22092083740234375, 0.22090956115722657, 0.2211768341064453, 0.2210160675048828, 0.22107545471191406, 0.22154035949707032, 0.22111436462402342, 0.22106521606445312, 0.22105599975585938, 0.22096998596191406, 0.22127615356445313, 0.22098329162597657, 0.22110617065429689, 0.2209003448486328, 0.22102117919921874, 0.22100376892089843, 0.2210365447998047, 0.22115635681152343, 0.2209239044189453, 0.22113996887207032, 0.22109490966796874, 0.22142259216308594, 0.22121983337402343, 0.22105599975585938, 0.2209105987548828, 0.221159423828125, 0.22130482482910158, 0.22137651062011718, 0.22106521606445312, 0.22213119506835938, 0.22088088989257812, 0.2209105987548828, 0.2210150451660156, 0.22098739624023436, 0.22129458618164063, 0.22125363159179687, 0.22113894653320312, 0.22119833374023437, 0.2218956756591797, 0.22099250793457031, 0.22140518188476563, 0.22130892944335936, 0.22184857177734374, 0.22106930541992187, 0.22177484130859376, 0.22129254150390626, 0.22115635681152343, 0.22096485900878907, 0.22116864013671875, 0.22148915100097658, 0.22126797485351563, 0.221384765625, 0.2214911346435547, 0.22094540405273438, 0.45877044677734374, 0.2209187774658203, 0.22111744689941407, 0.22087065124511718, 0.22097509765625, 0.22129971313476562, 0.2210744323730469, 0.2212843475341797, 0.22100991821289062, 0.22094744873046876, 0.2208368682861328, 0.22104473876953126, 0.22097305297851563, 0.2210682830810547, 0.2209914855957031, 0.2208757781982422, 0.2209628143310547, 0.2210426940917969, 0.22097509765625, 0.22097203063964843, 0.22091468811035156, 0.22095872497558594, 0.22083993530273438, 0.22126591491699218, 0.2211420135498047, 0.22217318725585938, 0.22125978088378906, 0.22093618774414062, 0.2212351989746094, 0.22171852111816406, 0.22144717407226563, 0.22102323913574218, 0.221486083984375, 0.220980224609375, 0.2209935302734375, 0.2213939208984375, 0.22103858947753907, 0.22095462036132812, 0.22107647705078126, 0.22094744873046876, 0.22103858947753907, 0.22127206420898438, 0.22109286499023437, 0.2211461181640625, 0.22218240356445312, 0.22110719299316406, 0.2209812469482422, 0.22119833374023437, 0.22108876037597655, 0.22112870788574218, 0.22101913452148436, 0.22104678344726564, 0.2209669189453125, 0.22112870788574218, 0.2210682830810547, 0.22112051391601562, 0.22107034301757814, 0.22104473876953126, 0.22089421081542968, 0.221011962890625, 0.2210160675048828, 0.2210846710205078, 0.2209628143310547, 0.45941656494140626, 0.2209495086669922, 0.221739013671875, 0.2209863739013672, 0.22102117919921874, 0.221048828125, 0.22226022338867188, 0.22110617065429689, 0.2219622344970703, 0.22180557250976562, 0.22233804321289063, 0.2211041259765625, 0.22118911743164063, 0.22098329162597657, 0.2222192687988281, 0.22136012268066407, 0.22214041137695312, 0.2213079071044922, 0.22128947448730468, 0.22100787353515625, 0.221127685546875, 0.22152703857421874, 0.2220943298339844, 0.2211031036376953, 0.22113792419433595, 0.22100889587402345, 0.22107034301757814, 0.22086758422851563, 0.2210048065185547, 0.22096588134765624, 0.2210846710205078, 0.22087680053710937, 0.221154296875, 0.22088607788085937, 0.22132730102539064, 0.2210529327392578, 0.22137548828125, 0.22103450012207032, 0.22103450012207032, 0.22075392150878906, 0.22111949157714844, 0.2211420135498047, 0.2211461181640625, 0.2211041259765625, 0.22117990112304686, 0.2209812469482422, 0.22081843566894532, 0.22104371643066406, 0.22111949157714844, 0.22109286499023437, 0.22118502807617188, 0.22213119506835938, 0.22129049682617188, 0.22118092346191406, 0.22175027465820313, 0.2221670379638672, 0.22145330810546876, 0.221517822265625, 0.22109490966796874, 0.22186087036132812, 0.2214615020751953, 0.22115122985839844, 0.22144102478027344, 0.46024600219726564, 0.2209976348876953, 0.22151475524902345, 0.22144717407226563, 0.2214246368408203, 0.22156083679199218, 0.2218076171875, 0.22113690185546875, 0.2215905303955078, 0.22111949157714844, 0.2214993896484375, 0.22210354614257813, 0.221412353515625, 0.22127104187011717, 0.22124339294433593, 0.22104165649414062, 0.22167552185058595, 0.22163148498535157, 0.2216048583984375, 0.22217523193359376, 0.22170419311523437, 0.22118092346191406, 0.22107034301757814, 0.22108876037597655, 0.22134988403320313, 0.22161715698242188, 0.22139698791503906, 0.22093209838867187, 0.22118707275390626, 0.22141542053222657, 0.22160383605957032, 0.2212833251953125, 0.22284288024902343, 0.2227220458984375, 0.22272000122070312, 0.22289511108398438, 0.22466764831542968, 0.2222950439453125, 0.22271078491210938, 0.22245887756347657, 0.22309580993652345, 0.2229698486328125, 0.22274867248535157, 0.22293504333496095, 0.22252748107910156, 0.22249267578125, 0.22271795654296875, 0.22294834899902344, 0.2227640380859375, 0.22101913452148436, 0.2211420135498047, 0.22163967895507813, 0.22111949157714844, 0.22116761779785157, 0.2211420135498047, 0.22103347778320312, 0.22134066772460936, 0.22165606689453124, 0.22108979797363282, 0.22110823059082033, 0.2213621826171875, 0.22120448303222656, 0.22122802734375, 0.46055218505859374, 0.22127308654785155, 0.22119218444824218, 0.22088088989257812, 0.22082456970214845, 0.22128128051757812, 0.22141439819335937, 0.22119935607910157, 0.22096383666992186, 0.22105804443359375, 0.22094137573242187, 0.2210590057373047, 0.22101094055175782, 0.22181581115722657, 0.22244248962402344, 0.22087271118164062, 0.22137753295898438, 0.22102732849121093, 0.2222520294189453, 0.22111129760742188, 0.22082662963867186, 0.22154853820800782, 0.22087986755371095, 0.2212034606933594, 0.22097920227050782, 0.22116659545898437, 0.22161100769042968, 0.22113075256347656, 0.2215413818359375, 0.22253465270996095, 0.22111949157714844, 0.2211102752685547, 0.2211788787841797, 0.22141952514648439, 0.22129766845703125, 0.22112562561035157, 0.22104063415527345, 0.22099046325683594, 0.22102835083007813, 0.22090444946289062, 0.22094438171386718, 0.2210181121826172, 0.22109907531738282, 0.22128121948242188, 0.22107647705078126, 0.22143795776367187, 0.2210846710205078, 0.22113792419433595, 0.22100274658203126, 0.22100889587402345, 0.2211266632080078, 0.22113587951660157, 0.2213580780029297, 0.2212833251953125, 0.221154296875, 0.22113587951660157, 0.2210672607421875, 0.22181581115722657, 0.22194586181640624, 0.22132325744628906, 0.22108572387695313, 0.22134576416015625, 0.22117170715332032]",tokens/s,4.448640633764008,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-125m,facebook/opt-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,1207.41888,879.230976,0.0,232.783872,169.719808,s,10,0.3329092750549316,0.03329092750549316,0.001090057705435725,0.0329870719909668,0.033882015228271486,0.0350985279083252,0.036071738052368164,"[0.03631504058837891, 0.03281526565551758, 0.03245974349975586, 0.03361167907714844, 0.03344601440429688, 0.033440895080566406, 0.032671966552734376, 0.033158878326416015, 0.03245904159545898, 0.0325307502746582]",tokens/s,7689.782748100329,kWh,3.9167167718924527e-07,2.1461758953218489e-07,8.304311279209363e-07,1.4367203946423665e-06,tokens/kWh,178183591.57052574,MB,1207.41888,879.230976,0.0,232.783872,199.792128,s,10,20.506330566406252,2.050633056640625,0.01795433438952847,2.0542513427734375,2.07115517578125,2.0720841796875,2.0728273828125,"[2.05074072265625, 2.0363714599609377, 2.0684541015625, 2.054240966796875, 2.07094873046875, 2.07301318359375, 2.05426171875, 2.0559150390625, 2.0203577880859376, 2.02202685546875]",tokens/s,30.72222004613905,kWh,2.4533342500768937e-05,1.3444361220557862e-05,5.009960961606961e-05,8.80773133373964e-05,tokens/kWh,715280.6734540922,,s,629,20.768719875335666,0.03301863255220301,0.003999927492502233,0.03273830413818359,0.033139096832275394,0.033337548065185546,0.06495629302978514,"[0.033740798950195314, 0.034253822326660154, 0.0341104621887207, 0.03412275314331055, 0.0335206413269043, 0.03323699188232422, 0.03351859283447266, 0.03322982406616211, 0.033159168243408206, 0.03280998229980469, 0.03333222579956055, 0.033258495330810545, 0.032756736755371094, 0.03297075271606445, 0.03364352035522461, 0.03362918472290039, 0.033584129333496096, 0.03300249481201172, 0.03202867126464844, 0.031971328735351565, 0.031821823120117186, 0.031848447799682614, 0.03192934417724609, 0.03202560043334961, 0.03206041717529297, 0.032089088439941404, 0.03197644805908203, 0.03203379058837891, 0.03201638412475586, 0.031941631317138675, 0.03214438247680664, 0.03217407989501953, 0.03201638412475586, 0.031987712860107424, 0.03198054313659668, 0.03194572830200195, 0.03189145660400391, 0.03199590492248535, 0.03189043235778809, 0.03177676773071289, 0.032366592407226565, 0.03329228973388672, 0.0331960334777832, 0.0321341438293457, 0.032966655731201173, 0.03283865737915039, 0.03283456039428711, 0.032866302490234374, 0.032574462890625, 0.03194470405578613, 0.032132095336914065, 0.032039936065673826, 0.03172454452514648, 0.03170099258422852, 0.03186278343200684, 0.03212492752075195, 0.031920127868652344, 0.032054271697998044, 0.03213926315307617, 0.03209830474853516, 0.03199795150756836, 0.03198873519897461, 0.06604799652099609, 0.031955968856811526, 0.03163852882385254, 0.03178700828552246, 0.03185971260070801, 0.03197337532043457, 0.03201945495605469, 0.031987712860107424, 0.032059391021728514, 0.03199795150756836, 0.03196211242675781, 0.03191500854492187, 0.03182694435119629, 0.031764480590820314, 0.031714303970336914, 0.0317573127746582, 0.031927295684814457, 0.03189145660400391, 0.03197235107421875, 0.03200921630859375, 0.03191500854492187, 0.03215462493896484, 0.03199795150756836, 0.03201126480102539, 0.031936511993408204, 0.03203071975708008, 0.03208294296264649, 0.031936511993408204, 0.032161792755126956, 0.03201331329345703, 0.031562751770019534, 0.03183923149108887, 0.0325928955078125, 0.03195187187194824, 0.03276287841796875, 0.03203276824951172, 0.03358924865722656, 0.032881664276123046, 0.033119232177734374, 0.03294617462158203, 0.03283251190185547, 0.03209011077880859, 0.032072704315185545, 0.03290934371948242, 0.03280073547363281, 0.032763904571533206, 0.03282329559326172, 0.03290419387817383, 0.032740352630615234, 0.032478206634521486, 0.03298611068725586, 0.03275059127807617, 0.033102848052978515, 0.03289395141601562, 0.03265331268310547, 0.032115745544433597, 0.03197641563415527, 0.03259494400024414, 0.03297894287109375, 0.03300044631958008, 0.03283865737915039, 0.03288780975341797, 0.03242803192138672, 0.06496460723876953, 0.03309568023681641, 0.03194470405578613, 0.03273830413818359, 0.03288678359985352, 0.03281919860839844, 0.03297689437866211, 0.032912384033203124, 0.032787487030029296, 0.033037311553955076, 0.03294307327270508, 0.032863231658935545, 0.03278847885131836, 0.03306496047973633, 0.032745471954345705, 0.03307212829589844, 0.03291852951049805, 0.03285094451904297, 0.03298406219482422, 0.03316223907470703, 0.03301683044433594, 0.0328611831665039, 0.03299532699584961, 0.03296051025390625, 0.03287449645996094, 0.0329615364074707, 0.03278745651245117, 0.032949249267578126, 0.033023998260498046, 0.03287449645996094, 0.032909313201904294, 0.03280179214477539, 0.0328089599609375, 0.03198464012145996, 0.032054271697998044, 0.03240140914916992, 0.03320729446411133, 0.033165313720703124, 0.03290726470947265, 0.032584705352783204, 0.032707584381103515, 0.032249855041503905, 0.03290521621704102, 0.0328458251953125, 0.032791553497314455, 0.03254579162597656, 0.03258879852294922, 0.03239424133300781, 0.03282227325439453, 0.032894977569580076, 0.03285811233520508, 0.03289708709716797, 0.032930816650390625, 0.03297683334350586, 0.033140735626220705, 0.032868350982666016, 0.03292364883422851, 0.032985088348388675, 0.033018878936767575, 0.0329431037902832, 0.032927745819091796, 0.032976993560791014, 0.03305363082885742, 0.06707913970947266, 0.03323494338989258, 0.03279257583618164, 0.033137664794921876, 0.0331776008605957, 0.03303014373779297, 0.033274879455566404, 0.0330967025756836, 0.033137664794921876, 0.03282534408569336, 0.03309056091308594, 0.033081344604492184, 0.03303424072265625, 0.033309696197509765, 0.03319500732421875, 0.03301171112060547, 0.033345535278320314, 0.03285094451904297, 0.03292876815795898, 0.03282124710083008, 0.03200307083129883, 0.0319866886138916, 0.032074752807617186, 0.031974399566650394, 0.03197644805908203, 0.03283967971801758, 0.03339263916015625, 0.033058815002441407, 0.032753662109375, 0.03252326583862305, 0.03210553741455078, 0.031921152114868165, 0.03210540771484375, 0.0323768310546875, 0.03219968032836914, 0.0321976318359375, 0.03212799835205078, 0.03213516616821289, 0.03212799835205078, 0.03211775970458984, 0.032985088348388675, 0.034108417510986325, 0.033040382385253905, 0.03294617462158203, 0.03294105529785156, 0.032906238555908206, 0.032069633483886716, 0.031744064331054686, 0.03175724792480469, 0.031899648666381834, 0.03206655883789063, 0.03191296005249023, 0.03192934417724609, 0.032126976013183595, 0.03269740676879883, 0.03293075180053711, 0.03260313415527344, 0.03196211242675781, 0.031904767990112305, 0.03212287902832031, 0.03203788757324219, 0.03282944107055664, 0.03291033554077148, 0.06675456237792969, 0.033050624847412106, 0.0328007698059082, 0.03285299301147461, 0.0328089599609375, 0.03288883209228516, 0.033067008972167966, 0.03282022476196289, 0.03274649429321289, 0.03173785591125488, 0.032020481109619144, 0.03192428779602051, 0.03253753662109375, 0.03285606384277344, 0.033326080322265625, 0.03281919860839844, 0.033073150634765625, 0.03273830413818359, 0.03295129776000977, 0.03292364883422851, 0.03303628921508789, 0.03282841491699219, 0.032919551849365236, 0.03291961669921875, 0.032923583984375, 0.032797695159912106, 0.03295948791503906, 0.03298303985595703, 0.032982078552246094, 0.03352467346191406, 0.032589824676513675, 0.033331199645996096, 0.03296255874633789, 0.032672767639160154, 0.032740352630615234, 0.03300864028930664, 0.03226521682739258, 0.03242803192138672, 0.03284275054931641, 0.03410636901855469, 0.03410943984985351, 0.032530433654785154, 0.032048126220703126, 0.032909313201904294, 0.03273625564575195, 0.03282022476196289, 0.033040382385253905, 0.03312844848632813, 0.033032222747802736, 0.03294512176513672, 0.032824321746826174, 0.032772159576416014, 0.032817089080810546, 0.03304959869384766, 0.032717823028564456, 0.0328007698059082, 0.03302195358276367, 0.03298713684082031, 0.03286937713623047, 0.032884735107421875, 0.032917503356933595, 0.03312947082519531, 0.0328243522644043, 0.06682825469970703, 0.032939006805419925, 0.03280998229980469, 0.0324136962890625, 0.032930816650390625, 0.03313868713378906, 0.03297382354736328, 0.032863231658935545, 0.03299327850341797, 0.03307212829589844, 0.03317145538330078, 0.033142784118652346, 0.032942081451416014, 0.033363967895507815, 0.03327897644042969, 0.03276595306396484, 0.03296255874633789, 0.033056766510009765, 0.03283967971801758, 0.03291648101806641, 0.03299225616455078, 0.032946239471435546, 0.032924606323242185, 0.03313868713378906, 0.03290521621704102, 0.03295641708374023, 0.03341516876220703, 0.0333199348449707, 0.03459379196166992, 0.03333631896972656, 0.032895999908447264, 0.03294412612915039, 0.03291648101806641, 0.03218227386474609, 0.032626686096191404, 0.0323061752319336, 0.03225804901123047, 0.03283148956298828, 0.033329151153564454, 0.03292879867553711, 0.03304956817626953, 0.03294822311401367, 0.03261030578613281, 0.03294105529785156, 0.032903167724609376, 0.03311206436157227, 0.03312025451660156, 0.03289190292358399, 0.03233280181884766, 0.032524288177490236, 0.03305779266357422, 0.03231129455566406, 0.0324505615234375, 0.03278950500488281, 0.03302707290649414, 0.03288678359985352, 0.032996353149414064, 0.03286937713623047, 0.03303014373779297, 0.03281203079223633, 0.03252633666992188, 0.032178176879882815, 0.03199084854125977, 0.065176513671875, 0.03205017471313477, 0.03209011077880859, 0.03212287902832031, 0.03207987213134766, 0.03216588973999023, 0.03170816040039062, 0.03165286445617676, 0.03220684814453125, 0.03201228713989258, 0.03200614547729492, 0.03260927963256836, 0.03219046401977539, 0.03249868774414062, 0.03227340698242188, 0.03234201431274414, 0.03262464141845703, 0.03326873779296875, 0.03293286514282227, 0.032851966857910156, 0.032830463409423825, 0.03336908721923828, 0.03311824035644531, 0.03295331192016602, 0.032176128387451174, 0.03207680130004883, 0.032168991088867185, 0.032793567657470706, 0.03295641708374023, 0.032803871154785155, 0.03245257568359375, 0.03248025512695313, 0.03281612777709961, 0.03295129776000977, 0.033175552368164066, 0.03283359909057617, 0.03288671875, 0.03277721786499024, 0.033051647186279294, 0.033258495330810545, 0.03284275054931641, 0.03292364883422851, 0.03303014373779297, 0.03285504150390625, 0.03174399948120117, 0.03240857696533203, 0.033037311553955076, 0.03285299301147461, 0.032894977569580076, 0.032917503356933595, 0.0321341438293457, 0.03308031845092774, 0.032146430969238284, 0.03266559982299805, 0.03310899353027344, 0.032740352630615234, 0.03243212890625, 0.03290828704833984, 0.032982017517089846, 0.032979969024658204, 0.032942081451416014, 0.032345088958740234, 0.03220172882080078, 0.06660710144042968, 0.032917537689208985, 0.032905185699462894, 0.032873470306396486, 0.033800193786621094, 0.0329615364074707, 0.03261542510986328, 0.03257753753662109, 0.0329881591796875, 0.03291545486450195, 0.032912384033203124, 0.032797695159912106, 0.03293183898925781, 0.033181697845458984, 0.03299020767211914, 0.03279564666748047, 0.03290726470947265, 0.03288883209228516, 0.03297177505493164, 0.03243929672241211, 0.032846847534179685, 0.03225600051879883, 0.03281510543823242, 0.033083393096923826, 0.0330332145690918, 0.032736320495605466, 0.032817089080810546, 0.03186278343200684, 0.03185766410827637, 0.03318483352661133, 0.03217606353759766, 0.03176038360595703, 0.032323585510253904, 0.03279052734375, 0.0322949104309082, 0.032471038818359374, 0.03288780975341797, 0.03274140930175781, 0.03304035186767578, 0.03282841491699219, 0.031971328735351565, 0.03230003356933594, 0.03287551879882813, 0.032753662109375, 0.03289907073974609, 0.03249356842041016, 0.03194367980957031, 0.0321710090637207, 0.03300352096557617, 0.03274649429321289, 0.0328540153503418, 0.03268710327148437, 0.03202764892578125, 0.032950271606445314, 0.03294515228271484, 0.03292671966552734, 0.03265740966796875, 0.03194675254821777, 0.03184127998352051, 0.03206860733032227, 0.03196108818054199, 0.03163750457763672, 0.031665151596069335, 0.0648058853149414, 0.03185663986206055, 0.031904767990112305, 0.031893503189086916, 0.032007167816162106, 0.0315729923248291, 0.031927295684814457, 0.03198259162902832, 0.0319682559967041, 0.03189967918395996, 0.03205014419555664, 0.031955968856811526, 0.03200614547729492, 0.03189657592773437, 0.03188121604919433, 0.03202150344848633, 0.0321638412475586, 0.031991840362548825, 0.03199894332885742, 0.03193343925476074, 0.03197542381286621, 0.0318525447845459, 0.03199078369140625, 0.03202252960205078, 0.03216281509399414, 0.03228876876831055, 0.032105472564697264, 0.0329697265625, 0.03331584167480469, 0.031987712860107424, 0.03201638412475586, 0.03187820816040039, 0.03171219253540039, 0.032074752807617186, 0.03211372756958008, 0.03188115119934082, 0.031936511993408204, 0.032328704833984374, 0.032024574279785153, 0.03211980819702148, 0.031898624420166014, 0.03198566436767578, 0.03184332847595215, 0.032132095336914065, 0.031905792236328126, 0.031921152114868165, 0.03194777679443359, 0.03189145660400391, 0.03162112045288086, 0.03145011138916016, 0.03202867126464844, 0.03241164779663086, 0.03273932647705078, 0.03275980758666992, 0.0327086067199707, 0.032747520446777346, 0.032508926391601564, 0.031954944610595705, 0.03173785591125488, 0.03213926315307617, 0.03191910362243652, 0.032194561004638675, 0.03189452743530274, 0.0649349136352539, 0.032023551940917966, 0.03198566436767578, 0.03202560043334961, 0.03201126480102539, 0.03183616065979004, 0.031854591369628905, 0.0317573127746582, 0.031936511993408204, 0.031903743743896484, 0.03194675254821777, 0.03200511932373047, 0.031974399566650394, 0.03211161422729492, 0.031942655563354495, 0.0321638412475586, 0.03193446350097656, 0.03189145660400391, 0.03212799835205078, 0.032, 0.03198361587524414, 0.031937536239624024, 0.032105472564697264, 0.03191193580627441, 0.031916032791137694, 0.03271680068969727, 0.0333383674621582, 0.03406950378417969, 0.03317657470703125, 0.03253247833251953, 0.03198975944519043, 0.03277926254272461, 0.03240140914916992, 0.03290521621704102, 0.03309465789794922, 0.032075775146484374, 0.03243110275268555, 0.03205017471313477, 0.03197747230529785, 0.03197235107421875, 0.031908863067626955, 0.03199283218383789, 0.03201126480102539, 0.03172659111022949, 0.03145113563537598, 0.031459327697753905, 0.031643648147583005, 0.03180441665649414, 0.032189441680908204, 0.031916032791137694, 0.031955968856811526, 0.031971328735351565, 0.03196723175048828, 0.03189657592773437, 0.031959039688110355, 0.03171327972412109, 0.031732736587524416, 0.03177369689941406, 0.032056320190429685, 0.03193343925476074, 0.03195084762573242, 0.03191910362243652, 0.03195084762573242]",tokens/s,30.28593017651422,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/gemma-7b,google/gemma-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/gemma-7b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aaa23-5541cdc42e463ac8623afa38;c5f6519e-d1d1-4373-8dd2-26816c758a44) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/gemma-7b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/gemma-7b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-7B,Qwen/Qwen-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 550, in from_pretrained + model_class = get_class_from_dynamic_module( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 498, in get_class_from_dynamic_module + final_module = get_cached_module_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 323, in get_cached_module_file + modules_needed = check_imports(resolved_module_file) + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 181, in check_imports + raise ImportError( +ImportError: This modeling file requires the following packages that were not found in your environment: transformers_stream_generator. Run `pip install transformers_stream_generator` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-72B,Qwen/Qwen1.5-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 96.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciLM-7B,,cuda,0,42,,,,,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,,MB,1662.808064,5516.034048,0.0,4869.586944,4743.593472,s,10,6.1237835083007806,0.6123783508300781,0.001253925145814982,0.6120383911132812,0.6140244384765625,0.6141366088867187,0.6142263452148438,"[0.6138787841796876, 0.614248779296875, 0.6107066650390625, 0.61129736328125, 0.6111465454101562, 0.6114146728515625, 0.6116507568359375, 0.612426025390625, 0.61399951171875, 0.613014404296875]",tokens/s,418.0422114089963,kWh,7.21943411562178e-06,3.95425479798766e-06,3.477342651153299e-05,4.594711542514243e-05,tokens/kWh,5571622.889299289,MB,1662.808064,5516.034048,0.0,4869.586944,4769.651712,s,10,360.87087890624997,36.087087890625,0.01748383820965051,36.082986328125,36.11191953125,36.119573046875,36.125695859375,"[36.11021875, 36.1272265625, 36.08849609375, 36.07920703125, 36.0845625, 36.08481640625, 36.08141015625, 36.0650625, 36.07158203125, 36.078296875]",tokens/s,1.7457767773045123,kWh,0.00042595356252458363,0.00023346088450391426,0.0019820382702698735,0.0026414527172983716,tokens/kWh,23850.51210170259,,s,629,365.80599053955063,0.5815675525271077,0.07276675093795772,0.5726207885742187,0.5739550537109376,0.5744299926757812,1.1846730029296875,"[0.5736171264648438, 0.573517822265625, 0.5738854370117188, 0.57415576171875, 0.5742704467773437, 0.5731041259765625, 0.57312255859375, 0.5737461547851562, 0.5739632568359375, 0.5725850219726563, 0.5728665161132812, 0.5735147705078125, 0.5726494750976563, 0.573000732421875, 0.57260546875, 0.5726546020507812, 0.572073974609375, 0.5728809204101563, 0.572885986328125, 0.5725767822265625, 0.5731676025390625, 0.5745838012695312, 0.5729105834960937, 0.5730211791992188, 0.5729566650390625, 0.5723678588867187, 0.5724866333007812, 0.5726197509765625, 0.5735526123046875, 0.5740676879882812, 0.5729720458984375, 0.5732095947265625, 0.5726760864257813, 0.5724497680664062, 0.5724518432617187, 0.572790771484375, 0.5723668823242187, 0.5727159423828125, 0.5731563720703124, 0.5741424560546875, 0.5723576049804687, 0.5724334106445312, 0.5723883666992188, 0.5724548950195313, 0.5725368041992187, 0.5722142944335937, 0.5739407348632812, 0.5740206298828125, 0.572927001953125, 0.573212646484375, 0.5730181274414062, 0.5728184204101563, 0.5730816040039063, 0.5726177368164063, 0.572484619140625, 0.5727293701171875, 0.5740543823242188, 0.5732669677734376, 0.57483056640625, 0.574424072265625, 0.5744517211914062, 0.5744885864257813, 1.18734130859375, 0.5729505004882812, 0.5725030517578125, 0.5725654907226563, 0.5738137817382812, 0.572379150390625, 0.5741465454101562, 0.574750732421875, 0.573475830078125, 0.5734666137695312, 0.5736826782226563, 0.5750292358398438, 0.5742151489257813, 0.5739274291992188, 0.5746913452148438, 0.5744281616210938, 0.5742459106445312, 0.5746903076171875, 0.5750845336914062, 0.5739735107421875, 0.5736888427734375, 0.574476318359375, 0.5740001220703125, 0.573032470703125, 0.5722992553710937, 0.57221630859375, 0.571994140625, 0.573065185546875, 0.573844482421875, 0.5739673461914062, 0.5742745361328125, 0.5743206787109375, 0.5743236694335937, 0.5738577880859375, 0.5742110595703125, 0.5753764038085938, 0.5745797119140625, 0.57402880859375, 0.5725982666015625, 0.5732260131835938, 0.5737594604492188, 0.5739530029296875, 0.5731553344726562, 0.572927978515625, 0.57257470703125, 0.5744312133789062, 0.5755975952148438, 0.5727283325195313, 0.5721149291992188, 0.572199951171875, 0.5723617553710938, 0.572105712890625, 0.57270068359375, 0.5733294067382813, 0.5723873291015625, 0.57191015625, 0.573065185546875, 0.5727866821289063, 0.5724027099609375, 0.5726883544921875, 0.5723258666992187, 0.5720657958984375, 0.5731727294921874, 1.1836395263671875, 0.5721917724609376, 0.5728533325195313, 0.5732484741210937, 0.5725409545898438, 0.5732628784179687, 0.5741773071289062, 0.5730109252929687, 0.5731737670898438, 0.573043701171875, 0.5734686889648437, 0.5732505493164063, 0.5726914672851563, 0.5729924926757812, 0.5729525756835937, 0.5732260131835938, 0.574719970703125, 0.5727836303710937, 0.5730283813476562, 0.5724508056640625, 0.5733519287109375, 0.5728767700195313, 0.5725736694335938, 0.5743431396484375, 0.5745357055664062, 0.573507568359375, 0.572695556640625, 0.5729177856445312, 0.573022216796875, 0.5733980102539062, 0.5731030883789062, 0.5730037841796874, 0.572927978515625, 0.5738895263671875, 0.573137939453125, 0.57310107421875, 0.5718549194335938, 0.5721885986328125, 0.57208935546875, 0.5725255737304688, 0.5722398681640625, 0.5733836669921875, 0.5725736694335938, 0.57238525390625, 0.5722880249023438, 0.5720064086914063, 0.5721456909179687, 0.57385986328125, 0.5721773681640625, 0.5720176391601562, 0.5720484008789063, 0.5730570068359375, 0.5720934448242188, 0.5720852661132813, 0.5719961547851562, 0.5719992065429688, 0.5727928466796876, 0.5721221313476562, 0.5721630859375, 0.572600341796875, 0.5723125610351563, 0.5724631958007812, 0.5727794799804687, 1.1843287353515626, 0.572779541015625, 0.57261669921875, 0.5734041748046875, 0.5739274291992188, 0.57350146484375, 0.57328125, 0.5731553344726562, 0.5729392700195313, 0.5729403076171875, 0.5732833251953126, 0.57289013671875, 0.5743707885742187, 0.5724887084960938, 0.5720043334960937, 0.5718681640625, 0.5722941284179688, 0.5721036987304687, 0.5721763916015625, 0.57200537109375, 0.5717247924804687, 0.572310546875, 0.5729740600585937, 0.5724682006835937, 0.5719664916992188, 0.5723975830078125, 0.5722286376953125, 0.5723289794921875, 0.572221435546875, 0.5728389282226563, 0.5737000732421875, 0.5728737182617187, 0.5730037841796874, 0.572178466796875, 0.5720411376953125, 0.5717974853515625, 0.5721036987304687, 0.5720719604492187, 0.5719695434570312, 0.5730037841796874, 0.5722941284179688, 0.5718763427734375, 0.5725706176757812, 0.5727713012695312, 0.5752719116210937, 0.5729822998046875, 0.5730355224609375, 0.5741107177734375, 0.5727109375, 0.5726207885742187, 0.5727815551757812, 0.5729136352539063, 0.572896240234375, 0.572600341796875, 0.5730191650390625, 0.5724456787109375, 0.573222900390625, 0.57379736328125, 0.5722828979492187, 0.572052490234375, 0.5719766845703125, 0.5719429321289062, 0.5717380981445312, 1.1848990478515624, 0.5739724731445313, 0.5733632202148438, 0.5733693237304688, 0.5734880981445313, 0.5738117065429688, 0.5729403076171875, 0.5730672607421875, 0.5729863891601562, 0.5730140380859375, 0.57394482421875, 0.5726064453125, 0.5722265625, 0.5720391845703126, 0.5731246337890625, 0.5720433349609375, 0.57220703125, 0.5742510375976563, 0.5736365966796875, 0.572663818359375, 0.5723699340820313, 0.5724794921875, 0.5720340576171875, 0.5722705688476563, 0.5724047241210938, 0.57236376953125, 0.5725368041992187, 0.5725255737304688, 0.5728256225585937, 0.5726392211914062, 0.5721354370117188, 0.5721804809570312, 0.5721774291992188, 0.5723484497070312, 0.5721793823242187, 0.572990478515625, 0.5730048217773438, 0.5721558837890625, 0.572242919921875, 0.5721978759765625, 0.5722470703125, 0.5720596313476562, 0.5724047241210938, 0.5724436645507812, 0.5722009887695313, 0.5737861328125, 0.57619970703125, 0.5726668701171875, 0.5722962036132813, 0.5730099487304687, 0.5725439453125, 0.5726085205078125, 0.5728368530273438, 0.5743206176757812, 0.5726044311523437, 0.5726791381835937, 0.5723678588867187, 0.5721937866210938, 0.572822509765625, 0.5727109375, 0.5722070922851562, 0.5722654418945312, 0.5724968872070313, 1.1851029052734374, 0.5723607177734376, 0.57244775390625, 0.5727620849609375, 0.5731502075195313, 0.5729341430664062, 0.5733734130859375, 0.5732301025390625, 0.572663818359375, 0.5735577392578125, 0.5724036865234375, 0.5722737426757812, 0.5723555297851562, 0.5722726440429687, 0.5723494262695312, 0.5725409545898438, 0.5730109252929687, 0.57223681640625, 0.5723402099609375, 0.5722613525390625, 0.572516357421875, 0.5727579956054687, 0.5724467163085938, 0.572221435546875, 0.5731195068359375, 0.5724169921875, 0.5723781127929688, 0.5720811767578124, 0.5718343505859375, 0.5718425903320312, 0.57302734375, 0.5730027465820312, 0.5725501708984375, 0.5730252685546875, 0.57335498046875, 0.5724067993164063, 0.5720145874023438, 0.5724354858398437, 0.5753599853515625, 0.572169189453125, 0.5720125732421875, 0.5726105346679687, 0.5732147216796875, 0.57269970703125, 0.5723483276367187, 0.572010498046875, 0.5722890014648437, 0.5727958984375, 0.5723873901367188, 0.57226953125, 0.5728460693359375, 0.57364892578125, 0.5736980590820312, 0.5732618408203125, 0.5729075317382812, 0.5730396118164063, 0.5730027465820312, 0.5737277221679687, 0.5737984008789062, 0.574740478515625, 0.5733519287109375, 0.5725675659179688, 0.572537841796875, 1.184806884765625, 0.5725839233398438, 0.5726617431640625, 0.573106201171875, 0.5745018920898437, 0.5735618286132812, 0.573259765625, 0.5744578857421875, 0.5749258422851562, 0.5728265991210938, 0.5725081787109375, 0.5727150268554687, 0.5735505981445312, 0.573075439453125, 0.5722562255859375, 0.57212109375, 0.5720760498046875, 0.5723033447265625, 0.5736908569335938, 0.5737195434570312, 0.5725010375976562, 0.57371337890625, 0.5728123168945313, 0.5738147583007812, 0.571978759765625, 0.5718763427734375, 0.5724252319335937, 0.5723299560546875, 0.5721712646484375, 0.5726679077148438, 0.57254296875, 0.5723995971679687, 0.5719193725585937, 0.5722777709960938, 0.571779052734375, 0.5716900024414062, 0.5719306030273438, 0.5717493896484375, 0.571821044921875, 0.5734307861328125, 0.5732904663085937, 0.5728604125976563, 0.572748779296875, 0.57318603515625, 0.5727365112304688, 0.5732413330078125, 0.57249072265625, 0.5729822998046875, 0.573169677734375, 0.5721641235351562, 0.5722808227539062, 0.5722900390625, 0.57253173828125, 0.5722101440429688, 0.5723197631835938, 0.5721200561523437, 0.5727498168945313, 0.5729525756835937, 0.5725030517578125, 0.5723596801757812, 0.5721190185546875, 0.5728788452148438, 0.57232177734375, 1.18618115234375, 0.5729382934570313, 0.5724763793945312, 0.5728051147460937, 0.572410888671875, 0.5722828979492187, 0.5726187744140625, 0.57223681640625, 0.5720278930664062, 0.5722142944335937, 0.5735720825195313, 0.5724139404296875, 0.5722695922851563, 0.5720924072265625, 0.572410888671875, 0.5720698852539062, 0.5738741455078125, 0.5721026611328125, 0.5729136352539063, 0.5728265991210938, 0.5724375, 0.5722224731445312, 0.57202587890625, 0.573212646484375, 0.572484619140625, 0.5720780639648437, 0.57230859375, 0.57369384765625, 0.5728798828125, 0.572020751953125, 0.5721231079101563, 0.5717554931640625, 0.572156982421875, 0.5722654418945312, 0.5718435668945312, 0.5726340942382813, 0.573137939453125, 0.5722542114257813, 0.5726156616210938, 0.572095458984375, 0.572031982421875, 0.5723658447265625, 0.5724303588867188, 0.5722726440429687, 0.5725307006835938, 0.5731655883789063, 0.5721272583007813, 0.5720043334960937, 0.572042236328125, 0.572242919921875, 0.5721856079101563, 0.5724548950195313, 0.5724661865234375, 0.5730764770507812, 0.5732301025390625, 0.5725450439453125, 0.5724866333007812, 0.5721734008789062, 0.5719572143554688, 0.5721549072265625, 0.572000244140625, 0.5720064086914063, 0.5725757446289063, 1.1873709716796874, 0.5724487915039063, 0.5732085571289063, 0.572284912109375, 0.5727989501953125, 0.5728409423828125, 0.5729740600585937, 0.573497314453125, 0.572516357421875, 0.572705810546875, 0.5721886596679687, 0.5721958618164062, 0.5724036865234375, 0.5723381958007813, 0.57270068359375, 0.5734564208984375, 0.5733170776367188, 0.5739089965820312, 0.5724610595703125, 0.5728818969726562, 0.5730293579101563, 0.57270068359375, 0.572495849609375, 0.5729935302734375, 0.5724405517578125, 0.572822509765625, 0.5726105346679687, 0.572315673828125, 0.5717718505859375, 0.5719408569335938, 0.5722603759765625, 0.5722347412109375, 0.5724713134765625, 0.5740123901367188, 0.5722296142578125, 0.5719521484375, 0.572314697265625, 0.572317626953125, 0.5720811767578124, 0.5719705810546875, 0.5717545166015625, 0.5723504638671875, 0.5726515502929688, 0.5724016723632812, 0.5723504638671875, 0.572231689453125, 0.572205078125, 0.5721190185546875, 0.572031982421875, 0.5724467163085938, 0.573043701171875, 0.5730468139648438, 0.5726412353515625, 0.5725030517578125, 0.5725399169921875, 0.5724713134765625, 0.5720606689453125, 0.5723914184570312, 0.5728767700195313, 0.573179931640625, 0.5726597290039063, 0.572295166015625, 0.5719317016601563, 1.1857969970703126, 0.5727313842773437, 0.572368896484375, 0.572347412109375, 0.5732669677734376, 0.5725573120117188, 0.5721907348632812, 0.5723095092773437, 0.5727498168945313, 0.5725081787109375, 0.5725491333007813, 0.5726720581054687, 0.5733457641601563, 0.5733345336914063, 0.5721354370117188, 0.5723043823242188, 0.5719183349609375, 0.5717770385742188, 0.5724968872070313, 0.5726105346679687, 0.573053955078125, 0.5741567993164063, 0.5731102905273437, 0.5720698852539062, 0.5720145874023438, 0.57215185546875, 0.574278564453125, 0.5722521362304688, 0.5720453491210937, 0.5730723266601563, 0.5723924560546875, 0.5727354736328125, 0.572337158203125, 0.5723197631835938, 0.5723668212890625, 0.572221435546875, 0.5729075317382812, 0.5726556396484375, 0.5724456787109375, 0.5738014526367188, 0.5731779174804688, 0.572559326171875, 0.5726812133789062, 0.5728235473632812, 0.572885986328125, 0.5726618041992187, 0.5722745971679688, 0.5731082153320313, 0.5732781982421875, 0.5720135498046875, 0.5721507568359375, 0.57227880859375, 0.5728726806640625, 0.572346435546875, 0.5726781005859375, 0.5728286743164063, 0.5733345336914063, 0.5729228515625, 0.572737548828125, 0.5724548950195313, 0.5741793212890625, 0.5722265625, 0.572494873046875]",tokens/s,1.7194907034525255,,,main,False,False,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-j-6b,,cuda,0,42,,,,,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,,MB,2054.77888,5566.365696,0.0,4919.918592,4635.537408,s,10,5.159750671386719,0.5159750671386718,0.0024202673095528563,0.5153813171386719,0.5170659973144531,0.5199656463623047,0.5222853656005859,"[0.5228652954101562, 0.5143955078125, 0.5157650146484375, 0.5144519653320313, 0.5155265502929688, 0.5141107177734375, 0.516309814453125, 0.5146680908203125, 0.516421630859375, 0.515236083984375]",tokens/s,496.14800463061573,kWh,6.087385805116759e-06,3.3354087369571064e-06,2.7887036198498415e-05,3.7309830740572285e-05,tokens/kWh,6861462.379179726,MB,2054.77888,5566.365696,0.0,4919.918592,4794.439168,s,10,301.012869140625,30.101286914062502,0.007474756939868433,30.101162109375,30.1092380859375,30.11128505859375,30.11292263671875,"[30.104474609375, 30.0874609375, 30.10366796875, 30.09865625, 30.095248046875, 30.1085703125, 30.108783203125, 30.11333203125, 30.097923828125, 30.094751953125]",tokens/s,2.0929337732257594,kWh,0.00035535723721815483,0.00019476662748246779,0.0016039330748120972,0.00215405693951272,tokens/kWh,29247.13773548231,,s,629,305.1713383178704,0.4851690593288888,0.06123176770881389,0.4777687072753906,0.47843901367187497,0.47872921142578123,0.9921051538085938,"[0.47712460327148437, 0.4769034118652344, 0.47702426147460936, 0.47709390258789064, 0.47730584716796876, 0.4771205139160156, 0.4773304443359375, 0.47765298461914063, 0.4770846862792969, 0.4785274963378906, 0.47798681640625, 0.47815167236328127, 0.4778588256835937, 0.4773570556640625, 0.477907958984375, 0.47804620361328126, 0.47764480590820313, 0.4777400207519531, 0.4774901733398437, 0.4775546875, 0.4790906982421875, 0.47841998291015625, 0.4775454711914062, 0.4774205322265625, 0.4777062377929687, 0.4773007507324219, 0.47782296752929687, 0.477517822265625, 0.47927194213867186, 0.47922592163085936, 0.4796824951171875, 0.4796200866699219, 0.478603271484375, 0.47771649169921876, 0.4775034790039063, 0.47786904907226563, 0.4779376525878906, 0.4778526611328125, 0.47831039428710936, 0.47768167114257815, 0.4777820129394531, 0.47838104248046875, 0.4779438171386719, 0.4779376525878906, 0.47762432861328125, 0.47779022216796874, 0.47766015625, 0.4779346008300781, 0.47730584716796876, 0.47776666259765627, 0.4777215881347656, 0.4779438171386719, 0.4776980590820312, 0.4786534423828125, 0.47755673217773437, 0.4775034790039063, 0.47755059814453127, 0.47741543579101564, 0.4771686401367187, 0.47783526611328125, 0.4779069519042969, 0.4776560668945313, 0.9921392822265624, 0.4773294067382812, 0.4772085876464844, 0.477370361328125, 0.4770396728515625, 0.4769504699707031, 0.47741748046875, 0.47746356201171875, 0.4771512451171875, 0.47819979858398437, 0.4786329650878906, 0.47734783935546876, 0.477412353515625, 0.4770017395019531, 0.47757003784179686, 0.47735910034179685, 0.47719219970703125, 0.4772116394042969, 0.4775659484863281, 0.47811892700195313, 0.4780257263183594, 0.4779304809570312, 0.4777492370605469, 0.4775045166015625, 0.4776427612304687, 0.4780738525390625, 0.4773498840332031, 0.47781170654296873, 0.4772178039550781, 0.47686654663085937, 0.4779991149902344, 0.47811175537109374, 0.47742669677734373, 0.4774471740722656, 0.47721368408203124, 0.4768757629394531, 0.47745944213867186, 0.4774481811523438, 0.4774799499511719, 0.4782315979003906, 0.4776785583496094, 0.4772024230957031, 0.477707275390625, 0.47777587890625, 0.47741030883789065, 0.47758950805664063, 0.47756494140625, 0.47724337768554687, 0.4777215881347656, 0.47767962646484374, 0.47741543579101564, 0.47842715454101564, 0.4785602111816406, 0.4774819946289062, 0.4774615173339844, 0.4775465087890625, 0.4775475158691406, 0.4778106994628906, 0.4775577697753906, 0.4773294067382812, 0.47762738037109376, 0.47781375122070313, 0.4781895751953125, 0.9939957885742188, 0.47789157104492186, 0.4779029235839844, 0.4775321044921875, 0.4773294067382812, 0.47682968139648435, 0.4777994384765625, 0.478455810546875, 0.4786780090332031, 0.47789056396484375, 0.47756390380859376, 0.4770979919433594, 0.47737344360351563, 0.4772915344238281, 0.4772536315917969, 0.4778649597167969, 0.47748095703125, 0.4771614685058594, 0.47814862060546875, 0.47750143432617187, 0.4773918762207031, 0.47743896484375, 0.4776662902832031, 0.47875582885742185, 0.4773284606933594, 0.4779539794921875, 0.4778567810058594, 0.47754238891601564, 0.47780154418945314, 0.47773382568359374, 0.4781598815917969, 0.4780472412109375, 0.4777594909667969, 0.477765625, 0.47762841796875, 0.47739596557617187, 0.47771444702148436, 0.47763250732421875, 0.4777687072753906, 0.4783943786621094, 0.47779736328125, 0.4777215881347656, 0.47773284912109376, 0.47762841796875, 0.4779315185546875, 0.4779735107421875, 0.4777461853027344, 0.4787445983886719, 0.47787722778320313, 0.4785080261230469, 0.4784373779296875, 0.47873638916015626, 0.47828274536132814, 0.4777001037597656, 0.47793560791015627, 0.4780277709960937, 0.47800115966796874, 0.47787213134765627, 0.47758950805664063, 0.47742156982421874, 0.47888998413085937, 0.478159912109375, 0.47799700927734373, 0.9920173950195312, 0.47721368408203124, 0.4778260498046875, 0.477939697265625, 0.47749530029296877, 0.4772659912109375, 0.4777287292480469, 0.4779949645996094, 0.47825204467773436, 0.4779581298828125, 0.4773775329589844, 0.4774225769042969, 0.47868109130859376, 0.4778076171875, 0.47798785400390625, 0.478139404296875, 0.47792538452148436, 0.4779366455078125, 0.4774615173339844, 0.47746356201171875, 0.477728759765625, 0.47751473999023436, 0.47717376708984377, 0.47695462036132813, 0.4774942626953125, 0.47761920166015626, 0.4774625244140625, 0.4780001220703125, 0.4778526611328125, 0.47788134765625, 0.4775301513671875, 0.47762127685546873, 0.4773498229980469, 0.4779581298828125, 0.4775413818359375, 0.47721881103515623, 0.47748504638671874, 0.4778824462890625, 0.47803488159179686, 0.47796633911132813, 0.47795098876953124, 0.47843328857421874, 0.4778895263671875, 0.47771749877929687, 0.47763766479492187, 0.47745123291015623, 0.47731610107421873, 0.47720550537109374, 0.47810763549804686, 0.47794686889648436, 0.478023681640625, 0.47817214965820315, 0.4775679931640625, 0.4772167663574219, 0.4781783142089844, 0.4780738525390625, 0.4782264404296875, 0.4780769348144531, 0.4779223022460938, 0.4776662902832031, 0.47806668090820315, 0.4780738525390625, 0.47803903198242187, 0.9936957397460937, 0.47736831665039064, 0.4772413330078125, 0.47760589599609377, 0.47733966064453126, 0.4780892028808594, 0.47794790649414065, 0.4774604797363281, 0.47752294921875, 0.4775792541503906, 0.47779022216796874, 0.4776407165527344, 0.4775096435546875, 0.4774523010253906, 0.4773304443359375, 0.4777021484375, 0.477939697265625, 0.4781055908203125, 0.4771205749511719, 0.477898681640625, 0.47731201171875, 0.47782196044921876, 0.4773775329589844, 0.4771747741699219, 0.4777697143554688, 0.47787213134765627, 0.4778188781738281, 0.4779683837890625, 0.47738983154296877, 0.4772915344238281, 0.4775802917480469, 0.47765914916992186, 0.47731201171875, 0.47755877685546877, 0.4773355407714844, 0.4775382995605469, 0.477918212890625, 0.47747378540039065, 0.4770508728027344, 0.4778793029785156, 0.477528076171875, 0.47719937133789064, 0.4775516052246094, 0.4772341613769531, 0.4773990478515625, 0.47757415771484374, 0.47758746337890623, 0.47739495849609376, 0.47834423828125, 0.4778536682128906, 0.4773058776855469, 0.47772262573242186, 0.4777635498046875, 0.4777185363769531, 0.4785080261230469, 0.4780687255859375, 0.4775168151855469, 0.4807618408203125, 0.47829605102539063, 0.47767245483398435, 0.4783206481933594, 0.4779980773925781, 0.4780625915527344, 0.9916989135742188, 0.47712973022460936, 0.4770652160644531, 0.4772781982421875, 0.47782296752929687, 0.47800833129882814, 0.47793356323242187, 0.47763150024414064, 0.47709390258789064, 0.47777484130859377, 0.4777667236328125, 0.47767550659179686, 0.47757720947265625, 0.4771798400878906, 0.47727410888671873, 0.47779531860351565, 0.4776089477539063, 0.4775792541503906, 0.47766937255859376, 0.47783627319335936, 0.47713690185546875, 0.47779226684570314, 0.4778567810058594, 0.4774615173339844, 0.4775833740234375, 0.47805950927734375, 0.47846502685546877, 0.478561279296875, 0.4782633056640625, 0.4781055908203125, 0.47865753173828124, 0.47780044555664064, 0.47805950927734375, 0.4787906494140625, 0.47858688354492185, 0.47910400390625, 0.47804107666015627, 0.4773498840332031, 0.47763455200195315, 0.4782264404296875, 0.47779736328125, 0.47735809326171874, 0.4779683837890625, 0.4779571228027344, 0.47788134765625, 0.47783627319335936, 0.4788439025878906, 0.47873126220703127, 0.47824075317382814, 0.47784039306640624, 0.47777279663085936, 0.47869952392578125, 0.4775126953125, 0.47748504638671874, 0.4778311767578125, 0.47821005249023435, 0.478708740234375, 0.4786903076171875, 0.4786278381347656, 0.4780533752441406, 0.47773080444335936, 0.47787213134765627, 0.47809127807617186, 0.9952440185546875, 0.4780052490234375, 0.47796734619140624, 0.47861553955078123, 0.47840972900390627, 0.478055419921875, 0.4771358642578125, 0.47734375, 0.4771768188476562, 0.47729766845703125, 0.477949951171875, 0.4772239379882813, 0.4777277526855469, 0.47779531860351565, 0.47714407348632815, 0.47731814575195314, 0.47762841796875, 0.47755877685546877, 0.477528076171875, 0.4772034606933594, 0.47703143310546875, 0.47779531860351565, 0.4779202575683594, 0.4790067138671875, 0.478593017578125, 0.4792412109375, 0.47822540283203124, 0.4779069519042969, 0.4779427795410156, 0.47770932006835937, 0.47803802490234376, 0.4777840576171875, 0.47797247314453123, 0.477939697265625, 0.47781991577148436, 0.4778240051269531, 0.47803289794921877, 0.47818240356445313, 0.47798175048828123, 0.4779959716796875, 0.4778680419921875, 0.4780902404785156, 0.47819674682617186, 0.47809945678710936, 0.4779765625, 0.479388671875, 0.4785356750488281, 0.4774143981933594, 0.4776642456054688, 0.47775238037109374, 0.4779253234863281, 0.4778547058105469, 0.4781363220214844, 0.47750860595703126, 0.4780830688476562, 0.47821728515625, 0.4779703674316406, 0.47782196044921876, 0.4780851135253906, 0.4781322326660156, 0.47786392211914064, 0.4776509399414062, 0.4776099853515625, 0.9929461669921875, 0.47778509521484375, 0.48122674560546874, 0.47778509521484375, 0.4777943115234375, 0.47731814575195314, 0.4780748901367188, 0.47792538452148436, 0.47784756469726564, 0.4781158447265625, 0.4780155029296875, 0.47748403930664063, 0.4777891845703125, 0.4779100036621094, 0.4776908874511719, 0.47778713989257815, 0.47860736083984373, 0.47720037841796875, 0.4777891845703125, 0.4777216491699219, 0.4775577087402344, 0.47795404052734375, 0.47750143432617187, 0.47806668090820315, 0.4786780090332031, 0.477744140625, 0.47708673095703125, 0.47796429443359373, 0.4774686584472656, 0.4776775817871094, 0.4779796447753906, 0.47754855346679687, 0.4775096435546875, 0.4782417907714844, 0.47757107543945315, 0.4774205322265625, 0.4781158447265625, 0.4779376525878906, 0.47806362915039063, 0.47810763549804686, 0.4782090148925781, 0.478413818359375, 0.47833804321289064, 0.4782417907714844, 0.4790927429199219, 0.4781537170410156, 0.47856845092773437, 0.4775301208496094, 0.4781506652832031, 0.4774615173339844, 0.4780472412109375, 0.4780318603515625, 0.47794891357421876, 0.47838311767578123, 0.47825408935546876, 0.4780349426269531, 0.47770111083984373, 0.4779857788085938, 0.47789266967773436, 0.47788128662109375, 0.4787712097167969, 0.47795306396484377, 0.47804415893554686, 0.9943838500976563, 0.4787189636230469, 0.47804107666015627, 0.47781375122070313, 0.477370361328125, 0.4773304443359375, 0.47816705322265624, 0.47811993408203124, 0.4777738342285156, 0.47763250732421875, 0.4774573974609375, 0.47708673095703125, 0.47772467041015626, 0.47725567626953125, 0.47783627319335936, 0.47736831665039064, 0.4771112976074219, 0.478445556640625, 0.47864935302734374, 0.4780984191894531, 0.47761102294921876, 0.4776365966796875, 0.4789073791503906, 0.4771778564453125, 0.47777484130859377, 0.47784756469726564, 0.4784169006347656, 0.4775045166015625, 0.47752191162109375, 0.47713177490234376, 0.47731610107421873, 0.47756390380859376, 0.4773212280273437, 0.47744512939453126, 0.4777543640136719, 0.477222900390625, 0.477707275390625, 0.47724850463867186, 0.47723724365234377, 0.4778823547363281, 0.477412353515625, 0.4772392883300781, 0.4777994384765625, 0.47849676513671874, 0.4779949951171875, 0.47791104125976563, 0.47725466918945314, 0.47723629760742187, 0.47771026611328127, 0.478139404296875, 0.4780195922851562, 0.478129150390625, 0.47857867431640627, 0.47747174072265625, 0.47793869018554686, 0.47770932006835937, 0.4779571228027344, 0.47765509033203124, 0.4775464172363281, 0.47778713989257815, 0.477918212890625, 0.4777820129394531, 0.47804107666015627, 0.9942221069335937, 0.47767550659179686, 0.4771686401367187, 0.477431884765625, 0.4771603698730469, 0.47666278076171875, 0.4774676513671875, 0.47758746337890623, 0.47736114501953125, 0.4776488952636719, 0.47701708984375, 0.47684402465820314, 0.47745944213867186, 0.477633544921875, 0.4778076171875, 0.4772536315917969, 0.47761715698242185, 0.47721881103515623, 0.4775362548828125, 0.47752294921875, 0.4779653015136719, 0.47805645751953124, 0.47783526611328125, 0.47724032592773435, 0.4773447570800781, 0.4780175476074219, 0.4793231506347656, 0.4781588439941406, 0.47761715698242185, 0.47719937133789064, 0.4781537170410156, 0.47752294921875, 0.4775311279296875, 0.47761920166015626, 0.4774993896484375, 0.4768399353027344, 0.4775946960449219, 0.47729248046875, 0.47762841796875, 0.4775372924804687, 0.4772720642089844, 0.4776089477539063, 0.4787261352539062, 0.47772262573242186, 0.47747378540039065, 0.47768780517578124, 0.477528076171875, 0.477149169921875, 0.4776355895996094, 0.47729766845703125, 0.4775628662109375, 0.47870361328125, 0.4778188781738281, 0.4776509399414062, 0.4779898986816406, 0.4782561340332031, 0.47833087158203125, 0.47798681640625, 0.4781035461425781, 0.477955078125, 0.47816192626953125, 0.4783472595214844, 0.47877734375]",tokens/s,2.061137207272146,,,main,False,False,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-0.5B,Qwen/Qwen1.5-0.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,1440.489472,1709.703168,0.0,1063.256064,942.605312,s,10,0.8710727081298828,0.08710727081298827,0.0020526611434407644,0.0869946403503418,0.08830768203735351,0.09043918418884277,0.09214438591003418,"[0.09257068634033203, 0.08689478302001953, 0.08709449768066406, 0.08734588623046875, 0.0876402587890625, 0.08528950500488282, 0.08547952270507812, 0.08543309020996094, 0.08549046325683594, 0.08783401489257812]",tokens/s,2938.9050719957663,kWh,1.0307077509193605e-06,5.647778186555472e-07,2.6262726324173835e-06,4.221758201992291e-06,tokens/kWh,60638243.06166823,MB,1440.817152,1709.703168,0.0,1063.256064,942.607872,s,10,54.10364501953125,5.410364501953125,0.04442393039305028,5.402393798828125,5.47523671875,5.4771017578125,5.478593789062501,"[5.402685546875, 5.4450302734375, 5.474822265625, 5.478966796875, 5.39001171875, 5.366296875, 5.3479423828125, 5.3592861328125, 5.4365009765625, 5.40210205078125]",tokens/s,11.64431712082563,kWh,6.414297001805283e-05,3.5154539095709204e-05,0.00015373626912418487,0.00025303377823794696,tokens/kWh,248978.61636779693,,s,629,54.788874206542985,0.08710472846827182,0.01042761753839409,0.08542822265625,0.08754749298095703,0.08829050750732421,0.17079922241210937,"[0.08673382568359375, 0.08457215881347656, 0.08464179229736328, 0.08488550567626953, 0.08813362884521485, 0.08856473541259766, 0.08671949005126953, 0.08648601531982422, 0.08687923431396484, 0.08737894439697266, 0.08506470489501954, 0.0844554214477539, 0.0844400634765625, 0.08486911773681641, 0.08474524688720703, 0.08481378936767578, 0.08675328063964843, 0.08531455993652344, 0.08472268676757813, 0.08468172454833985, 0.08468172454833985, 0.0845322265625, 0.08479436492919921, 0.08471552276611329, 0.08475545501708984, 0.08474521636962891, 0.08502886199951172, 0.0846899871826172, 0.0848424301147461, 0.084569091796875, 0.08475955200195312, 0.08495104217529297, 0.08461721801757813, 0.08463571166992187, 0.08458643341064453, 0.08476876831054687, 0.08473804473876953, 0.08473190307617187, 0.0844554214477539, 0.08442777252197266, 0.08484761810302735, 0.08466124725341796, 0.08562483215332031, 0.08599142456054687, 0.08450764465332031, 0.0840970230102539, 0.08538521575927735, 0.08838041687011719, 0.0868485107421875, 0.08694374084472656, 0.08736255645751953, 0.08686080169677735, 0.08700518035888671, 0.08699289703369141, 0.08753254699707032, 0.0868485107421875, 0.08696627044677735, 0.08728985595703125, 0.08770150756835937, 0.08642969512939454, 0.08510566711425781, 0.08681779479980468, 0.1740953674316406, 0.08652799987792968, 0.08564736175537109, 0.0842209243774414, 0.08706150054931641, 0.08500326538085938, 0.08664473724365235, 0.08596275329589843, 0.0853391342163086, 0.08664268493652344, 0.08708096313476563, 0.08690073394775391, 0.08696934509277343, 0.08694783782958984, 0.08664575958251954, 0.08597196960449219, 0.08684748840332031, 0.08683110046386719, 0.08509439849853516, 0.08638566589355469, 0.08632217407226563, 0.08716287994384765, 0.087119873046875, 0.08713215637207031, 0.08506470489501954, 0.08651168060302734, 0.08632109069824219, 0.08708403015136719, 0.08612662506103516, 0.08650748443603516, 0.08707481384277344, 0.08641535949707031, 0.08600678253173828, 0.08589107513427735, 0.08514252471923828, 0.08699494171142579, 0.08681676483154296, 0.08650956726074219, 0.0871383056640625, 0.08712806701660156, 0.08630374145507813, 0.08593612670898437, 0.08639590454101563, 0.08506681823730469, 0.08687712097167968, 0.08704512023925781, 0.08696115112304688, 0.08838451385498047, 0.08482816314697265, 0.08464895629882813, 0.08431206512451171, 0.08453427124023437, 0.08795238494873046, 0.08794931030273437, 0.08684134674072265, 0.08686592102050782, 0.08699801635742188, 0.08754073333740234, 0.087947265625, 0.08687513732910156, 0.08657100677490234, 0.08599350738525391, 0.08582653045654297, 0.17401548767089844, 0.0865423355102539, 0.08694374084472656, 0.08704204559326172, 0.08744652557373046, 0.08659967803955078, 0.08703897857666015, 0.08700109100341796, 0.08715776062011718, 0.08681574249267578, 0.08705433654785157, 0.08619929504394531, 0.08701548767089844, 0.086857666015625, 0.08725811004638671, 0.08656588745117187, 0.08733184051513672, 0.0871731185913086, 0.08719257354736328, 0.08709120178222657, 0.0869713897705078, 0.0868485107421875, 0.08620236968994141, 0.08700415802001953, 0.08701337432861328, 0.0868331527709961, 0.08713420867919922, 0.08696627044677735, 0.08672972869873047, 0.08645836639404297, 0.08661094665527344, 0.08728678131103515, 0.08693452453613282, 0.08721206665039062, 0.08688022613525391, 0.08789299011230468, 0.08672870635986328, 0.08694477081298828, 0.087731201171875, 0.08747007751464844, 0.08717005157470703, 0.08711270141601563, 0.08729190063476562, 0.08698271942138672, 0.08669484710693359, 0.08654745483398438, 0.08669286346435547, 0.08840806579589844, 0.08711475372314453, 0.08676044464111328, 0.08631705474853515, 0.08654956817626953, 0.08599648284912109, 0.08647679901123047, 0.08694989013671875, 0.08696524810791016, 0.08698060607910156, 0.08548556518554687, 0.08556953430175782, 0.0868823013305664, 0.08658534240722657, 0.08709939575195312, 0.08710451507568359, 0.17466163635253906, 0.08715673828125, 0.08728371429443359, 0.08680345916748047, 0.08702361297607422, 0.0872099838256836, 0.08695299530029296, 0.0865228500366211, 0.08702054595947266, 0.08711270141601563, 0.08681267547607421, 0.0862402572631836, 0.08722022247314454, 0.08714342498779297, 0.08641433715820312, 0.08614297485351563, 0.08621363067626953, 0.08743526458740235, 0.0874260482788086, 0.08645123291015624, 0.08728675079345703, 0.08692940521240235, 0.08707891082763672, 0.08719055938720703, 0.08709014129638672, 0.08714854431152344, 0.08714035034179687, 0.08686592102050782, 0.0871751708984375, 0.08751411437988281, 0.08707379150390625, 0.08701952362060547, 0.08471449279785156, 0.08613683319091797, 0.08813772583007813, 0.08714240264892578, 0.08702566528320313, 0.0872959976196289, 0.08773222351074218, 0.08883302307128907, 0.08757759857177734, 0.08561151885986328, 0.08676659393310547, 0.08834764862060547, 0.08476467132568359, 0.08778854370117188, 0.08729497528076172, 0.08688441467285156, 0.08712287902832032, 0.0869744644165039, 0.0868136978149414, 0.08646246337890626, 0.08621875, 0.08678707122802734, 0.08692428588867188, 0.08732262420654296, 0.08703692626953125, 0.08800665283203125, 0.08684236907958984, 0.08649215698242188, 0.0869048309326172, 0.08684236907958984, 0.08680857849121094, 0.1740042266845703, 0.08717619323730469, 0.0870635528564453, 0.0868219223022461, 0.08698159790039063, 0.08682291412353516, 0.08556953430175782, 0.08729804992675781, 0.08690278625488282, 0.08721305847167969, 0.08680857849121094, 0.08707686614990234, 0.08708812713623047, 0.0862003173828125, 0.08698265838623047, 0.08703385925292968, 0.08684031677246094, 0.08691814422607422, 0.08724281311035156, 0.08704608154296875, 0.08688742065429687, 0.08603648376464844, 0.08850739288330078, 0.08509849548339844, 0.08658636474609376, 0.08696934509277343, 0.08461007690429688, 0.08453526306152344, 0.08452095794677734, 0.08469811248779296, 0.084959228515625, 0.08464383697509766, 0.08490086364746094, 0.08458649444580078, 0.08620851135253907, 0.08480665588378906, 0.08471449279785156, 0.08450867462158203, 0.08490086364746094, 0.0845486068725586, 0.08454962921142578, 0.08446873474121094, 0.0846909408569336, 0.0845506591796875, 0.08456806182861328, 0.08460185241699218, 0.08472268676757813, 0.08459366607666016, 0.08461414337158203, 0.0848005142211914, 0.08458854675292969, 0.08446669006347657, 0.08458854675292969, 0.08470323181152344, 0.08453427124023437, 0.08449433898925782, 0.08460192108154296, 0.08460384368896484, 0.08456192016601563, 0.08444825744628906, 0.08477184295654297, 0.0844421157836914, 0.08439193725585938, 0.17255935668945313, 0.08706764984130859, 0.0871720962524414, 0.0869908447265625, 0.08572518157958985, 0.08475033569335938, 0.0848353271484375, 0.08454962921142578, 0.08473804473876953, 0.08482406616210937, 0.08465305328369141, 0.08468479919433594, 0.08473395538330078, 0.08460902404785156, 0.08453529357910156, 0.08463053131103515, 0.0845455322265625, 0.08432128143310547, 0.08473190307617187, 0.08471449279785156, 0.084578369140625, 0.08457823944091797, 0.08466944122314453, 0.08652288055419922, 0.08498278045654296, 0.08480255889892578, 0.08473702239990234, 0.08468787384033204, 0.08633650970458985, 0.0850851821899414, 0.08458751678466797, 0.08468991851806641, 0.08474726104736328, 0.0846909408569336, 0.0885032958984375, 0.08480870056152344, 0.0858071060180664, 0.08694374084472656, 0.08694989013671875, 0.08657920074462891, 0.08678604888916015, 0.08691814422607422, 0.084853759765625, 0.08535858917236328, 0.08599654388427734, 0.08458137512207031, 0.08471347045898438, 0.084495361328125, 0.08461516571044922, 0.0845322265625, 0.08443698883056641, 0.0841891860961914, 0.08461619567871094, 0.0845137939453125, 0.0844031982421875, 0.08491827392578125, 0.08468889617919922, 0.08446873474121094, 0.08431718444824218, 0.0844595489501953, 0.08526640319824219, 0.08521421051025391, 0.0846397476196289, 0.1709055938720703, 0.08506777954101563, 0.08515379333496094, 0.08487731170654297, 0.08472576141357421, 0.08464383697509766, 0.08466124725341796, 0.08488448333740234, 0.08489778900146484, 0.08490598297119141, 0.08465203094482422, 0.0848189468383789, 0.08532991790771484, 0.08477286529541016, 0.08459468841552735, 0.08518246459960938, 0.08490803527832032, 0.0846561279296875, 0.08462028503417969, 0.08444313812255859, 0.0862208023071289, 0.08462643432617188, 0.08517222595214843, 0.08487423706054688, 0.08629145812988281, 0.08506470489501954, 0.0851937255859375, 0.08485785675048828, 0.08452505493164063, 0.0856289291381836, 0.08493059539794921, 0.08512406158447265, 0.08508620452880859, 0.08450457763671874, 0.084515869140625, 0.08459465789794922, 0.08446975708007813, 0.08475852966308593, 0.08493772888183594, 0.08463161468505859, 0.08454649353027344, 0.0845322265625, 0.08463667297363281, 0.08475753784179688, 0.08451376342773438, 0.08450662231445312, 0.08457318115234375, 0.08449842834472657, 0.08456294250488282, 0.08457523345947265, 0.08489676666259766, 0.08472576141357421, 0.087193603515625, 0.08506368255615235, 0.08472064208984376, 0.08463565063476562, 0.08460390472412109, 0.08499712371826172, 0.08570368194580077, 0.08466124725341796, 0.08457625579833984, 0.08450355529785156, 0.08474214172363281, 0.17052569580078125, 0.0847298583984375, 0.08458444976806641, 0.08468275451660157, 0.08454041290283203, 0.08447081756591797, 0.08448406219482422, 0.08457523345947265, 0.0845301742553711, 0.0844247055053711, 0.08477286529541016, 0.08456502532958984, 0.08428745269775391, 0.08627097320556641, 0.08500633239746094, 0.08606105804443359, 0.08684646606445312, 0.08513126373291016, 0.08685874938964844, 0.08501042938232421, 0.08475545501708984, 0.0846346206665039, 0.08462438201904297, 0.08665497589111328, 0.08509542083740235, 0.08484454345703125, 0.08662322998046874, 0.08476467132568359, 0.08460492706298828, 0.0845998077392578, 0.08459878540039062, 0.08473702239990234, 0.08481484985351563, 0.08530022430419922, 0.08468275451660157, 0.08468685150146485, 0.08445645141601563, 0.08475548553466797, 0.08516194915771484, 0.08527359771728515, 0.08474931335449219, 0.08479542541503907, 0.08579888153076172, 0.08668876647949218, 0.08599346923828124, 0.08607129669189453, 0.08541696166992188, 0.08659865570068359, 0.08641024017333984, 0.08462847900390626, 0.08492339324951172, 0.0848875503540039, 0.084674560546875, 0.08466534423828125, 0.08485273742675781, 0.08462643432617188, 0.08465408325195313, 0.08468479919433594, 0.08457011413574218, 0.0847984619140625, 0.08467558288574219, 0.08471142578125, 0.08483334350585937, 0.1702665557861328, 0.08738406372070312, 0.08923648071289063, 0.08826573181152343, 0.08826470184326173, 0.08866099548339844, 0.08817049407958985, 0.08779468536376953, 0.08849612426757812, 0.08879206085205078, 0.08808345794677734, 0.08809164428710937, 0.0881233901977539, 0.08886784362792968, 0.08884019470214843, 0.08862515258789062, 0.08821043395996093, 0.08844905853271484, 0.08820121765136718, 0.08817353820800782, 0.08852275085449218, 0.08854227447509766, 0.08476563262939453, 0.0872243194580078, 0.08712806701660156, 0.08852275085449218, 0.08827391815185547, 0.08493363189697266, 0.08474214172363281, 0.08446873474121094, 0.08490598297119141, 0.08460697937011719, 0.08487014770507813, 0.08471858978271485, 0.08479743957519531, 0.0852490234375, 0.08817971038818359, 0.0883599395751953, 0.08766566467285156, 0.0848189468383789, 0.08454348754882812, 0.08476876831054687, 0.08462950134277344, 0.08455782318115235, 0.0845998077392578, 0.08457523345947265, 0.08428851318359375, 0.08459468841552735, 0.08465920257568359, 0.08506470489501954, 0.08472268676757813, 0.08413491058349609, 0.08476467132568359, 0.08465715026855469, 0.08452915191650391, 0.08448102569580078, 0.0847267837524414, 0.08448614501953125, 0.08460594940185547, 0.08444416046142578, 0.08624127960205077, 0.08508006286621093, 0.08459878540039062, 0.1760184326171875, 0.08723353576660156, 0.08769331359863282, 0.08806092834472656, 0.08757453155517578, 0.08732978820800781, 0.08919449615478516, 0.08813977813720703, 0.08797286224365235, 0.08746189117431641, 0.08733798217773438, 0.08667750549316407, 0.08574054718017578, 0.08622182464599609, 0.08452095794677734, 0.08488345336914062, 0.08462950134277344, 0.08516403198242188, 0.08760736083984375, 0.08629344177246094, 0.08519782257080079, 0.08494694519042968, 0.08468275451660157, 0.08495308685302734, 0.08589212799072266, 0.08511488342285156, 0.08629449462890625, 0.08508927917480469, 0.08481075286865235, 0.08436326599121094, 0.08498995208740234, 0.08503705596923829, 0.08542822265625, 0.08469503784179687, 0.08443392181396485, 0.08501554870605468, 0.0846376953125, 0.08496742248535157, 0.08612351989746093, 0.08490290832519531, 0.08461516571044922, 0.08477081298828125, 0.0848005142211914, 0.08621977233886718, 0.08505651092529297, 0.08459571075439454, 0.08459878540039062, 0.08486918640136719, 0.08474515533447266, 0.08450969696044922, 0.08503501129150391, 0.0847267837524414, 0.08468889617919922, 0.08508927917480469, 0.0886118392944336, 0.08830156707763671, 0.08738508605957031, 0.08487321472167969, 0.0855920639038086, 0.08490496063232422, 0.08478208160400391, 0.08481689453125, 0.08506368255615235]",tokens/s,11.480433009606973,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-160m,EleutherAI/pythia-160m,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,1312.858112,1030.22592,0.0,383.778816,312.280064,s,10,0.275944766998291,0.0275944766998291,0.0013350624992567249,0.0271289119720459,0.027794796180725095,0.02969363794326782,0.031212711353302003,"[0.03159247970581055, 0.027166879653930665, 0.027137279510498047, 0.02737283134460449, 0.02710416030883789, 0.02712054443359375, 0.027108831405639647, 0.027081024169921874, 0.027172767639160156, 0.027087968826293947]",tokens/s,9277.218871905097,kWh,3.2455721808658853e-07,1.778424162839056e-07,7.707522954780324e-07,1.2731519298485266e-06,tokens/kWh,201075766.3702066,MB,1312.858112,1030.22592,0.0,383.778816,347.089408,s,10,16.864729003906252,1.6864729003906251,0.020323100201592524,1.6794135742187501,1.6910009521484375,1.7188808471679686,1.7411847631835937,"[1.7467607421875, 1.67984912109375, 1.6774495849609374, 1.678377685546875, 1.67897802734375, 1.6768824462890626, 1.6750291748046875, 1.684479248046875, 1.684805419921875, 1.6821175537109374]",tokens/s,37.356070165970515,kWh,1.9882176955741114e-05,1.0895638286132514e-05,4.512764553432279e-05,7.590546077619641e-05,tokens/kWh,829979.8111462949,,s,629,17.07843275070189,0.027151721384263763,0.003315017386101685,0.026595327377319337,0.0273623046875,0.02783129653930664,0.05417021270751953,"[0.029247488021850586, 0.027447296142578126, 0.028238847732543947, 0.029016063690185546, 0.02889727973937988, 0.028690431594848635, 0.028497919082641602, 0.029652992248535157, 0.028964864730834962, 0.02876518440246582, 0.029016063690185546, 0.02860851287841797, 0.02900480079650879, 0.029263872146606446, 0.02835353660583496, 0.028241920471191406, 0.02977484893798828, 0.031457279205322264, 0.028201984405517577, 0.02774220848083496, 0.027778047561645508, 0.02750464057922363, 0.027625471115112304, 0.027877376556396483, 0.027576320648193358, 0.027623424530029295, 0.027853824615478515, 0.027623424530029295, 0.02755583953857422, 0.027467775344848632, 0.027467775344848632, 0.027464704513549806, 0.02751692771911621, 0.027671552658081053, 0.028444671630859376, 0.02751081657409668, 0.027527168273925783, 0.02693833541870117, 0.02656768035888672, 0.02656051254272461, 0.026586111068725587, 0.026575904846191406, 0.02656355285644531, 0.026597375869750976, 0.02668339157104492, 0.026562559127807618, 0.0265799674987793, 0.02705510330200195, 0.02689945602416992, 0.02673766326904297, 0.026646528244018555, 0.026635263442993166, 0.02674892807006836, 0.026631168365478516, 0.026613759994506835, 0.026697727203369142, 0.026650623321533205, 0.02710323143005371, 0.026925119400024414, 0.02659014320373535, 0.027423744201660157, 0.027077632904052733, 0.05460070419311523, 0.026619903564453123, 0.026565631866455077, 0.026629119873046874, 0.026554399490356446, 0.02660246467590332, 0.026628095626831053, 0.02656768035888672, 0.026635263442993166, 0.0266527042388916, 0.026569696426391603, 0.026598463058471678, 0.026578880310058593, 0.0265799674987793, 0.02658406448364258, 0.026687488555908204, 0.026688512802124024, 0.02656358337402344, 0.02654003143310547, 0.026614784240722656, 0.026625024795532228, 0.02652672004699707, 0.026619903564453123, 0.028004352569580077, 0.02753945541381836, 0.02736332893371582, 0.027281408309936524, 0.02712678337097168, 0.026574848175048828, 0.026603519439697267, 0.02654310417175293, 0.026565631866455077, 0.0265031681060791, 0.026557439804077147, 0.02653388786315918, 0.026556415557861326, 0.02654412841796875, 0.02655948829650879, 0.02668441581726074, 0.026625024795532228, 0.02655948829650879, 0.026588159561157225, 0.02654310417175293, 0.026572799682617186, 0.02655129623413086, 0.02656768035888672, 0.026604543685913085, 0.02655955123901367, 0.026596288681030273, 0.026562559127807618, 0.026596351623535155, 0.026594303131103517, 0.026558464050292968, 0.02655232048034668, 0.02659328079223633, 0.02651238441467285, 0.026549247741699217, 0.026588159561157225, 0.02657587242126465, 0.02659328079223633, 0.026556415557861326, 0.02650726318359375, 0.02656870460510254, 0.0541736946105957, 0.026597375869750976, 0.026587135314941408, 0.02655232048034668, 0.026550271987915038, 0.026600448608398438, 0.026652671813964843, 0.026689535140991212, 0.02655129623413086, 0.02653081512451172, 0.02658406448364258, 0.026677248001098632, 0.02651548767089844, 0.02671510314941406, 0.027616256713867186, 0.027520000457763674, 0.026843135833740234, 0.026626047134399415, 0.026536991119384765, 0.02662294387817383, 0.026621952056884765, 0.02655129623413086, 0.026534912109375, 0.026634239196777345, 0.026594303131103517, 0.02670182418823242, 0.0265799674987793, 0.02651136016845703, 0.02654003143310547, 0.02653388786315918, 0.02654412841796875, 0.026617855072021485, 0.026968063354492186, 0.02676736068725586, 0.026615808486938477, 0.02655539131164551, 0.02651545524597168, 0.026661888122558593, 0.026573823928833007, 0.026454015731811522, 0.02652470397949219, 0.026513376235961915, 0.0265031681060791, 0.02651852798461914, 0.026694656372070313, 0.026492927551269533, 0.026570751190185548, 0.026505216598510743, 0.026565631866455077, 0.026583040237426758, 0.026574848175048828, 0.02650931167602539, 0.02651033592224121, 0.026664960861206056, 0.026570751190185548, 0.02651238441467285, 0.026582015991210937, 0.02654515266418457, 0.02649087905883789, 0.02654719924926758, 0.026638336181640625, 0.02657689666748047, 0.026598400115966796, 0.05530112075805664, 0.027283456802368163, 0.027225088119506836, 0.02709503936767578, 0.026944511413574217, 0.02655436706542969, 0.02654310417175293, 0.026695680618286134, 0.026656768798828126, 0.0265850887298584, 0.026588191986083986, 0.02655023956298828, 0.026674175262451173, 0.026564607620239256, 0.026580991744995116, 0.026611711502075194, 0.026570751190185548, 0.0265533447265625, 0.026611711502075194, 0.026570751190185548, 0.026529792785644532, 0.02669977569580078, 0.026549247741699217, 0.02660767936706543, 0.02659424018859863, 0.02671001625061035, 0.02650111961364746, 0.02655539131164551, 0.026587135314941408, 0.02655232048034668, 0.026657791137695314, 0.0265533447265625, 0.026541055679321288, 0.026558464050292968, 0.02651238441467285, 0.026514432907104493, 0.026467327117919923, 0.026565631866455077, 0.02692403221130371, 0.026690559387207033, 0.02655436706542969, 0.026600448608398438, 0.026719232559204102, 0.02657689666748047, 0.026662912368774414, 0.02651136016845703, 0.026564607620239256, 0.026537984848022462, 0.026505216598510743, 0.02651238441467285, 0.02656358337402344, 0.02652262306213379, 0.026663936614990235, 0.026497024536132813, 0.026646528244018555, 0.02652876853942871, 0.026637311935424804, 0.02657587242126465, 0.02652672004699707, 0.02657177543640137, 0.026604543685913085, 0.026583040237426758, 0.026484735488891603, 0.054171646118164066, 0.02660147285461426, 0.026550271987915038, 0.02656972885131836, 0.02654515266418457, 0.02658406448364258, 0.026532863616943358, 0.02650726318359375, 0.02657792091369629, 0.02667622375488281, 0.0265799674987793, 0.02656153678894043, 0.026580991744995116, 0.026556480407714845, 0.02659833526611328, 0.026598400115966796, 0.026611711502075194, 0.026615808486938477, 0.026514432907104493, 0.02654207992553711, 0.02656768035888672, 0.02656870460510254, 0.026633216857910157, 0.026589183807373046, 0.02657177543640137, 0.02655436706542969, 0.026660863876342773, 0.026614784240722656, 0.026558464050292968, 0.02655436706542969, 0.026617855072021485, 0.026604543685913085, 0.026554399490356446, 0.026611679077148436, 0.026650623321533205, 0.026488832473754883, 0.02629631996154785, 0.02631372833251953, 0.0275230712890625, 0.027797504425048827, 0.02753023910522461, 0.02730803108215332, 0.02698854446411133, 0.026604543685913085, 0.02653081512451172, 0.026624000549316407, 0.02660147285461426, 0.026764287948608398, 0.026658815383911134, 0.02656870460510254, 0.026557439804077147, 0.02654617691040039, 0.026639360427856446, 0.026580991744995116, 0.026534912109375, 0.02653593635559082, 0.026594303131103517, 0.026620927810668944, 0.026596351623535155, 0.02657587242126465, 0.026604543685913085, 0.026604543685913085, 0.02653388786315918, 0.05420851135253906, 0.0265677433013916, 0.02660960006713867, 0.02667519950866699, 0.026580991744995116, 0.026594303131103517, 0.026562559127807618, 0.026665983200073243, 0.026590208053588867, 0.02653900718688965, 0.026637311935424804, 0.026638336181640625, 0.02654617691040039, 0.026590208053588867, 0.026562559127807618, 0.026696704864501954, 0.026629119873046874, 0.02652774429321289, 0.02652057647705078, 0.02653081512451172, 0.0265533447265625, 0.02654207992553711, 0.026694656372070313, 0.02658406448364258, 0.026640384674072266, 0.02657792091369629, 0.026534912109375, 0.02651545524597168, 0.026521600723266602, 0.02656153678894043, 0.02651852798461914, 0.026580991744995116, 0.02653388786315918, 0.026529792785644532, 0.0265799674987793, 0.026592256546020508, 0.026693632125854492, 0.02651545524597168, 0.026639360427856446, 0.027520000457763674, 0.02674380874633789, 0.026613759994506835, 0.026694656372070313, 0.026521600723266602, 0.026641408920288087, 0.026607616424560547, 0.026558464050292968, 0.026610687255859376, 0.026630144119262695, 0.02669260787963867, 0.026634271621704102, 0.02657072067260742, 0.02654003143310547, 0.02657587242126465, 0.026605567932128905, 0.026534912109375, 0.026697727203369142, 0.026587135314941408, 0.02652060890197754, 0.026654687881469727, 0.026598400115966796, 0.026714111328125, 0.02658406448364258, 0.054166526794433595, 0.02671615982055664, 0.026641408920288087, 0.026625024795532228, 0.0265994873046875, 0.026596288681030273, 0.02654719924926758, 0.026572799682617186, 0.026620927810668944, 0.026605567932128905, 0.02669977569580078, 0.02654316711425781, 0.026589120864868164, 0.026582015991210937, 0.026604543685913085, 0.026550271987915038, 0.026619903564453123, 0.02656051254272461, 0.026521600723266602, 0.026631168365478516, 0.02628505516052246, 0.02677350425720215, 0.026702848434448243, 0.02657177543640137, 0.02660966491699219, 0.02656358337402344, 0.02651545524597168, 0.0265533447265625, 0.026504192352294922, 0.02660147285461426, 0.02653900718688965, 0.026639360427856446, 0.0265850887298584, 0.026626047134399415, 0.02654003143310547, 0.026556447982788087, 0.026569696426391603, 0.026558464050292968, 0.02659328079223633, 0.026556415557861326, 0.02654617691040039, 0.02668339157104492, 0.026589183807373046, 0.026498048782348634, 0.026669055938720702, 0.026529792785644532, 0.026572799682617186, 0.02655129623413086, 0.0265164794921875, 0.026590208053588867, 0.02657792091369629, 0.026616832733154298, 0.026598400115966796, 0.026660863876342773, 0.026573823928833007, 0.026595327377319337, 0.026570751190185548, 0.0265164794921875, 0.02639366340637207, 0.026553279876708986, 0.02658406448364258, 0.026657791137695314, 0.02649497604370117, 0.05412044906616211, 0.026598400115966796, 0.026583040237426758, 0.026651647567749022, 0.02652262306213379, 0.026586111068725587, 0.02655129623413086, 0.026594303131103517, 0.02712678337097168, 0.02734489631652832, 0.026644479751586913, 0.026637311935424804, 0.026594303131103517, 0.026646528244018555, 0.026602495193481446, 0.027348991394042968, 0.027696128845214843, 0.027305984497070314, 0.027241472244262696, 0.02698854446411133, 0.026565631866455077, 0.026608640670776368, 0.027296768188476563, 0.027320320129394532, 0.027320320129394532, 0.027268096923828124, 0.02708684730529785, 0.026829824447631836, 0.026608640670776368, 0.02674073600769043, 0.02660147285461426, 0.026590208053588867, 0.02655129623413086, 0.02654207992553711, 0.02661075210571289, 0.0265860481262207, 0.02667622375488281, 0.026572799682617186, 0.026582015991210937, 0.026586111068725587, 0.02653081512451172, 0.026566656112670898, 0.02655232048034668, 0.026702848434448243, 0.026602495193481446, 0.02654617691040039, 0.026550271987915038, 0.026645503997802734, 0.026550271987915038, 0.026616832733154298, 0.02660966491699219, 0.026596351623535155, 0.026595327377319337, 0.026586111068725587, 0.02654617691040039, 0.02654617691040039, 0.026786815643310546, 0.02656153678894043, 0.026608640670776368, 0.026602495193481446, 0.026529792785644532, 0.026586111068725587, 0.026638336181640625, 0.05429862213134766, 0.026634239196777345, 0.026537984848022462, 0.026639360427856446, 0.02672127914428711, 0.02692915153503418, 0.027014144897460936, 0.026755071640014647, 0.026573823928833007, 0.026562559127807618, 0.026572799682617186, 0.026690559387207033, 0.02657177543640137, 0.026617855072021485, 0.026610687255859376, 0.026606592178344726, 0.026598400115966796, 0.026573823928833007, 0.026597375869750976, 0.026514432907104493, 0.026582015991210937, 0.02659328079223633, 0.026673152923583986, 0.027404319763183593, 0.026896352767944335, 0.026578943252563478, 0.026537984848022462, 0.026552352905273437, 0.02658710479736328, 0.026514432907104493, 0.02651238441467285, 0.026514432907104493, 0.02652262306213379, 0.02651238441467285, 0.02653081512451172, 0.026626047134399415, 0.026565631866455077, 0.026716224670410155, 0.026623935699462892, 0.026580991744995116, 0.02654617691040039, 0.026556415557861326, 0.026479616165161132, 0.02652876853942871, 0.026670080184936523, 0.026643455505371092, 0.026827775955200195, 0.026611711502075194, 0.027028480529785157, 0.027226112365722657, 0.027289600372314454, 0.02755583953857422, 0.0273623046875, 0.0273623046875, 0.02728550338745117, 0.02791116714477539, 0.027173887252807616, 0.0267775993347168, 0.026597375869750976, 0.02654515266418457, 0.02656153678894043, 0.026599424362182617, 0.026617855072021485, 0.05418803024291992, 0.026620927810668944, 0.026647552490234375, 0.02657689666748047, 0.02651238441467285, 0.026595327377319337, 0.026607616424560547, 0.026603519439697267, 0.026594303131103517, 0.026619903564453123, 0.02672230339050293, 0.026630144119262695, 0.026639360427856446, 0.02667622375488281, 0.026677248001098632, 0.026631168365478516, 0.026617855072021485, 0.026565631866455077, 0.026587135314941408, 0.0265164794921875, 0.026616832733154298, 0.026588159561157225, 0.026612735748291014, 0.026612735748291014, 0.02656768035888672, 0.026558464050292968, 0.026629119873046874, 0.026508287429809572, 0.026702848434448243, 0.02671615982055664, 0.0265533447265625, 0.026586111068725587, 0.026626047134399415, 0.02655232048034668, 0.02658406448364258, 0.02654412841796875, 0.02653081512451172, 0.0265799674987793, 0.026612735748291014, 0.026603519439697267, 0.02635366439819336, 0.026727424621582032, 0.02656153678894043, 0.026595327377319337, 0.026583040237426758, 0.026492927551269533, 0.026572799682617186, 0.02653593635559082, 0.026678272247314453, 0.02653900718688965, 0.02690559959411621, 0.02734284782409668, 0.02688102340698242, 0.026624000549316407, 0.02656358337402344, 0.026578943252563478, 0.027030527114868166, 0.027415552139282227, 0.02731724739074707, 0.027236352920532225, 0.027314176559448244, 0.027259904861450194, 0.027412479400634765]",tokens/s,36.83007739537158,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Salesforce/codegen-16B-nl,Salesforce/codegen-16B-nl,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1469, in _autoset_attn_implementation + cls._check_and_enable_flash_attn_2( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1560, in _check_and_enable_flash_attn_2 + raise ValueError( +ValueError: CodeGenForCausalLM does not support Flash Attention 2.0 yet. Please request to add support where the model is hosted, on its model hub page: https://huggingface.co//tmp/tmpirdn8wy3/no_weights_model/discussions/new or in the Transformers GitHub repo: https://github.com/huggingface/transformers/issues/new + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-MoE-A2.7B,Qwen/Qwen1.5-MoE-A2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3640, in from_pretrained + hf_quantizer.preprocess_model( + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 182, in preprocess_model + return self._process_model_before_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 85, in _process_model_before_weight_loading + model, has_been_replaced = replace_with_awq_linear( + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 178, in replace_with_awq_linear + _, has_been_replaced = replace_with_awq_linear( + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 178, in replace_with_awq_linear + _, has_been_replaced = replace_with_awq_linear( + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 178, in replace_with_awq_linear + _, has_been_replaced = replace_with_awq_linear( + [Previous line repeated 1 more time] + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 165, in replace_with_awq_linear + model._modules[name] = target_cls( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemm.py"", line 102, in __init__ + assert out_features % (32 // self.w_bit) == 0 +AssertionError + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-1.3b,EleutherAI/pythia-1.3b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,,,MB,1610.354688,2254.962688,0.0,1608.515584,1463.6928,s,10,1.209205665588379,0.1209205665588379,0.0011005379634326184,0.1205662727355957,0.1215480598449707,0.12274063606262207,0.12369469703674317,"[0.12393321228027344, 0.12128304290771484, 0.12008969879150391, 0.12004278564453125, 0.12018284606933594, 0.12025001525878906, 0.12034620666503906, 0.12078633880615235, 0.12113209533691406, 0.121159423828125]",tokens/s,2117.0922969124094,kWh,1.4171419990441157e-06,7.762810317477228e-07,6.305967346357218e-06,8.499390377149056e-06,tokens/kWh,30119807.261502665,MB,1610.952704,2254.962688,0.0,1608.515584,1560.974848,s,10,70.14807763671874,7.0148077636718735,0.00472223733563277,7.012995361328125,7.0220806640625,7.0226084472656245,7.023030673828125,"[7.017875, 7.0128837890625, 7.0119638671875, 7.01192919921875, 7.01310693359375, 7.0105234375, 7.0078984375, 7.02196337890625, 7.01679736328125, 7.02313623046875]",tokens/s,8.981001635748731,kWh,8.273579535030184e-05,4.5345370896494765e-05,0.0003660393860852451,0.0004941205523320418,tokens/kWh,127499.25034015773,,s,629,71.1210855636597,0.11307008833650187,0.01439647456429658,0.11126681518554687,0.11167272796630859,0.11198136138916015,0.23187720581054685,"[0.11166719818115234, 0.11130470275878906, 0.11149517059326172, 0.11213107299804688, 0.11133849334716797, 0.11115827178955077, 0.11096371459960938, 0.11128832244873046, 0.11117977905273438, 0.11118182373046875, 0.11122688293457031, 0.1112647705078125, 0.11119513702392578, 0.11100774383544922, 0.111087646484375, 0.11108963012695312, 0.11133439636230469, 0.11160678100585937, 0.11127910614013672, 0.1123737564086914, 0.1114972152709961, 0.11121971130371094, 0.11113676452636718, 0.1111009292602539, 0.11132621002197265, 0.11129344177246094, 0.11116851043701172, 0.11124531555175782, 0.1112442855834961, 0.11126681518554687, 0.11126579284667969, 0.11136102294921875, 0.11114291381835938, 0.11164262390136719, 0.1114081268310547, 0.11137843322753906, 0.11116134643554687, 0.11157810974121093, 0.1113333740234375, 0.11127091217041016, 0.11125965118408203, 0.11129241943359375, 0.11120745849609374, 0.11137942504882813, 0.11125247955322265, 0.11137229156494141, 0.11129138946533203, 0.11126988983154297, 0.11214335632324218, 0.11184333038330078, 0.11213414764404298, 0.11166515350341796, 0.11144300842285157, 0.11137529754638673, 0.11148185729980468, 0.11138457489013671, 0.11156582641601563, 0.1114777603149414, 0.1114081268310547, 0.11125965118408203, 0.11127603149414063, 0.11128425598144531, 0.2332733154296875, 0.11126271820068359, 0.11134259033203125, 0.1112136001586914, 0.11113475036621094, 0.1112872314453125, 0.11119516754150391, 0.11118179321289062, 0.11123712158203125, 0.11110195159912109, 0.11118694305419922, 0.11108454132080078, 0.11117158508300781, 0.1111357421875, 0.1112248306274414, 0.11135078430175781, 0.1113016357421875, 0.11121049499511719, 0.11146649932861329, 0.11143270111083985, 0.11139379119873047, 0.11125862121582031, 0.11111116790771484, 0.11113369750976562, 0.11115213012695313, 0.11124736022949219, 0.11156070709228516, 0.11131289672851563, 0.11127193450927735, 0.11115929412841796, 0.11136819458007813, 0.11125145721435546, 0.11122073364257813, 0.11129753875732422, 0.11136614227294922, 0.11128217315673829, 0.11125145721435546, 0.11142047882080078, 0.11116230773925781, 0.11125759887695312, 0.1110487060546875, 0.11122994995117187, 0.11132518768310547, 0.1113026885986328, 0.11126576232910156, 0.11120947265625, 0.11332198333740234, 0.11140608215332032, 0.11135590362548828, 0.11126579284667969, 0.11120845031738281, 0.11143065643310547, 0.1113917465209961, 0.11126988983154297, 0.11125350189208984, 0.11131187438964844, 0.11131903839111328, 0.11127091217041016, 0.11129446411132812, 0.11129446411132812, 0.11138969421386719, 0.1114603500366211, 0.11127808380126954, 0.23181004333496094, 0.11102105712890625, 0.1111910400390625, 0.11175424194335938, 0.11164672088623047, 0.11138969421386719, 0.11125247955322265, 0.11111219024658203, 0.11101798248291016, 0.11089920043945313, 0.11102413177490235, 0.11136723327636719, 0.1116630401611328, 0.11121561431884766, 0.11134259033203125, 0.11122585296630859, 0.11111116790771484, 0.11116339111328125, 0.11112857818603515, 0.11112242889404297, 0.11111014556884766, 0.11112966156005859, 0.11107218933105469, 0.11174297332763672, 0.11127603149414063, 0.11119821166992187, 0.11121663665771485, 0.11118080139160157, 0.11127808380126954, 0.1111583023071289, 0.11111011505126953, 0.11111014556884766, 0.11128832244873046, 0.11131903839111328, 0.1111562271118164, 0.11147058868408204, 0.11126681518554687, 0.1112279052734375, 0.11108051300048828, 0.11122681427001953, 0.11114086151123047, 0.11119923400878906, 0.11140201568603515, 0.11115004730224609, 0.1115494384765625, 0.11138969421386719, 0.11126886749267578, 0.11192524719238281, 0.111351806640625, 0.11120127868652344, 0.11122994995117187, 0.11163238525390624, 0.11214540863037109, 0.11136921691894532, 0.11118592071533204, 0.11130879974365235, 0.11120845031738281, 0.11130470275878906, 0.1112442855834961, 0.11167743682861328, 0.11148799896240234, 0.11126988983154297, 0.11132012939453124, 0.23181817626953125, 0.11116236877441406, 0.11126585388183594, 0.11114182281494141, 0.11116748809814453, 0.1111695327758789, 0.11111014556884766, 0.11164057922363281, 0.11159449768066407, 0.11120127868652344, 0.11121766662597657, 0.11119206237792968, 0.11146649932861329, 0.11107635498046875, 0.11124121856689453, 0.11109375762939454, 0.11119718170166015, 0.11117465972900391, 0.11109478759765624, 0.11113471984863281, 0.11132415771484375, 0.11108249664306641, 0.11122585296630859, 0.11124838256835938, 0.11113267517089843, 0.11119821166992187, 0.11118386840820313, 0.11122380828857421, 0.11127808380126954, 0.11117056274414062, 0.11127500915527344, 0.111388671875, 0.11147878265380859, 0.111246337890625, 0.11121868896484376, 0.1112616958618164, 0.1112811508178711, 0.11174400329589844, 0.11160575866699218, 0.11130368041992188, 0.11123001861572265, 0.11139884948730469, 0.11166207885742188, 0.11131187438964844, 0.11219455718994141, 0.11126377868652344, 0.11129955291748046, 0.11126886749267578, 0.11122278594970703, 0.11126886749267578, 0.1113169937133789, 0.11135282897949218, 0.11117362976074219, 0.1112442855834961, 0.11130060577392578, 0.11121766662597657, 0.11142655944824219, 0.111246337890625, 0.11133132934570313, 0.11140409851074219, 0.11129542541503906, 0.11128422546386718, 0.11133952331542969, 0.23190016174316405, 0.11121459197998047, 0.11119718170166015, 0.11123302459716797, 0.11124736022949219, 0.1111551971435547, 0.1117624282836914, 0.11134259033203125, 0.11116851043701172, 0.11123609924316406, 0.11112960052490234, 0.11113983917236328, 0.11123097229003906, 0.11135692596435547, 0.11133030700683594, 0.1111756820678711, 0.11099750518798829, 0.11165388488769531, 0.11119923400878906, 0.11110399627685547, 0.11134361267089844, 0.11165695953369141, 0.11130470275878906, 0.11117874908447266, 0.11137843322753906, 0.11113881683349609, 0.11112754821777343, 0.11104460906982422, 0.11110195159912109, 0.1110456314086914, 0.11133542633056641, 0.11121971130371094, 0.11121459197998047, 0.11156275177001954, 0.11139590454101563, 0.11135379028320312, 0.11121868896484376, 0.11118592071533204, 0.11130982208251954, 0.11158528137207031, 0.11178291320800782, 0.11122176361083984, 0.11130470275878906, 0.11126681518554687, 0.111246337890625, 0.11125452423095702, 0.11126374053955078, 0.11133030700683594, 0.11145932769775391, 0.11136000061035156, 0.11132621002197265, 0.11127705383300782, 0.11152998352050782, 0.1112442855834961, 0.11130265808105469, 0.11128729248046874, 0.11133952331542969, 0.11251200103759766, 0.11131084442138672, 0.1113221435546875, 0.11128931427001953, 0.11129036712646484, 0.11130879974365235, 0.2321817626953125, 0.11112652587890624, 0.11125247955322265, 0.11120230102539062, 0.11118284606933594, 0.11110912322998047, 0.11104768371582031, 0.11113471984863281, 0.11123814392089844, 0.11114495849609375, 0.11111321258544922, 0.11121971130371094, 0.11117874908447266, 0.11118592071533204, 0.11127603149414063, 0.11113369750976562, 0.11118796539306641, 0.11104364776611328, 0.11120121765136719, 0.11109683227539062, 0.11134873962402343, 0.11129241943359375, 0.11124531555175782, 0.11129138946533203, 0.1111695327758789, 0.11111936187744141, 0.11120230102539062, 0.11098214721679688, 0.11109478759765624, 0.11121772766113282, 0.11126573181152344, 0.11151564788818359, 0.11180134582519531, 0.1111551971435547, 0.11124018859863281, 0.11106508636474609, 0.11120845031738281, 0.11125452423095702, 0.11243007659912109, 0.11165286254882813, 0.11120947265625, 0.11119308471679687, 0.11124224090576172, 0.11133747100830078, 0.1112647705078125, 0.11123200225830078, 0.11110912322998047, 0.11123097229003906, 0.11112140655517579, 0.11116134643554687, 0.11120230102539062, 0.11124838256835938, 0.11131597137451171, 0.11115929412841796, 0.11126681518554687, 0.11124326324462891, 0.11124940490722657, 0.11136819458007813, 0.1113364486694336, 0.11219558715820313, 0.11167436981201172, 0.11133542633056641, 0.11132006072998046, 0.23260365295410157, 0.1113016357421875, 0.11109069061279297, 0.11130675506591797, 0.11117772674560547, 0.11111116790771484, 0.11099852752685548, 0.11096575927734376, 0.11118592071533204, 0.11106201934814453, 0.11124940490722657, 0.1109964828491211, 0.11128729248046874, 0.11114086151123047, 0.11112041473388672, 0.11092476654052734, 0.11116646575927734, 0.11122585296630859, 0.11110297393798828, 0.11115929412841796, 0.1110282211303711, 0.11134361267089844, 0.11134054565429688, 0.11161497497558594, 0.11118796539306641, 0.11125145721435546, 0.1112965087890625, 0.11122898864746093, 0.11125344085693359, 0.11107839965820313, 0.1114081268310547, 0.11136511993408203, 0.11113164520263671, 0.11111936187744141, 0.11107635498046875, 0.11111628723144532, 0.1109964828491211, 0.11101900482177735, 0.11115827178955077, 0.11171942138671875, 0.11166413116455078, 0.11109580993652343, 0.11116649627685547, 0.11109782409667969, 0.11119821166992187, 0.11115110778808594, 0.11119206237792968, 0.11109580993652343, 0.1113333740234375, 0.111388671875, 0.11111014556884766, 0.11112242889404297, 0.11119926452636719, 0.11117769622802734, 0.11110707092285156, 0.11125350189208984, 0.111246337890625, 0.11176140594482421, 0.11179315185546874, 0.11137229156494141, 0.1111234588623047, 0.1113917465209961, 0.11129138946533203, 0.23285554504394532, 0.11228876495361328, 0.1119969253540039, 0.11132012939453124, 0.1113087387084961, 0.1110118408203125, 0.11107635498046875, 0.11119821166992187, 0.11114905548095703, 0.11114291381835938, 0.11154227447509765, 0.111283203125, 0.11145011138916015, 0.11127091217041016, 0.11112140655517579, 0.11115110778808594, 0.11166719818115234, 0.11149517059326172, 0.11157708740234375, 0.11116441345214843, 0.11162931060791016, 0.11222220611572266, 0.11127398681640625, 0.11119721221923828, 0.11121660614013672, 0.11111833953857422, 0.11153817749023437, 0.11167436981201172, 0.11123817443847656, 0.11156681823730469, 0.11157299041748046, 0.11123200225830078, 0.11113267517089843, 0.11120339202880859, 0.11119302368164062, 0.11129446411132812, 0.11130265808105469, 0.11117874908447266, 0.11140914916992188, 0.11178803253173829, 0.11143987274169923, 0.111388671875, 0.11134054565429688, 0.11126271820068359, 0.11172557067871093, 0.11178189086914063, 0.11163648223876953, 0.11228876495361328, 0.11165491485595704, 0.11137843322753906, 0.11123712158203125, 0.11235433959960937, 0.11163645172119141, 0.11138969421386719, 0.11162419128417969, 0.11138253021240234, 0.11193548583984375, 0.11135289764404296, 0.11133433532714844, 0.11140198516845704, 0.11156582641601563, 0.11123814392089844, 0.11139584350585938, 0.23310745239257813, 0.11154329681396484, 0.11122585296630859, 0.11103334045410156, 0.11112242889404297, 0.11111321258544922, 0.11110502624511719, 0.11125759887695312, 0.11132006072998046, 0.11105689239501954, 0.11160371398925781, 0.11119308471679687, 0.11138662719726562, 0.11129039764404297, 0.11160777282714844, 0.1111203842163086, 0.11147058868408204, 0.11210137939453126, 0.11113471984863281, 0.11226214599609376, 0.11134054565429688, 0.11112960052490234, 0.11109683227539062, 0.11101696014404297, 0.11132927703857422, 0.11150438690185546, 0.11192934417724609, 0.11167231750488281, 0.11156684875488282, 0.11118489837646485, 0.11123814392089844, 0.11113683319091797, 0.11117254638671875, 0.11120333099365234, 0.11124940490722657, 0.11127398681640625, 0.11138355255126953, 0.11188531494140624, 0.11153510284423829, 0.11116134643554687, 0.11122892761230468, 0.11117874908447266, 0.11126892852783203, 0.11124422454833985, 0.11132723236083984, 0.111246337890625, 0.1125038070678711, 0.11145523071289062, 0.1111695327758789, 0.11121663665771485, 0.11127193450927735, 0.11120537567138672, 0.11170098876953125, 0.11128729248046874, 0.11114701080322266, 0.112110595703125, 0.11148185729980468, 0.11123916625976563, 0.1111910400390625, 0.11130879974365235, 0.11141222381591796, 0.11127603149414063, 0.11140608215332032, 0.23328665161132814, 0.11162009429931641, 0.11125350189208984, 0.11111116790771484, 0.11114086151123047, 0.1111234588623047, 0.11114803314208985, 0.11112454223632813, 0.11118073272705079, 0.11184435272216797, 0.11149311828613281, 0.1110835189819336, 0.11113676452636718, 0.11192729949951172, 0.11148902130126953, 0.1111900177001953, 0.1112074203491211, 0.11114393615722656, 0.11168256378173828, 0.11129446411132812, 0.11139584350585938, 0.11125043487548827, 0.11120435333251953, 0.11126886749267578, 0.11185561370849609, 0.11130675506591797, 0.11252838134765625, 0.11186688232421875, 0.11209420776367188, 0.11134873962402343, 0.11136819458007813, 0.11132006072998046, 0.11117056274414062, 0.1115688934326172, 0.111425537109375, 0.11140096282958985, 0.11195801544189453, 0.11133439636230469, 0.11169280242919923, 0.11158016204833984, 0.11150745391845703, 0.11149209594726563, 0.11174915313720703, 0.11140297698974609, 0.11168153381347656, 0.11251609802246093, 0.11149823760986328, 0.1112125473022461, 0.11124947357177735, 0.11128211212158202, 0.11113369750976562, 0.11132518768310547, 0.11279666900634766, 0.11159654235839844, 0.11184844970703126, 0.11136102294921875, 0.11125350189208984, 0.11140096282958985, 0.11135078430175781, 0.11131084442138672, 0.11126579284667969, 0.11146348571777344, 0.111246337890625]",tokens/s,8.844071979708314,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-2.7b,EleutherAI/pythia-2.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,2221.408256,3330.801664,0.0,2684.35456,2447.595008,s,10,2.311121505737305,0.2311121505737305,0.0008602880707090342,0.23093637084960938,0.23197196044921875,0.23249637908935547,0.23291591400146486,"[0.2330207977294922, 0.23185542297363282, 0.23051023864746092, 0.23036329650878906, 0.2304237823486328, 0.2300004119873047, 0.23067584228515625, 0.2311968994140625, 0.231619384765625, 0.23145542907714844]",tokens/s,1107.6873256749418,kWh,2.719861554979074e-06,1.4903567604051204e-06,1.2427491002590687e-05,1.6637709317974884e-05,tokens/kWh,15386733.54050159,MB,2222.546944,3330.801664,0.0,2684.35456,2597.68064,s,10,135.98273144531248,13.598273144531248,0.005626597165862265,13.59927783203125,13.6021646484375,13.60570927734375,13.60854498046875,"[13.60925390625, 13.5985986328125, 13.5982265625, 13.6012080078125, 13.59995703125, 13.5903544921875, 13.5950361328125, 13.600318359375, 13.601376953125, 13.5884013671875]",tokens/s,4.63294120734267,kWh,0.00016051297066488652,8.797306729544289e-05,0.0007245357930120089,0.0009730218309723382,tokens/kWh,64746.74873126358,,s,629,137.8431466979981,0.21914649713513207,0.027476947243743017,0.21579168701171875,0.2161471496582031,0.2163871795654297,0.4466235400390625,"[0.2167510986328125, 0.21612850952148438, 0.2157936706542969, 0.21595852661132814, 0.21579776000976564, 0.21578341674804688, 0.21598515319824219, 0.21577626037597655, 0.21600563049316407, 0.21667738342285156, 0.2159656982421875, 0.21651968383789064, 0.21600767517089844, 0.21581004333496093, 0.21593600463867188, 0.21574656677246093, 0.2157589111328125, 0.21606707763671876, 0.2157782440185547, 0.21572096252441406, 0.2157322235107422, 0.21579058837890625, 0.21588172912597656, 0.21599232482910155, 0.2161080322265625, 0.21578341674804688, 0.21588172912597656, 0.21578341674804688, 0.21624217224121095, 0.21602406311035155, 0.2159482879638672, 0.21581619262695312, 0.21592169189453125, 0.21585302734375, 0.2157742004394531, 0.21580697631835938, 0.2158233642578125, 0.21585305786132813, 0.216163330078125, 0.2158858184814453, 0.21590425109863282, 0.2159718475341797, 0.2159482879638672, 0.21587660217285157, 0.2160394287109375, 0.21591244506835938, 0.2157660217285156, 0.21589913940429686, 0.21592576599121094, 0.21606399536132812, 0.21596365356445313, 0.2158428192138672, 0.21607833862304687, 0.21651866149902343, 0.21638246154785157, 0.21635379028320312, 0.21680844116210937, 0.21638656616210938, 0.21633331298828126, 0.21572709655761718, 0.21570457458496095, 0.21581517028808594, 0.448078857421875, 0.21560012817382812, 0.21530323791503905, 0.21556626892089845, 0.21550796508789063, 0.2154915771484375, 0.21565951538085937, 0.21555711364746094, 0.21552742004394532, 0.21577728271484375, 0.2157127685546875, 0.21635789489746093, 0.21575782775878907, 0.21568818664550782, 0.2154915771484375, 0.2156810302734375, 0.21565440368652344, 0.21555404663085936, 0.21562060546875, 0.21577728271484375, 0.2157557830810547, 0.21564006042480469, 0.21555815124511718, 0.21571583557128907, 0.2157373504638672, 0.21600973510742189, 0.21567079162597655, 0.2159482879638672, 0.21565235900878907, 0.2156820526123047, 0.21579469299316406, 0.21556941223144532, 0.215583740234375, 0.21569842529296876, 0.21576191711425782, 0.21580294799804686, 0.21571781921386718, 0.215583740234375, 0.21599334716796875, 0.2168739776611328, 0.21618482971191405, 0.21625753784179688, 0.21591142272949218, 0.21590835571289063, 0.21615206909179688, 0.21587557983398437, 0.21748121643066406, 0.21573324584960937, 0.21585816955566406, 0.21574758911132813, 0.21579161071777345, 0.21602613830566406, 0.21589602661132812, 0.21573837280273436, 0.21600767517089844, 0.2161459197998047, 0.216195068359375, 0.21681561279296874, 0.21595545959472656, 0.2159800262451172, 0.21614183044433594, 0.21579263305664062, 0.21579168701171875, 0.44681414794921875, 0.21564927673339843, 0.21552024841308592, 0.2156093444824219, 0.21581517028808594, 0.21569024658203126, 0.21543324279785156, 0.2153850555419922, 0.21555917358398438, 0.21604454040527343, 0.2156441650390625, 0.21566259765625, 0.2157373504638672, 0.21574348449707031, 0.2156165771484375, 0.2157229461669922, 0.21567692565917967, 0.21569024658203126, 0.2158540802001953, 0.21572198486328126, 0.2158008270263672, 0.21594009399414063, 0.21594009399414063, 0.2173450164794922, 0.21601280212402343, 0.21582643127441406, 0.21562265014648438, 0.216121337890625, 0.21586329650878905, 0.21575065612792968, 0.21581721496582032, 0.21577626037597655, 0.21595135498046875, 0.21587046813964844, 0.2158223419189453, 0.21579263305664062, 0.21578854370117187, 0.21617765808105469, 0.21579571533203126, 0.2159615936279297, 0.2160906219482422, 0.2161459197998047, 0.21604864501953125, 0.2158551025390625, 0.2156943359375, 0.21593498229980468, 0.2158254089355469, 0.21669273376464843, 0.21583258056640625, 0.21568307495117187, 0.21580288696289063, 0.21581210327148437, 0.21588070678710938, 0.21590118408203124, 0.2158018493652344, 0.2157803497314453, 0.2158192596435547, 0.2158192596435547, 0.2157670440673828, 0.21574758911132813, 0.21588890075683595, 0.2157137908935547, 0.21567079162597655, 0.44661248779296875, 0.215657470703125, 0.21567999267578125, 0.21604351806640626, 0.2158858184814453, 0.21600665283203124, 0.21593600463867188, 0.2156145324707031, 0.21560723876953125, 0.21591346740722656, 0.21581619262695312, 0.21606809997558593, 0.216342529296875, 0.21595750427246094, 0.21586636352539063, 0.21600154113769532, 0.2158540802001953, 0.21593907165527343, 0.21594522094726562, 0.21608345031738282, 0.21599334716796875, 0.21589401245117187, 0.21608038330078125, 0.21600665283203124, 0.21591448974609376, 0.21582131958007814, 0.21565542602539062, 0.2157004852294922, 0.21572921752929688, 0.21575570678710937, 0.215878662109375, 0.21577113342285156, 0.2156390380859375, 0.21572709655761718, 0.21576191711425782, 0.215878662109375, 0.21596774291992188, 0.21590835571289063, 0.21587660217285157, 0.21585305786132813, 0.2158356475830078, 0.21571685791015624, 0.21579263305664062, 0.21578956604003907, 0.21685247802734375, 0.2158223419189453, 0.21567692565917967, 0.21570661926269533, 0.21589503479003908, 0.21621554565429688, 0.2161643524169922, 0.21638758850097656, 0.215878662109375, 0.21575885009765625, 0.21584588623046874, 0.21589401245117187, 0.2158745574951172, 0.21597900390625, 0.21583769226074218, 0.21566361999511718, 0.2158182373046875, 0.21573017883300782, 0.21579776000976564, 0.44662783813476564, 0.21568511962890624, 0.2158233642578125, 0.21589605712890625, 0.2156615753173828, 0.21600154113769532, 0.21570355224609375, 0.2157639617919922, 0.21581312561035157, 0.21569024658203126, 0.215552001953125, 0.21569638061523438, 0.21593087768554686, 0.2157076416015625, 0.21581004333496093, 0.21580902099609375, 0.21563699340820314, 0.21653094482421875, 0.21587660217285157, 0.21565440368652344, 0.21584588623046874, 0.2159288330078125, 0.21573426818847657, 0.21571583557128907, 0.21572096252441406, 0.21561036682128906, 0.21595750427246094, 0.21581210327148437, 0.215841796875, 0.21570559692382812, 0.21621452331542967, 0.2162554931640625, 0.216015869140625, 0.21625958251953126, 0.21600767517089844, 0.2159052734375, 0.2159800262451172, 0.21584077453613282, 0.21585714721679689, 0.2158612518310547, 0.21581517028808594, 0.2157496337890625, 0.21570252990722658, 0.21573837280273436, 0.21577215576171874, 0.21601689147949218, 0.21586534118652342, 0.21566053771972657, 0.21617971801757813, 0.215804931640625, 0.2159964141845703, 0.215910400390625, 0.21567692565917967, 0.21574143981933594, 0.215993408203125, 0.21583967590332032, 0.21577626037597655, 0.21723341369628907, 0.2157936706542969, 0.21567181396484375, 0.21573017883300782, 0.2158602294921875, 0.21591346740722656, 0.44625918579101564, 0.2154598388671875, 0.21547109985351562, 0.215942138671875, 0.2155018310546875, 0.2154967041015625, 0.21560525512695314, 0.21534719848632813, 0.2155335693359375, 0.21555609130859374, 0.21548646545410155, 0.21554893493652344, 0.21556838989257812, 0.2154239959716797, 0.21581210327148437, 0.21597080993652343, 0.2159831085205078, 0.21576191711425782, 0.21572813415527345, 0.21552543640136718, 0.21562054443359374, 0.21571994018554688, 0.21557862854003906, 0.21559500122070313, 0.21570867919921874, 0.21560115051269532, 0.21548851013183593, 0.21564210510253906, 0.21561856079101563, 0.21550079345703124, 0.2157137908935547, 0.21574348449707031, 0.21571481323242186, 0.21626162719726563, 0.21590016174316407, 0.2157127685546875, 0.21570867919921874, 0.21562982177734374, 0.21705215454101562, 0.21569229125976563, 0.21560012817382812, 0.2157373504638672, 0.21579058837890625, 0.21566464233398439, 0.21566361999511718, 0.21576191711425782, 0.21585101318359376, 0.21570867919921874, 0.21573529052734375, 0.21557554626464845, 0.2158305206298828, 0.2157742004394531, 0.2156083221435547, 0.2155704345703125, 0.21572813415527345, 0.21567082214355468, 0.21573321533203124, 0.2156513214111328, 0.21562777709960937, 0.21566265869140624, 0.21600965881347656, 0.21582028198242187, 0.21637837219238282, 0.4470773620605469, 0.2156072998046875, 0.21567079162597655, 0.21566566467285156, 0.21550079345703124, 0.21559091186523438, 0.21565542602539062, 0.21559500122070313, 0.2156697540283203, 0.21570661926269533, 0.21575474548339843, 0.21560426330566407, 0.2156011199951172, 0.21565235900878907, 0.21573939514160156, 0.2161090545654297, 0.2160199737548828, 0.21577523803710938, 0.21591552734375, 0.21555815124511718, 0.2156441650390625, 0.21569842529296876, 0.21553868103027343, 0.21665382385253906, 0.21579571533203126, 0.21555711364746094, 0.2155888671875, 0.21576499938964844, 0.215762939453125, 0.21582028198242187, 0.2158582458496094, 0.2157608337402344, 0.21596368408203126, 0.21594313049316408, 0.2157445068359375, 0.2157639617919922, 0.21581414794921874, 0.21576502990722657, 0.21569635009765625, 0.215762939453125, 0.2158223419189453, 0.21592268371582032, 0.21613157653808593, 0.21586329650878905, 0.21606501770019532, 0.21634661865234375, 0.21581517028808594, 0.21568716430664062, 0.215804931640625, 0.21572096252441406, 0.215689208984375, 0.21578239440917968, 0.21577523803710938, 0.21575167846679688, 0.21579776000976564, 0.21582028198242187, 0.21593600463867188, 0.2157178955078125, 0.21569024658203126, 0.21573324584960937, 0.21574758911132813, 0.21596876525878905, 0.21589605712890625, 0.44737637329101565, 0.21560012817382812, 0.2154035186767578, 0.21556224060058593, 0.21559091186523438, 0.21557760620117186, 0.2155878448486328, 0.2156195831298828, 0.21551922607421875, 0.21568511962890624, 0.21572813415527345, 0.21568818664550782, 0.21565542602539062, 0.21562265014648438, 0.21565440368652344, 0.21580799865722655, 0.21636402893066406, 0.2157178955078125, 0.2158018493652344, 0.21578341674804688, 0.21576499938964844, 0.21569740295410156, 0.21566464233398439, 0.21570970153808594, 0.2157936706542969, 0.21613772583007812, 0.21595237731933595, 0.21639680480957033, 0.21639474487304688, 0.21601791381835939, 0.2160025634765625, 0.21584999084472656, 0.2159288330078125, 0.21633024597167969, 0.21584793090820312, 0.2159964141845703, 0.21607936096191407, 0.2160148468017578, 0.21588992309570312, 0.21595852661132814, 0.21601689147949218, 0.21620632934570314, 0.21618380737304688, 0.216595458984375, 0.21603021240234374, 0.21584690856933594, 0.21581619262695312, 0.21571994018554688, 0.21570252990722658, 0.21567692565917967, 0.21569638061523438, 0.21571072387695311, 0.2158970947265625, 0.2156810302734375, 0.2158981170654297, 0.2158602294921875, 0.21611932373046874, 0.21608546447753907, 0.21588479614257813, 0.21581004333496093, 0.21603225708007812, 0.21623910522460937, 0.21606809997558593, 0.44815972900390627, 0.21594316101074218, 0.21585305786132813, 0.21599130249023438, 0.21577932739257813, 0.21586534118652342, 0.21586431884765625, 0.21613055419921876, 0.21584895324707032, 0.21639578247070312, 0.21571994018554688, 0.215657470703125, 0.21565951538085937, 0.2156748809814453, 0.21572607421875, 0.21584690856933594, 0.21561138916015626, 0.21668966674804688, 0.2158008270263672, 0.21577113342285156, 0.21580902099609375, 0.2158673858642578, 0.21579673767089844, 0.21592063903808595, 0.21624319458007812, 0.21564927673339843, 0.21612953186035155, 0.21587046813964844, 0.2156810302734375, 0.21577830505371093, 0.2156943359375, 0.2155827178955078, 0.21565542602539062, 0.21575372314453126, 0.21602610778808592, 0.2159052734375, 0.2158039093017578, 0.21560421752929687, 0.21574656677246093, 0.2161326141357422, 0.21603021240234374, 0.21574041748046874, 0.21587251281738282, 0.21601791381835939, 0.21597900390625, 0.2160343017578125, 0.21582028198242187, 0.21569126892089843, 0.2157424621582031, 0.21565235900878907, 0.21565440368652344, 0.2158602294921875, 0.21615206909179688, 0.21636915588378905, 0.2160148468017578, 0.21692825317382813, 0.21595960998535157, 0.21593696594238282, 0.21585101318359376, 0.21585101318359376, 0.21587251281738282, 0.2156134338378906, 0.21572813415527345, 0.44772250366210936, 0.2154598388671875, 0.21554896545410157, 0.21563491821289063, 0.21536154174804686, 0.2154977264404297, 0.21561651611328125, 0.2155335693359375, 0.21565338134765624, 0.21574867248535157, 0.21558367919921875, 0.2155714874267578, 0.21574960327148437, 0.2156390380859375, 0.21553868103027343, 0.21560838317871095, 0.21563385009765626, 0.2156810302734375, 0.2155704345703125, 0.2155704345703125, 0.21559603881835937, 0.21553663635253906, 0.21553152465820313, 0.21560525512695314, 0.21560421752929687, 0.21559397888183593, 0.21667225646972657, 0.21568818664550782, 0.21559910583496095, 0.21570457458496095, 0.2156513214111328, 0.21557554626464845, 0.21559193420410155, 0.21564210510253906, 0.21565338134765624, 0.21612953186035155, 0.2163251190185547, 0.21550592041015626, 0.215910400390625, 0.21579469299316406, 0.2157455291748047, 0.215762939453125, 0.21571891784667968, 0.21559295654296876, 0.21559091186523438, 0.2155847625732422, 0.21548442077636717, 0.21573426818847657, 0.21578341674804688, 0.21575270080566405, 0.21571583557128907, 0.2156134338378906, 0.2155878448486328, 0.21560012817382812, 0.21608242797851562, 0.2159646759033203, 0.2156615753173828, 0.21582438659667968, 0.21574348449707031, 0.215689208984375, 0.21561856079101563, 0.21556121826171876, 0.2156380157470703]",tokens/s,4.563157582132701,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-14B,Qwen/Qwen1.5-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,3927.220224,12732.33408,0.0,12085.886976,11337.364992,s,10,11.005286376953126,1.1005286376953127,0.00219951912246486,1.1001186523437498,1.1034800170898438,1.1035955383300782,1.1036879553222656,"[1.1009342041015624, 1.102259765625, 1.0977261962890625, 1.099001708984375, 1.0989405517578126, 1.0975576171875, 1.0993031005859375, 1.103454345703125, 1.1023978271484376, 1.1037110595703126]",tokens/s,232.615482443152,kWh,1.2971816278166244e-05,7.107321095700172e-06,6.386246775661464e-05,8.394160513048104e-05,tokens/kWh,3049739.156191579,MB,3927.220224,12732.33408,0.0,12085.886976,11686.79936,s,10,645.4452890625,64.54452890625001,0.006590572212624524,64.54356250000001,64.553694140625,64.5538998046875,64.5540643359375,"[64.5536484375, 64.5513828125, 64.538359375, 64.55410546875, 64.5438203125, 64.5392109375, 64.53455859375, 64.5433046875, 64.5480546875, 64.53884375]",tokens/s,0.9760703357445925,kWh,0.0007619009707371396,0.00041758989640207795,0.003727131037257969,0.004906621904397187,tokens/kWh,12839.791047184834,,s,629,654.3247843627931,1.0402619783192257,0.1307385835871094,1.0244464111328124,1.0251733154296874,1.0253723876953125,2.12362228515625,"[1.0239293212890626, 1.0247884521484374, 1.0239590454101561, 1.0238433227539063, 1.0245233154296876, 1.0249482421875, 1.0245938720703125, 1.024648193359375, 1.0243778076171874, 1.0244761962890625, 1.0242918701171875, 1.02411474609375, 1.0238392333984374, 1.02443310546875, 1.0240072021484374, 1.025123291015625, 1.0243983154296874, 1.0239702758789062, 1.0238289794921875, 1.024879638671875, 1.0246707763671874, 1.0245980224609375, 1.0249052734375, 1.0246092529296875, 1.0252646484375, 1.025154052734375, 1.0244208984375, 1.0245867919921876, 1.025153076171875, 1.0249769287109376, 1.025039306640625, 1.0247618408203125, 1.0240020751953125, 1.0246666259765624, 1.024626708984375, 1.024693359375, 1.0244617919921875, 1.024501708984375, 1.02472802734375, 1.0254991455078124, 1.024680908203125, 1.025005615234375, 1.0249237060546874, 1.0251907958984374, 1.0256405029296876, 1.0250250244140624, 1.025184814453125, 1.02483251953125, 1.024786376953125, 1.0252471923828126, 1.0244935302734375, 1.0253773193359375, 1.024436279296875, 1.025090576171875, 1.025364990234375, 1.0249923095703124, 1.0241658935546876, 1.024015380859375, 1.0247249755859376, 1.0251048583984375, 1.0238597412109376, 1.02401123046875, 2.12693603515625, 1.0248826904296875, 1.024394287109375, 1.02513671875, 1.02491650390625, 1.024427978515625, 1.02502294921875, 1.0245028076171876, 1.0245499267578124, 1.024806884765625, 1.024963623046875, 1.024541748046875, 1.02464306640625, 1.0238269653320313, 1.024679931640625, 1.025101806640625, 1.0239959106445313, 1.0239354858398437, 1.024395263671875, 1.0241566162109375, 1.0247188720703124, 1.0248836669921875, 1.0245191650390626, 1.0240450439453126, 1.024563232421875, 1.0246707763671874, 1.0240665283203125, 1.02398974609375, 1.02415869140625, 1.024153564453125, 1.0244813232421874, 1.0248365478515624, 1.0246881103515626, 1.0250660400390625, 1.024818115234375, 1.02487451171875, 1.0241402587890625, 1.0241505126953125, 1.0248140869140625, 1.0247342529296875, 1.0252052001953125, 1.024384033203125, 1.0249267578125, 1.0247813720703125, 1.0250096435546876, 1.02519091796875, 1.0286304931640624, 1.0247679443359374, 1.0250260009765626, 1.02523291015625, 1.0252830810546876, 1.024384033203125, 1.0238484497070313, 1.0244556884765625, 1.0247547607421874, 1.0238433227539063, 1.0240604248046874, 1.0238505249023437, 1.0239692993164062, 1.02425390625, 1.0245499267578124, 1.0240184326171875, 1.0239478149414063, 2.123869140625, 1.024112548828125, 1.0244178466796876, 1.024606201171875, 1.024331787109375, 1.024970703125, 1.0241341552734375, 1.0240655517578126, 1.0246256103515625, 1.024216064453125, 1.02445361328125, 1.0246973876953125, 1.02462158203125, 1.0241884765625, 1.024816162109375, 1.0244024658203126, 1.0240604248046874, 1.0241331787109376, 1.02439013671875, 1.024101318359375, 1.02460107421875, 1.0241126708984376, 1.0239580078125, 1.0260316162109375, 1.024153564453125, 1.024626708984375, 1.025184814453125, 1.024384033203125, 1.0247044677734376, 1.0242744140625, 1.0245263671875, 1.0240440673828124, 1.0243983154296874, 1.0242017822265626, 1.025048583984375, 1.02445263671875, 1.0241658935546876, 1.024067626953125, 1.0245570068359375, 1.0243236083984375, 1.0241505126953125, 1.0240552978515625, 1.0239324340820313, 1.02377880859375, 1.0250537109375, 1.0241873779296875, 1.0242979736328126, 1.0260521240234375, 1.0243245849609375, 1.0240809326171876, 1.024226318359375, 1.02384228515625, 1.0239518432617187, 1.02460107421875, 1.0242713623046875, 1.023963134765625, 1.0244454345703125, 1.02419970703125, 1.0243572998046875, 1.025553466796875, 1.0245345458984374, 1.0243184814453126, 1.02401123046875, 2.12367041015625, 1.0241259765625, 1.0242652587890626, 1.0244208984375, 1.0240543212890625, 1.0244075927734375, 1.0240450439453126, 1.024143310546875, 1.024236572265625, 1.0248232421875, 1.0240543212890625, 1.0240615234375, 1.0252451171875, 1.0245201416015626, 1.02453857421875, 1.024142333984375, 1.024111572265625, 1.0240758056640624, 1.0245355224609376, 1.0249923095703124, 1.0250772705078126, 1.025503173828125, 1.024585693359375, 1.025353759765625, 1.0251724853515625, 1.025302490234375, 1.0245509033203124, 1.02447412109375, 1.025395751953125, 1.0254765625, 1.025158203125, 1.02477001953125, 1.025292236328125, 1.024942138671875, 1.024954345703125, 1.024868408203125, 1.025138671875, 1.0246287841796875, 1.0256414794921875, 1.025081298828125, 1.0248232421875, 1.023910888671875, 1.0245693359375, 1.0238443603515626, 1.027009521484375, 1.0242816162109376, 1.0239989624023438, 1.0243726806640625, 1.0255267333984375, 1.024779296875, 1.0242899169921875, 1.0245653076171874, 1.0248477783203125, 1.0242303466796876, 1.025016845703125, 1.024173095703125, 1.0242969970703124, 1.024927734375, 1.0247515869140624, 1.02410546875, 1.0243441162109375, 1.024101318359375, 1.024564208984375, 2.12349853515625, 1.0242017822265626, 1.024427978515625, 1.0245703125, 1.0245938720703125, 1.025349609375, 1.02506396484375, 1.02497900390625, 1.0247310791015625, 1.0254193115234376, 1.0243072509765625, 1.02436865234375, 1.02426416015625, 1.024543701171875, 1.0246932373046875, 1.0242508544921876, 1.0247291259765625, 1.0247679443359374, 1.0248099365234375, 1.0243450927734374, 1.0237440185546876, 1.0239437255859376, 1.02409619140625, 1.0248201904296874, 1.024573486328125, 1.0249298095703125, 1.0247445068359375, 1.024711669921875, 1.0242037353515625, 1.025076171875, 1.0243707275390626, 1.0241719970703125, 1.0248785400390625, 1.0241934814453124, 1.0238074951171876, 1.0240870361328125, 1.0238863525390625, 1.024194580078125, 1.0243511962890626, 1.023699951171875, 1.0239385375976562, 1.0239354858398437, 1.0245355224609376, 1.0244464111328124, 1.02446484375, 1.0242958984375, 1.024153564453125, 1.0246614990234375, 1.0249881591796874, 1.0237716674804687, 1.02404296875, 1.0248468017578125, 1.0250352783203125, 1.024362548828125, 1.0239723510742187, 1.02413623046875, 1.0246624755859375, 1.029676025390625, 1.02432568359375, 1.0238546142578124, 1.0238064575195311, 1.0237429809570313, 1.02468603515625, 2.1217158203125, 1.0237695922851562, 1.024711669921875, 1.02485498046875, 1.0252984619140626, 1.0250035400390625, 1.0246727294921876, 1.024594970703125, 1.0253424072265624, 1.024754638671875, 1.0237081298828126, 1.024232421875, 1.024865234375, 1.0245396728515626, 1.02403173828125, 1.02407275390625, 1.02489599609375, 1.0252052001953125, 1.0250966796875, 1.023773681640625, 1.02441162109375, 1.02495947265625, 1.0239385375976562, 1.0248038330078124, 1.0244832763671874, 1.0240286865234376, 1.0238095092773438, 1.02410546875, 1.024607177734375, 1.0238873901367187, 1.0241719970703125, 1.0245919189453125, 1.0244515380859376, 1.0247874755859374, 1.0245714111328126, 1.02497998046875, 1.0246953125, 1.024564208984375, 1.02820654296875, 1.02466455078125, 1.02426220703125, 1.02436962890625, 1.0242867431640625, 1.0247608642578125, 1.023847412109375, 1.0239989624023438, 1.0239672241210938, 1.024320556640625, 1.0242447509765624, 1.02379931640625, 1.02389453125, 1.024280517578125, 1.0238443603515626, 1.0239979248046875, 1.02392626953125, 1.02389453125, 1.0241024169921875, 1.0241402587890625, 1.024089111328125, 1.0238341064453125, 1.0239702758789062, 1.0240225830078125, 1.02510595703125, 2.124275634765625, 1.02431640625, 1.0250086669921874, 1.0248448486328126, 1.0252420654296874, 1.024489501953125, 1.0244403076171875, 1.0246031494140626, 1.0247711181640624, 1.0237265625, 1.0236641235351562, 1.0239344482421875, 1.0241915283203125, 1.023867919921875, 1.02379931640625, 1.0237245483398438, 1.024006103515625, 1.023836181640625, 1.024227294921875, 1.02468603515625, 1.0239273071289063, 1.0240594482421874, 1.0241719970703125, 1.024067626953125, 1.023867919921875, 1.0238515014648437, 1.025666015625, 1.0245263671875, 1.024973876953125, 1.02382080078125, 1.0239006958007812, 1.0243861083984376, 1.0252000732421875, 1.025047607421875, 1.0245989990234374, 1.0247177734375, 1.023931396484375, 1.0238443603515626, 1.0247506103515625, 1.0237757568359376, 1.0241033935546875, 1.0239365844726562, 1.0244884033203125, 1.0237399291992189, 1.024362548828125, 1.024405517578125, 1.024089111328125, 1.0244864501953126, 1.02445263671875, 1.02447509765625, 1.024986083984375, 1.0248734130859376, 1.0243768310546875, 1.024090087890625, 1.024077880859375, 1.0242088623046874, 1.0244178466796876, 1.0243809814453124, 1.024343017578125, 1.024501708984375, 1.0247445068359375, 1.024690185546875, 1.0249298095703125, 2.128819091796875, 1.02439111328125, 1.0247823486328125, 1.0247977294921875, 1.02436767578125, 1.0251878662109375, 1.025364990234375, 1.0248714599609374, 1.024216064453125, 1.023978515625, 1.024204833984375, 1.024175048828125, 1.0241976318359376, 1.02432568359375, 1.024288818359375, 1.024332763671875, 1.02521142578125, 1.0247762451171876, 1.0244351806640626, 1.0237880249023437, 1.0244085693359375, 1.024279541015625, 1.024522216796875, 1.0245447998046875, 1.0244136962890624, 1.02485302734375, 1.0252779541015624, 1.0244249267578125, 1.0243072509765625, 1.024364501953125, 1.0249554443359374, 1.02441064453125, 1.02491748046875, 1.025059814453125, 1.0243563232421875, 1.024710693359375, 1.0254488525390626, 1.0243583984375, 1.024216064453125, 1.0244423828125, 1.024673828125, 1.0236497802734374, 1.0240399169921874, 1.02411572265625, 1.0241024169921875, 1.0244669189453126, 1.024669677734375, 1.0245672607421874, 1.0250875244140625, 1.0249993896484375, 1.024100341796875, 1.0239754028320311, 1.0242611083984374, 1.0236723022460938, 1.0240758056640624, 1.025005615234375, 1.0246123046875, 1.024385009765625, 1.024511962890625, 1.0244013671875, 1.0243123779296874, 1.0240665283203125, 1.0242939453125, 2.1268427734375, 1.0253404541015625, 1.0243931884765625, 1.0240972900390626, 1.0241934814453124, 1.0252298583984376, 1.024215087890625, 1.0247833251953125, 1.024521240234375, 1.023973388671875, 1.0242017822265626, 1.0258052978515626, 1.024546875, 1.0249359130859375, 1.025333251953125, 1.0250526123046875, 1.02436962890625, 1.0253035888671875, 1.0246348876953124, 1.0245201416015626, 1.02516015625, 1.0252349853515625, 1.02502294921875, 1.024257080078125, 1.0243369140625, 1.0239395751953124, 1.02425390625, 1.0249779052734376, 1.024067626953125, 1.0241177978515625, 1.0244976806640624, 1.025036376953125, 1.02519189453125, 1.02483349609375, 1.0254356689453126, 1.0247762451171876, 1.024606201171875, 1.0247310791015625, 1.024310302734375, 1.0240450439453126, 1.024733154296875, 1.0244976806640624, 1.0241402587890625, 1.0240235595703124, 1.025154052734375, 1.023847412109375, 1.0244290771484375, 1.0241136474609376, 1.0241822509765626, 1.0238689575195312, 1.0249298095703125, 1.0245478515625, 1.0244771728515625, 1.0245509033203124, 1.024611328125, 1.0240921630859374, 1.024890869140625, 1.0240205078125, 1.024279541015625, 1.0245560302734376, 1.0247720947265626, 1.024206787109375, 1.0244300537109374, 2.128713623046875, 1.024427001953125, 1.0249287109375, 1.025007568359375, 1.0242989501953126, 1.025701904296875, 1.0250526123046875, 1.024421875, 1.02443115234375, 1.0251038818359375, 1.024236572265625, 1.0237388916015624, 1.0245765380859375, 1.023847412109375, 1.0237470703125, 1.0244495849609374, 1.024459716796875, 1.024206787109375, 1.02432666015625, 1.024251953125, 1.0239293212890626, 1.024359375, 1.02411669921875, 1.0238505249023437, 1.024151611328125, 1.025427490234375, 1.0240450439453126, 1.024953369140625, 1.02416796875, 1.0247802734375, 1.0241229248046875, 1.024373779296875, 1.024141357421875, 1.0242396240234375, 1.02449560546875, 1.0251766357421874, 1.024015380859375, 1.0237122802734375, 1.0239969482421876, 1.0238975830078125, 1.0237726440429689, 1.02384130859375, 1.02431640625, 1.0243133544921874, 1.0245919189453125, 1.024943115234375, 1.0246246337890625, 1.024611328125, 1.024716796875, 1.024035888671875, 1.0247413330078126, 1.024973876953125, 1.024796630859375, 1.0246318359375, 1.0259609375, 1.0238064575195311, 1.0238228759765624, 1.02377880859375, 1.02478955078125, 1.02438916015625, 1.0245723876953126, 1.0243604736328125, 1.02432568359375]",tokens/s,0.9612963088545465,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen2-beta-14B,Qwen/Qwen2-beta-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,3956.158464,12732.33408,0.0,12085.886976,11337.370624,s,10,10.987366088867189,1.098736608886719,0.001768964710489349,1.0985178833007812,1.1007381103515623,1.101008154296875,1.101224189453125,"[1.1001180419921874, 1.1012781982421875, 1.0972359619140626, 1.0976529541015625, 1.097395263671875, 1.09569873046875, 1.0975172119140626, 1.0993828125, 1.1004088134765626, 1.1006781005859374]",tokens/s,232.99487605076595,kWh,1.295269936323166e-05,7.0975890679255816e-06,6.178085498018682e-05,8.183114341134406e-05,tokens/kWh,3128393.2904756567,MB,3956.158464,12732.33408,0.0,12085.886976,11686.804992,s,10,644.0755078125,64.40755078125,0.02043927244380867,64.406017578125,64.42852109374999,64.437791796875,64.445208359375,"[64.41949609375, 64.4215078125, 64.3984453125, 64.3965, 64.3855703125, 64.37932421875, 64.38755078125, 64.41358984375, 64.4470625, 64.4264609375]",tokens/s,0.9781461837288221,kWh,0.0007606077267395126,0.000416880381825722,0.003642305219397435,0.00481979332796267,tokens/kWh,13071.09988191758,,s,629,652.9415617065439,1.038062896194822,0.13045675837303516,1.0222479248046874,1.02322421875,1.0235658081054688,2.119142734375,"[1.0216539916992187, 1.0224589233398438, 1.0223789672851562, 1.0224598999023438, 1.0228684692382812, 1.0230978393554688, 1.0232852783203126, 1.0228684692382812, 1.0223707885742188, 1.0226544799804687, 1.0221055908203125, 1.022455810546875, 1.0223565063476563, 1.0220953369140624, 1.02287255859375, 1.0226544799804687, 1.0230180053710938, 1.0228029174804687, 1.022360595703125, 1.0219008178710938, 1.02221826171875, 1.0227445678710938, 1.0227865600585937, 1.0227056884765624, 1.0226657104492187, 1.0222479248046874, 1.0227005615234375, 1.0229114990234376, 1.0229288940429688, 1.0229381103515625, 1.0223861694335938, 1.0226493530273437, 1.0219724731445312, 1.02327294921875, 1.0225264892578125, 1.021897705078125, 1.0218536987304687, 1.0230333251953125, 1.0221629638671874, 1.022266357421875, 1.0232105102539062, 1.0222919921875, 1.0218916015625, 1.0232176513671876, 1.0224496459960937, 1.0222622680664062, 1.0223595581054687, 1.0233876342773438, 1.0225960693359375, 1.0226697998046874, 1.0229954833984376, 1.0221025390625, 1.0217092895507813, 1.0235074462890625, 1.0223370361328126, 1.0225868530273436, 1.0227589111328126, 1.0223380737304688, 1.0219468994140626, 1.02196533203125, 1.0226288452148438, 1.0231378173828125, 2.124905517578125, 1.0222356567382813, 1.021960205078125, 1.022509033203125, 1.0222704467773438, 1.0222008056640626, 1.0229309692382813, 1.0228899536132812, 1.0222418212890625, 1.022761962890625, 1.0235012817382811, 1.0224527587890626, 1.0228264770507813, 1.0230067138671874, 1.022603271484375, 1.022150634765625, 1.0228582153320311, 1.0227486572265625, 1.0225458984375, 1.023267822265625, 1.0228500366210938, 1.0230599975585937, 1.02280908203125, 1.0225940551757813, 1.0225387573242188, 1.0220328979492188, 1.0222684326171876, 1.0227660522460937, 1.0227394409179686, 1.0236610717773438, 1.0233159790039061, 1.02287158203125, 1.0244444580078125, 1.0236958618164063, 1.022750732421875, 1.0225233764648438, 1.0231818237304688, 1.0220595092773437, 1.0220052490234375, 1.0232534790039063, 1.0219489135742188, 1.02171337890625, 1.0222592163085937, 1.0224414672851563, 1.023140869140625, 1.0223493041992187, 1.0228746337890624, 1.0218157958984375, 1.02228173828125, 1.02193359375, 1.0220175170898438, 1.0220697631835938, 1.02232470703125, 1.0216427612304688, 1.0218946533203126, 1.0215690307617187, 1.022439453125, 1.02179736328125, 1.0224251098632813, 1.0220114135742187, 1.0223197021484376, 1.0227127685546875, 1.0225541381835936, 2.119232421875, 1.0225018920898437, 1.022055419921875, 1.0218680419921875, 1.0223175659179689, 1.0221229858398437, 1.0219366455078125, 1.0225018920898437, 1.0220114135742187, 1.0218792724609376, 1.0220257568359374, 1.0222643432617187, 1.0220001220703125, 1.0221516723632813, 1.0221260986328125, 1.0216611938476563, 1.0218567504882812, 1.0222837524414063, 1.022792724609375, 1.0223380737304688, 1.0225889282226563, 1.0225274658203125, 1.0219089965820312, 1.0221567993164062, 1.0223544311523438, 1.0222837524414063, 1.0220943603515624, 1.022055419921875, 1.0220431518554687, 1.0217420654296876, 1.022118896484375, 1.0222120971679687, 1.02186083984375, 1.0227660522460937, 1.0224302368164062, 1.0219202270507812, 1.0222950439453125, 1.022434326171875, 1.0217748413085936, 1.0216837158203125, 1.021929443359375, 1.0220933227539062, 1.0220226440429687, 1.0222172241210938, 1.021971435546875, 1.0217420654296876, 1.0220123901367189, 1.0219243774414062, 1.0223441772460937, 1.022044189453125, 1.022455810546875, 1.0223093872070312, 1.0240828857421875, 1.0233630981445312, 1.0228223876953124, 1.0220809936523438, 1.022814208984375, 1.0219632568359376, 1.0219100341796874, 1.021834228515625, 1.0227333374023437, 1.0218475341796875, 1.0219889526367187, 2.119702392578125, 1.0218157958984375, 1.0222847900390626, 1.0225991821289062, 1.0219386596679687, 1.0222909545898438, 1.0225694580078124, 1.02200732421875, 1.0223073120117188, 1.0221957397460937, 1.0222427978515625, 1.0221383666992188, 1.0221312255859376, 1.022075927734375, 1.0216263427734376, 1.0212208862304688, 1.02229296875, 1.0224076538085938, 1.023636474609375, 1.0234900512695313, 1.023056884765625, 1.0234654541015624, 1.0236641235351562, 1.02160791015625, 1.0215352172851562, 1.0216673583984375, 1.021822998046875, 1.0217000732421875, 1.022213134765625, 1.02185986328125, 1.0224988403320312, 1.022687255859375, 1.0223042602539063, 1.0217113647460938, 1.0225919799804688, 1.022482421875, 1.022055419921875, 1.0219806518554688, 1.023088623046875, 1.0223472900390624, 1.021929443359375, 1.0223493041992187, 1.0222120971679687, 1.0214901733398438, 1.0219417724609374, 1.0221721801757813, 1.0221773071289062, 1.0221752319335937, 1.0220165405273438, 1.0218884887695312, 1.0218833618164063, 1.0219745483398437, 1.0215157470703125, 1.0211686401367188, 1.0218997802734375, 1.0223994750976562, 1.0220114135742187, 1.022298095703125, 1.02169189453125, 1.0218690795898437, 1.021822998046875, 1.0221168823242188, 1.0219120483398438, 2.118912109375, 1.0223964233398437, 1.0216151123046875, 1.022171142578125, 1.0217677001953125, 1.0219642944335938, 1.0219970703125, 1.0222827758789061, 1.0216611938476563, 1.021549560546875, 1.0215291137695313, 1.0223114013671875, 1.022255126953125, 1.0225100708007813, 1.0217799682617188, 1.02200830078125, 1.0224179077148436, 1.0221260986328125, 1.0218782958984376, 1.0218884887695312, 1.0225499877929687, 1.022076904296875, 1.0219366455078125, 1.0222387084960938, 1.0217011108398437, 1.0217328491210937, 1.02257666015625, 1.0215403442382813, 1.0214144287109375, 1.0214072265625, 1.0221701049804688, 1.0217277221679688, 1.0222387084960938, 1.0218997802734375, 1.0216539916992187, 1.0216837158203125, 1.0219857788085938, 1.0216980590820313, 1.0217205810546874, 1.0214840087890624, 1.02214453125, 1.0218076171875, 1.0222633056640624, 1.0221834106445313, 1.0223790283203125, 1.0221107177734374, 1.022223388671875, 1.0214297485351562, 1.0217717895507812, 1.0216028442382812, 1.0226585693359376, 1.0215946044921875, 1.0224056396484376, 1.0217778930664063, 1.0220472412109376, 1.0222387084960938, 1.0221731567382812, 1.0224568481445313, 1.0223892211914063, 1.021928466796875, 1.0224486694335937, 1.0220421142578124, 1.022455810546875, 2.117295166015625, 1.0219315185546876, 1.0214788818359375, 1.0215782470703125, 1.021528076171875, 1.0218690795898437, 1.0219069213867187, 1.0218506469726563, 1.0224097290039063, 1.0219745483398437, 1.0218424072265626, 1.021686767578125, 1.0215823364257812, 1.0215782470703125, 1.021676513671875, 1.0219089965820312, 1.02179736328125, 1.0221465454101561, 1.0219561157226562, 1.021676513671875, 1.0220635986328126, 1.0227415161132813, 1.0216591186523438, 1.0214799194335937, 1.021432861328125, 1.0216959838867188, 1.021422607421875, 1.0225029296875, 1.0220267333984374, 1.021981689453125, 1.022392333984375, 1.0219192504882812, 1.021675537109375, 1.0215321655273437, 1.0215844116210937, 1.0223329467773437, 1.0216427612304688, 1.0223380737304688, 1.0223411254882813, 1.0224465942382812, 1.0220513305664063, 1.021770751953125, 1.0217267456054688, 1.0211819458007811, 1.0219478759765626, 1.0219786376953126, 1.0217195434570312, 1.02196533203125, 1.0216980590820313, 1.0216949462890625, 1.0220328979492188, 1.02213427734375, 1.0221680908203126, 1.02171337890625, 1.0219458618164063, 1.021823974609375, 1.022129150390625, 1.0220892333984375, 1.0217769165039063, 1.022181396484375, 1.0221486206054688, 1.0219581298828124, 1.0223165283203124, 2.1193359375, 1.021507568359375, 1.0222807006835937, 1.0218096923828126, 1.0217789306640626, 1.0224578857421875, 1.0224793701171875, 1.0222202758789063, 1.0224015502929686, 1.0220513305664063, 1.0227291870117188, 1.0218936157226564, 1.0227210083007812, 1.0215659790039062, 1.021507568359375, 1.0218588256835937, 1.0215946044921875, 1.0212608032226562, 1.0221383666992188, 1.021644775390625, 1.0220400390625, 1.0219151611328126, 1.0219458618164063, 1.0218168334960938, 1.0220155029296876, 1.0220830688476563, 1.0217257080078126, 1.0215249633789063, 1.0215782470703125, 1.0214522705078124, 1.021971435546875, 1.0216908569335938, 1.0222796630859374, 1.0217891845703124, 1.0220093383789062, 1.0221598510742187, 1.0217533569335937, 1.021591552734375, 1.021834228515625, 1.0217297973632813, 1.0220697631835938, 1.0221373291015625, 1.021681640625, 1.021823974609375, 1.022286865234375, 1.0229125366210938, 1.0224752807617188, 1.0218772583007814, 1.0224363403320313, 1.02299853515625, 1.021749267578125, 1.022983154296875, 1.0218731689453124, 1.0215403442382813, 1.022688232421875, 1.0222745361328125, 1.0221066284179687, 1.021812744140625, 1.0228449096679688, 1.0219735107421875, 1.0219089965820312, 1.0222427978515625, 1.0222807006835937, 2.121678955078125, 1.0220349731445313, 1.0220318603515626, 1.022308349609375, 1.0220912475585937, 1.0223062744140625, 1.0220667114257813, 1.0221168823242188, 1.0219243774414062, 1.0236375122070311, 1.0226729125976564, 1.022350341796875, 1.0221209716796875, 1.0218065795898437, 1.0220042114257812, 1.02285107421875, 1.0233190307617188, 1.0220626220703124, 1.0221178588867188, 1.0219304809570313, 1.0225458984375, 1.0225377197265626, 1.0221055908203125, 1.0222172241210938, 1.0222172241210938, 1.0232381591796875, 1.0236713256835936, 1.02211376953125, 1.0223145141601562, 1.0224046020507813, 1.0224127807617187, 1.0221404418945312, 1.0220615844726562, 1.0223544311523438, 1.0224671020507812, 1.0220564575195312, 1.02242919921875, 1.0219437866210936, 1.0222633056640624, 1.0225182495117187, 1.0220728149414062, 1.0216908569335938, 1.0225111083984375, 1.0221690673828125, 1.0233764038085937, 1.0224005126953124, 1.0228797607421876, 1.022582763671875, 1.0221362915039063, 1.0220574951171875, 1.0226176147460937, 1.023177734375, 1.022688232421875, 1.0220369873046875, 1.0229730224609375, 1.0224833984375, 1.0232719116210938, 1.0233231201171875, 1.0232688598632813, 1.0227711791992187, 1.02242919921875, 1.0218782958984376, 1.0227630004882813, 2.123450439453125, 1.0225910034179688, 1.022224365234375, 1.0221588745117187, 1.0227947387695313, 1.0223759155273437, 1.0220062866210937, 1.021928466796875, 1.0225244140625, 1.0222254028320312, 1.0220328979492188, 1.0240491943359376, 1.0243778076171874, 1.0222274780273437, 1.0225735473632813, 1.02270361328125, 1.0226463012695313, 1.0230241088867187, 1.023151123046875, 1.0230661010742188, 1.022940185546875, 1.0235166625976562, 1.0226575317382813, 1.0224097290039063, 1.02308349609375, 1.0247874755859374, 1.0235719604492188, 1.0233917236328125, 1.023362060546875, 1.0236497802734374, 1.0237828979492187, 1.0238555908203124, 1.024322509765625, 1.0232586059570312, 1.0231644287109376, 1.023604736328125, 1.0232422485351562, 1.0228469848632813, 1.0233712768554688, 1.0233978881835937, 1.0237593383789063, 1.02187109375, 1.0220492553710938, 1.0218803100585938, 1.023657958984375, 1.0220001220703125, 1.0220830688476563, 1.0218035278320312, 1.02250390625, 1.02404296875, 1.0239928588867186, 1.0232227783203125, 1.0231644287109376, 1.0232197265625, 1.0240625, 1.0235238647460938, 1.022983154296875, 1.0219345703125, 1.0223042602539063, 1.022192626953125, 1.0235565795898438, 1.0230742797851562, 1.0222684326171876, 2.122977294921875, 1.021991943359375, 1.0222418212890625, 1.022287841796875, 1.0231869506835938, 1.0230241088867187, 1.022814208984375, 1.022866455078125, 1.0224425048828125, 1.0229033203125, 1.0226390991210939, 1.0222633056640624, 1.0229155883789063, 1.0221915893554687, 1.0231552124023438, 1.0233661499023436, 1.0227711791992187, 1.0222151489257814, 1.0229217529296875, 1.0231818237304688, 1.0223114013671875, 1.0220543823242187, 1.02232373046875, 1.0222356567382813, 1.0230077514648437, 1.0234613647460937, 1.0228264770507813, 1.0229319458007813, 1.024480224609375, 1.0225131225585937, 1.0220369873046875, 1.0220902099609375, 1.0221383666992188, 1.0218895263671874, 1.0225633544921875, 1.0228480224609375, 1.0221834106445313, 1.0225643310546875, 1.0235330810546874, 1.02322998046875, 1.0229186401367187, 1.0226974487304688, 1.0228060302734374, 1.022993408203125, 1.023494140625, 1.0230015869140625, 1.0222899169921875, 1.0222520141601563, 1.0227855224609375, 1.0226903076171876, 1.022087158203125, 1.0217963256835938, 1.0218424072265626, 1.0221260986328125, 1.0221178588867188, 1.0227445678710938, 1.023115234375, 1.0224527587890626, 1.0226483154296875, 1.022645263671875, 1.022792724609375, 1.0225899658203126, 1.0227425537109376]",tokens/s,0.9633327649660273,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-30b,facebook/opt-30b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,5945.012224,19933.954048,0.0,19287.506944,18376.2688,s,10,24.42857763671875,2.442857763671875,0.001982296286527604,2.4421226806640624,2.4459630615234373,2.446148400878906,2.4462966723632813,"[2.44311767578125, 2.441929443359375, 2.441054443359375, 2.44091455078125, 2.441591064453125, 2.440709716796875, 2.44231591796875, 2.446333740234375, 2.445921875, 2.444689208984375]",tokens/s,104.79529500530754,kWh,2.881985050108698e-05,1.5794228899012522e-05,0.00013673797050139935,0.00018135204990149886,tokens/kWh,1411619.0036950014,MB,5946.937344,19933.954048,0.0,19287.506944,18871.985152,s,10,1453.1662812499999,145.316628125,0.01677016451823821,145.3091796875,145.3392140625,145.34826328125,145.35550265625,"[145.31590625, 145.308921875, 145.3094375, 145.306953125, 145.3573125, 145.337203125, 145.304828125, 145.303359375, 145.319953125, 145.30240625]",tokens/s,0.4335360709430169,kWh,0.0017155244359870752,0.0009402592492990879,0.00808164824309121,0.010737431928377374,tokens/kWh,5867.324740238933,,s,629,1472.8442895507828,2.3415648482524345,0.29007180854476533,2.306472900390625,2.307840771484375,2.308159619140625,4.7478279296874994,"[2.30830078125, 2.30712939453125, 2.305491943359375, 2.305958984375, 2.306070556640625, 2.306164794921875, 2.306714599609375, 2.306167724609375, 2.30651708984375, 2.306343994140625, 2.306281494140625, 2.3058994140625, 2.306171875, 2.305965087890625, 2.3060673828125, 2.305934326171875, 2.306259033203125, 2.30677197265625, 2.30823828125, 2.3065693359375, 2.3062998046875, 2.3065107421875, 2.305712158203125, 2.3060634765625, 2.30664404296875, 2.30611865234375, 2.305919921875, 2.306153564453125, 2.30573876953125, 2.307145751953125, 2.307577880859375, 2.30742529296875, 2.3060234375, 2.306620361328125, 2.30689892578125, 2.306671630859375, 2.30721337890625, 2.306621337890625, 2.30584423828125, 2.306515869140625, 2.306025390625, 2.307158935546875, 2.307966064453125, 2.3080498046875, 2.30725927734375, 2.30751123046875, 2.30584326171875, 2.306826171875, 2.30643505859375, 2.306669677734375, 2.3066572265625, 2.306987060546875, 2.3064228515625, 2.30631640625, 2.3060048828125, 2.3066552734375, 2.306758544921875, 2.306385986328125, 2.306724853515625, 2.30693994140625, 2.3067626953125, 2.306996337890625, 4.747927734375, 2.3060458984375, 2.306315185546875, 2.305916015625, 2.30605615234375, 2.30666455078125, 2.306251708984375, 2.306472900390625, 2.306552734375, 2.30586376953125, 2.3063408203125, 2.306269287109375, 2.306483154296875, 2.3060849609375, 2.306608154296875, 2.306598876953125, 2.306947021484375, 2.306627685546875, 2.30656298828125, 2.306156494140625, 2.305994873046875, 2.305953857421875, 2.3070966796875, 2.307031005859375, 2.306716552734375, 2.30727587890625, 2.305818603515625, 2.30561181640625, 2.306303955078125, 2.30622607421875, 2.30601513671875, 2.306629638671875, 2.306018310546875, 2.307220458984375, 2.307166259765625, 2.30818310546875, 2.30658251953125, 2.306062255859375, 2.3065908203125, 2.3057265625, 2.306356201171875, 2.30654150390625, 2.306720703125, 2.306249755859375, 2.307949462890625, 2.307349609375, 2.30730859375, 2.306469970703125, 2.306974609375, 2.30639501953125, 2.30609716796875, 2.30603369140625, 2.306250732421875, 2.306716552734375, 2.306428955078125, 2.3060244140625, 2.306080810546875, 2.306798583984375, 2.306280517578125, 2.306587646484375, 2.306314208984375, 2.3067607421875, 2.307210205078125, 4.74750048828125, 2.30618017578125, 2.306073486328125, 2.306304931640625, 2.3058740234375, 2.306080810546875, 2.3063828125, 2.30624560546875, 2.305426513671875, 2.306010009765625, 2.3058330078125, 2.306188232421875, 2.305746826171875, 2.30639501953125, 2.305995849609375, 2.305857421875, 2.305976318359375, 2.306387939453125, 2.307175537109375, 2.3076884765625, 2.30548486328125, 2.305490966796875, 2.3058544921875, 2.30586474609375, 2.306364501953125, 2.30583203125, 2.306301025390625, 2.306060302734375, 2.306914306640625, 2.308274169921875, 2.307685302734375, 2.3084912109375, 2.306903076171875, 2.30611669921875, 2.306157470703125, 2.30603466796875, 2.306026611328125, 2.30809912109375, 2.30824658203125, 2.30763623046875, 2.30674853515625, 2.306740234375, 2.30609521484375, 2.305721435546875, 2.30651806640625, 2.307560546875, 2.30752880859375, 2.306269287109375, 2.30721630859375, 2.306631591796875, 2.30655908203125, 2.3057724609375, 2.30613818359375, 2.306503662109375, 2.306617431640625, 2.307072021484375, 2.30647802734375, 2.305734619140625, 2.3066552734375, 2.306449462890625, 2.30620361328125, 2.30637158203125, 2.306595947265625, 4.74802880859375, 2.306186279296875, 2.305531982421875, 2.305669189453125, 2.306135986328125, 2.3058759765625, 2.306019287109375, 2.306431884765625, 2.306522216796875, 2.30567529296875, 2.3055830078125, 2.306220947265625, 2.30667578125, 2.306291748046875, 2.306946044921875, 2.30719287109375, 2.3062763671875, 2.30605517578125, 2.3061484375, 2.306335693359375, 2.30618115234375, 2.30607666015625, 2.30662646484375, 2.30637158203125, 2.30767724609375, 2.308170654296875, 2.30727978515625, 2.306090087890625, 2.3074775390625, 2.30679248046875, 2.308010009765625, 2.306175048828125, 2.3065908203125, 2.306902099609375, 2.307230712890625, 2.30744677734375, 2.307564453125, 2.306438232421875, 2.306532470703125, 2.306220947265625, 2.306926513671875, 2.30791162109375, 2.30597021484375, 2.30626416015625, 2.306064453125, 2.305848388671875, 2.30631005859375, 2.305681396484375, 2.306680908203125, 2.3066787109375, 2.306912353515625, 2.30727587890625, 2.30600390625, 2.306438232421875, 2.306324462890625, 2.305838134765625, 2.305875, 2.305974365234375, 2.30620166015625, 2.306227294921875, 2.3058759765625, 2.306018310546875, 2.306186279296875, 4.7475712890625, 2.3056015625, 2.305901611328125, 2.30658251953125, 2.307598388671875, 2.307072998046875, 2.30600390625, 2.308192138671875, 2.307829833984375, 2.30786865234375, 2.307883056640625, 2.306185302734375, 2.305741943359375, 2.30852392578125, 2.307919921875, 2.30782763671875, 2.30862841796875, 2.308886474609375, 2.308350830078125, 2.308662353515625, 2.308884521484375, 2.30721337890625, 2.307564453125, 2.307115966796875, 2.307040283203125, 2.30601025390625, 2.307072998046875, 2.305795166015625, 2.3058740234375, 2.306260009765625, 2.3076474609375, 2.3075, 2.305462158203125, 2.305919921875, 2.306641845703125, 2.308442138671875, 2.30645654296875, 2.306227294921875, 2.306621337890625, 2.306598876953125, 2.3067822265625, 2.306314208984375, 2.307274658203125, 2.306165771484375, 2.30636962890625, 2.306336669921875, 2.3068681640625, 2.307556396484375, 2.307939208984375, 2.306641845703125, 2.308420654296875, 2.30788720703125, 2.30803857421875, 2.307458984375, 2.308116455078125, 2.3085107421875, 2.30814306640625, 2.309021728515625, 2.30691845703125, 2.3075400390625, 2.30881787109375, 2.30786962890625, 2.308601806640625, 4.747927734375, 2.30775, 2.307747802734375, 2.306754638671875, 2.30704638671875, 2.30609521484375, 2.305871826171875, 2.305828857421875, 2.306450439453125, 2.3074384765625, 2.30639404296875, 2.306155517578125, 2.30721240234375, 2.3061943359375, 2.306260986328125, 2.3063408203125, 2.305946533203125, 2.30626708984375, 2.3059384765625, 2.306785400390625, 2.30590576171875, 2.307837890625, 2.308127685546875, 2.307727294921875, 2.30624755859375, 2.30803759765625, 2.308209716796875, 2.306083740234375, 2.308106201171875, 2.30796484375, 2.307409912109375, 2.307322998046875, 2.3070556640625, 2.30601416015625, 2.30601220703125, 2.30685400390625, 2.3067841796875, 2.3062333984375, 2.30788818359375, 2.3078798828125, 2.307981201171875, 2.308041748046875, 2.307943359375, 2.3078779296875, 2.306785400390625, 2.306217041015625, 2.30694189453125, 2.3059833984375, 2.30669921875, 2.30706787109375, 2.307828857421875, 2.3062119140625, 2.30769970703125, 2.30689990234375, 2.306872314453125, 2.30632763671875, 2.307541015625, 2.307304443359375, 2.306280517578125, 2.305987548828125, 2.3064677734375, 2.30799365234375, 2.30702490234375, 4.74871826171875, 2.306419677734375, 2.30594775390625, 2.305490966796875, 2.3053935546875, 2.306144287109375, 2.305890380859375, 2.30618310546875, 2.30565380859375, 2.305838134765625, 2.3064453125, 2.306163818359375, 2.305967041015625, 2.30632958984375, 2.30599365234375, 2.30660205078125, 2.306093017578125, 2.30550927734375, 2.3062958984375, 2.30859375, 2.307629150390625, 2.305786865234375, 2.30675244140625, 2.30698291015625, 2.307533935546875, 2.306155517578125, 2.30588916015625, 2.30687744140625, 2.30660205078125, 2.30624658203125, 2.30635205078125, 2.30655078125, 2.30631005859375, 2.307313720703125, 2.306417724609375, 2.306740234375, 2.306871337890625, 2.30619140625, 2.3060859375, 2.305734619140625, 2.3065068359375, 2.30648828125, 2.306944091796875, 2.30637060546875, 2.3063173828125, 2.305910888671875, 2.306567138671875, 2.305996826171875, 2.306691162109375, 2.306502685546875, 2.306821044921875, 2.306130859375, 2.30672900390625, 2.307013671875, 2.306767822265625, 2.306404296875, 2.30620458984375, 2.306543701171875, 2.30641552734375, 2.30656298828125, 2.306662353515625, 2.306578369140625, 2.30736083984375, 4.75277099609375, 2.306112548828125, 2.305987548828125, 2.306069580078125, 2.305553466796875, 2.305196044921875, 2.306450439453125, 2.306493408203125, 2.306361328125, 2.305670166015625, 2.305954833984375, 2.30636328125, 2.305942626953125, 2.305912841796875, 2.3060244140625, 2.306492431640625, 2.306491455078125, 2.3065087890625, 2.306021484375, 2.30662255859375, 2.306044921875, 2.305946533203125, 2.306109375, 2.3060673828125, 2.30651708984375, 2.306368408203125, 2.306166748046875, 2.305617919921875, 2.3070966796875, 2.30624462890625, 2.30612060546875, 2.30616259765625, 2.306585693359375, 2.3061943359375, 2.305982421875, 2.30618212890625, 2.30691015625, 2.307801025390625, 2.30774169921875, 2.305775634765625, 2.30598974609375, 2.307757080078125, 2.307713134765625, 2.3067412109375, 2.306273193359375, 2.306452392578125, 2.30667578125, 2.306123779296875, 2.307871826171875, 2.3065498046875, 2.306324462890625, 2.3063818359375, 2.306296875, 2.306356201171875, 2.306536376953125, 2.30616162109375, 2.306343994140625, 2.3068017578125, 2.307167236328125, 2.3066982421875, 2.3068271484375, 2.30626611328125, 2.306765869140625, 4.75146533203125, 2.305406005859375, 2.306103271484375, 2.306335693359375, 2.306105224609375, 2.305650634765625, 2.305935302734375, 2.3065302734375, 2.306123779296875, 2.306839599609375, 2.306525146484375, 2.3065908203125, 2.3057275390625, 2.3065927734375, 2.306326416015625, 2.305699951171875, 2.306766845703125, 2.306298828125, 2.306177978515625, 2.306017333984375, 2.3064453125, 2.307205078125, 2.3069912109375, 2.30652099609375, 2.307852294921875, 2.30791357421875, 2.307249267578125, 2.30723388671875, 2.308533203125, 2.307859375, 2.307365966796875, 2.30736474609375, 2.306230224609375, 2.306595947265625, 2.306797607421875, 2.306817138671875, 2.307945556640625, 2.306934814453125, 2.306532470703125, 2.306154541015625, 2.306238525390625, 2.306093994140625, 2.306994140625, 2.30653125, 2.30648828125, 2.30630810546875, 2.306410400390625, 2.30582177734375, 2.30702392578125, 2.306610107421875, 2.30685888671875, 2.30723388671875, 2.307060791015625, 2.30736083984375, 2.307283935546875, 2.306974609375, 2.307143798828125, 2.30702685546875, 2.307147705078125, 2.306912353515625, 2.306466796875, 2.30637060546875, 2.305808349609375, 4.75029296875, 2.3062138671875, 2.305847412109375, 2.305751953125, 2.306163818359375, 2.305503173828125, 2.306188232421875, 2.305998779296875, 2.306451416015625, 2.305594482421875, 2.30626513671875, 2.30557080078125, 2.306021484375, 2.305986572265625, 2.30660498046875, 2.306206787109375, 2.305995849609375, 2.306838623046875, 2.306428955078125, 2.306533447265625, 2.306038818359375, 2.306441162109375, 2.306575439453125, 2.30774267578125, 2.30721435546875, 2.307527587890625, 2.30657421875, 2.3058974609375, 2.30637255859375, 2.305725341796875, 2.30624560546875, 2.306049072265625, 2.3060224609375, 2.306079833984375, 2.306041748046875, 2.305879150390625, 2.30664404296875, 2.306240478515625, 2.30664697265625, 2.306007080078125, 2.306346923828125, 2.3059384765625, 2.305849365234375, 2.306935791015625, 2.306491455078125, 2.306400146484375, 2.3062744140625, 2.306458740234375, 2.306747314453125, 2.307177490234375, 2.306862060546875, 2.306808837890625, 2.306905029296875, 2.30681298828125, 2.306693115234375, 2.30618017578125, 2.30659375, 2.306906005859375, 2.30719482421875, 2.306628662109375, 2.306923583984375, 2.306654296875, 2.3069306640625]",tokens/s,0.4270648326252095,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-1.3B,,cuda,0,42,,,,,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,,MB,1554.096128,1957.167104,0.0,1310.72,1163.82464,s,10,1.286997444152832,0.1286997444152832,0.0010709154419692667,0.12870168304443358,0.13024390106201172,0.13026507034301757,0.13028200576782226,"[0.13028623962402344, 0.13023919677734375, 0.1282391662597656, 0.12805363464355468, 0.1269244155883789, 0.12729046630859375, 0.12865869140625, 0.1287446746826172, 0.1295413055419922, 0.1290196533203125]",tokens/s,1989.1259393177147,kWh,1.4998026454163673e-06,8.218217431655106e-07,6.2844740852399835e-06,8.60609847382186e-06,tokens/kWh,29746347.985524923,MB,1554.096128,1959.264256,0.0,1312.817152,1232.774656,s,10,75.16664355468751,7.516664355468751,0.019324999879587898,7.517852294921875,7.5454459960937506,7.548604345703125,7.551131025390625,"[7.5517626953125, 7.5180322265625, 7.52255224609375, 7.52390966796875, 7.544744140625, 7.51767236328125, 7.49461669921875, 7.49106689453125, 7.50185205078125, 7.5004345703125]",tokens/s,8.3813772999142,kWh,8.89009176429398e-05,4.872418928564946e-05,0.00036540068437377195,0.0005030257913023612,tokens/kWh,125242.08716393956,,s,629,76.20247959136968,0.12114861620249542,0.01533479707935131,0.1192816619873047,0.12045271148681641,0.12072119903564453,0.24720068908691412,"[0.12236902618408203, 0.12192870330810547, 0.12110336303710938, 0.12021862030029297, 0.12056371307373047, 0.11989708709716797, 0.11929497528076172, 0.11829862213134766, 0.11859661102294922, 0.11861504364013672, 0.11821469116210938, 0.11968099212646484, 0.11959193420410157, 0.12045926666259765, 0.12021043395996094, 0.1199452133178711, 0.11970047760009765, 0.12003225708007813, 0.120163330078125, 0.11992678070068359, 0.11977011108398437, 0.11847372436523437, 0.11856588745117187, 0.11883724975585938, 0.11842662048339844, 0.11851676940917968, 0.11948745727539062, 0.11911475372314453, 0.119552001953125, 0.1215068130493164, 0.120257568359375, 0.1204623031616211, 0.12012236785888672, 0.12004659271240234, 0.12008038330078125, 0.12001689910888672, 0.11995954895019531, 0.12004761505126953, 0.12005683135986328, 0.12010189056396485, 0.12003635406494141, 0.11968307495117188, 0.12004863739013671, 0.12036300659179687, 0.11988992309570312, 0.1200558090209961, 0.1202534408569336, 0.12022169494628906, 0.12032921600341796, 0.12026265716552734, 0.12077056121826171, 0.12018585968017578, 0.11901235198974609, 0.1193861083984375, 0.11910348510742187, 0.11925606536865234, 0.11975270080566407, 0.12020941162109375, 0.11999846649169922, 0.12033843231201172, 0.11914854431152344, 0.12003942108154297, 0.2506158142089844, 0.12027391815185547, 0.11959091186523438, 0.11991551971435546, 0.1192816619873047, 0.11801292419433594, 0.11855052947998047, 0.11803955078125, 0.11798016357421875, 0.11803241729736329, 0.11873174285888671, 0.12019200134277344, 0.11996057891845703, 0.11952947235107422, 0.11857612609863281, 0.11973939514160156, 0.11993702697753907, 0.11889049530029297, 0.11839385223388672, 0.11943730926513672, 0.11985203552246093, 0.11852082824707032, 0.11822898864746094, 0.11862739562988281, 0.11842758178710938, 0.11852185821533204, 0.1185054702758789, 0.11835596466064453, 0.11835699462890625, 0.11841331481933594, 0.11834674835205078, 0.11860070037841797, 0.12062003326416015, 0.11992473602294922, 0.12043059539794922, 0.12157746887207031, 0.12022579193115235, 0.11964927673339844, 0.12050534057617188, 0.12049407958984375, 0.12015615844726563, 0.11910553741455078, 0.12046438598632812, 0.1204510726928711, 0.12023603057861328, 0.11975885009765624, 0.12176998138427735, 0.12023705291748046, 0.12054937744140624, 0.11995750427246094, 0.12132864379882813, 0.11925917053222657, 0.1186283187866211, 0.11835699462890625, 0.11851776123046875, 0.11839794921875, 0.11879730987548828, 0.11951107025146485, 0.11866006469726563, 0.11900109100341796, 0.11836723327636718, 0.11853414154052734, 0.11843583679199218, 0.24882893371582032, 0.11982233428955077, 0.11939225769042969, 0.12049919891357422, 0.12002508544921875, 0.1198919677734375, 0.12023910522460937, 0.11994217681884765, 0.11870614624023437, 0.11836006164550782, 0.11821260833740234, 0.11851570892333985, 0.11971686553955078, 0.12000460815429688, 0.11984076690673828, 0.11931136322021485, 0.11905126190185547, 0.11986329650878906, 0.11975373077392579, 0.11980595397949219, 0.1196267547607422, 0.11956735992431641, 0.12024422454833984, 0.12106342315673828, 0.11977728271484375, 0.1197127685546875, 0.11865599822998046, 0.11831404876708984, 0.11899692535400391, 0.11852082824707032, 0.11829145812988281, 0.11852595520019531, 0.11846348571777343, 0.11836518096923829, 0.11844812774658203, 0.11833757019042969, 0.11830780792236328, 0.11964620971679688, 0.11897650909423828, 0.11840921783447265, 0.11968511962890625, 0.11841024017333984, 0.11969535827636718, 0.11970559692382812, 0.12029644775390624, 0.12001078033447266, 0.11961955261230468, 0.12051967620849609, 0.11861196899414063, 0.11842150115966797, 0.11837133026123046, 0.12028313446044922, 0.11990322875976563, 0.1188751983642578, 0.12033939361572266, 0.11918643188476563, 0.1187583999633789, 0.12003942108154297, 0.11934515380859376, 0.12007730865478515, 0.12025138854980469, 0.12019916534423829, 0.12015615844726563, 0.24848793029785157, 0.11997798156738282, 0.11991763305664062, 0.12037625885009766, 0.12008345794677734, 0.11944652557373046, 0.12011827087402344, 0.12008448028564453, 0.11987149047851563, 0.12019404602050782, 0.12016851043701172, 0.11977823638916016, 0.11987865447998047, 0.11973426818847656, 0.12008038330078125, 0.11942092895507812, 0.11997491455078126, 0.11880242919921875, 0.11979673767089843, 0.1197127685546875, 0.11939532470703125, 0.11976703643798828, 0.11997494506835937, 0.12019094085693359, 0.12064665222167968, 0.12001996612548828, 0.12060364532470703, 0.12040294647216797, 0.11997388458251954, 0.11948646545410156, 0.11998822021484375, 0.11993907165527344, 0.11955097961425781, 0.11953561401367188, 0.11927756500244141, 0.11897036743164062, 0.11999334716796875, 0.11905126190185547, 0.11846553802490234, 0.11841843414306641, 0.11854847717285157, 0.11835596466064453, 0.11830995178222656, 0.11828428649902344, 0.1203987808227539, 0.11818905639648437, 0.1193707504272461, 0.11814604949951171, 0.11847782135009766, 0.1183078384399414, 0.11799350738525391, 0.11804771423339844, 0.11817372894287109, 0.11848700714111328, 0.11852082824707032, 0.11849932861328125, 0.12020735931396484, 0.120089599609375, 0.12004761505126953, 0.12012134552001953, 0.11869286346435547, 0.11855974578857421, 0.11863142395019531, 0.24572006225585938, 0.1183477783203125, 0.11846041870117187, 0.11811737823486328, 0.1181470718383789, 0.11830067443847657, 0.11928985595703125, 0.11911065673828125, 0.1185771484375, 0.11835187530517578, 0.12018994903564453, 0.12008550262451172, 0.11990534210205078, 0.1200219497680664, 0.12004659271240234, 0.11994931030273437, 0.12035686492919923, 0.12041522979736329, 0.12053298950195312, 0.12031999969482422, 0.12058419036865234, 0.12013158416748047, 0.12075929260253906, 0.11833036804199219, 0.1206794204711914, 0.12085862731933594, 0.12161433410644532, 0.11919667053222656, 0.1187041244506836, 0.11844300842285156, 0.11833650970458984, 0.11833856201171875, 0.1186355209350586, 0.11820236968994141, 0.12034559631347656, 0.12019712066650391, 0.12016947174072265, 0.11850035095214843, 0.12005375671386719, 0.12051763153076171, 0.11887615966796874, 0.11807539367675782, 0.11823411560058594, 0.11905741119384766, 0.12032307434082032, 0.1189775390625, 0.12029132843017579, 0.12084019470214843, 0.12052992248535156, 0.1207357406616211, 0.12054220581054688, 0.12078284454345703, 0.12055244445800781, 0.12059852600097656, 0.12064358520507812, 0.12082486724853515, 0.1207326431274414, 0.12082278442382813, 0.12053196716308594, 0.12073065948486328, 0.12090262603759766, 0.12064870452880859, 0.12058112335205078, 0.24601292419433593, 0.11817881774902343, 0.11808255767822265, 0.11800780487060547, 0.118255615234375, 0.11830067443847657, 0.11829145812988281, 0.11941478729248046, 0.12005375671386719, 0.1198202896118164, 0.12005785369873047, 0.11806412506103516, 0.11867750549316407, 0.11829145812988281, 0.11833548736572265, 0.11830989074707031, 0.11806822204589844, 0.11833241271972657, 0.11837747192382812, 0.11845120239257813, 0.11834572601318359, 0.11841228485107422, 0.11827609252929687, 0.11805286407470703, 0.11896729278564454, 0.11836313629150391, 0.11981619262695313, 0.12392044830322266, 0.1205656967163086, 0.12015821075439453, 0.11990016174316406, 0.12003533172607422, 0.11934003448486329, 0.1199974365234375, 0.1193154525756836, 0.11959500885009766, 0.11863040161132812, 0.1188106231689453, 0.12067327880859376, 0.11914444732666016, 0.11830477142333984, 0.11853721618652344, 0.1185269775390625, 0.11941990661621094, 0.11858329772949219, 0.1186693115234375, 0.12036608123779297, 0.12004557037353515, 0.1200558090209961, 0.12050841522216797, 0.12022681427001954, 0.12039065551757812, 0.12011212921142578, 0.12010495758056641, 0.12036608123779297, 0.1201817626953125, 0.12043673706054688, 0.12033843231201172, 0.12007526397705078, 0.12000972747802735, 0.12006505584716796, 0.11996975708007812, 0.11998822021484375, 0.2494791717529297, 0.12030156707763671, 0.11901952362060547, 0.11824127960205078, 0.1192816619873047, 0.11887308502197266, 0.11826175689697266, 0.11869388580322265, 0.11826687622070313, 0.11836211395263672, 0.11810918426513672, 0.11887104034423829, 0.11972198486328126, 0.11960012817382812, 0.1192468490600586, 0.118002685546875, 0.11875942230224609, 0.11941478729248046, 0.11890585327148437, 0.11843382263183594, 0.1191526107788086, 0.11959500885009766, 0.11970668792724609, 0.11957036590576171, 0.11933695983886719, 0.11909327697753906, 0.11842966461181641, 0.11905228424072266, 0.11911888122558593, 0.11894780731201172, 0.11848089599609375, 0.1181890869140625, 0.11840406036376953, 0.1190297622680664, 0.11982848358154297, 0.11972096252441407, 0.11864371490478516, 0.11924582672119141, 0.11929804992675781, 0.11825459289550781, 0.11951718139648437, 0.11916287994384765, 0.11947417449951171, 0.11800064086914062, 0.11845017242431641, 0.11818089294433594, 0.11930210876464843, 0.1199810562133789, 0.11853517150878906, 0.11870310211181641, 0.11945164489746093, 0.11844096374511719, 0.1184194564819336, 0.118508544921875, 0.11926220703125, 0.12017356872558593, 0.11892838287353516, 0.1193021469116211, 0.11871849822998047, 0.11844911956787109, 0.11832115173339844, 0.11867545318603516, 0.11844403076171875, 0.24766259765625, 0.11916802978515625, 0.12123133087158203, 0.12028313446044922, 0.11920588684082031, 0.11884953308105468, 0.11862528228759765, 0.11994931030273437, 0.11886182403564453, 0.11885363006591797, 0.12013670349121094, 0.11913011169433593, 0.11992473602294922, 0.11888435363769531, 0.11946598052978516, 0.11882707214355469, 0.1207070083618164, 0.11865702056884765, 0.11839590454101563, 0.11839078521728516, 0.11947724914550781, 0.11919769287109375, 0.11886080169677735, 0.11848601531982422, 0.11876966094970703, 0.11810099029541016, 0.11880960083007812, 0.11817369842529298, 0.11927142333984375, 0.11965235137939453, 0.118761474609375, 0.11863654327392578, 0.11901439666748047, 0.11827814483642578, 0.11898368072509766, 0.11854233551025391, 0.11844915008544922, 0.11848397064208985, 0.11839078521728516, 0.11844608306884766, 0.11859455871582031, 0.11865190124511718, 0.11820441436767579, 0.11907686614990234, 0.11970867156982422, 0.11921715545654298, 0.11824230194091796, 0.11921817779541016, 0.11916390228271484, 0.11929804992675781, 0.11828326416015625, 0.11838361358642578, 0.11848499298095704, 0.11860991668701172, 0.11808153533935548, 0.11826588439941406, 0.11823715209960937, 0.11888639831542969, 0.11884031677246094, 0.11812454223632812, 0.11833446502685546, 0.11847577667236328, 0.11946905517578126, 0.24866712951660155, 0.11849318695068359, 0.11980188751220704, 0.11915773010253906, 0.11846041870117187, 0.1203589096069336, 0.1205381088256836, 0.11978342437744141, 0.11976601409912109, 0.11947417449951171, 0.11962163543701172, 0.11930316925048828, 0.11854541015625, 0.11897344207763672, 0.11885158538818359, 0.11875020599365234, 0.1181839370727539, 0.11931443023681641, 0.11823616027832032, 0.11900313568115234, 0.11816345977783203, 0.11854847717285157, 0.11989708709716797, 0.11902668762207032, 0.11941683197021484, 0.11941580963134765, 0.11852185821533204, 0.11982540893554687, 0.11974553680419922, 0.11874918365478515, 0.11918950653076171, 0.11970355224609375, 0.11946189117431641, 0.11857100677490234, 0.11914035034179687, 0.11940966033935548, 0.1185689926147461, 0.11831394958496094, 0.11820441436767579, 0.11945574188232422, 0.11997388458251954, 0.1203240966796875, 0.11903180694580077, 0.11818495941162109, 0.11810201263427735, 0.11788288116455078, 0.11807539367675782, 0.11814604949951171, 0.11880038452148438, 0.11919872283935547, 0.11881983947753906, 0.11975987243652343, 0.11975373077392579, 0.11953568267822266, 0.12003424072265625, 0.11991961669921875, 0.12044185638427735, 0.11880550384521485, 0.11889356994628907, 0.11842253112792969, 0.11843379211425781, 0.11815936279296875, 0.11811634826660156, 0.24846543884277345, 0.11927139282226562, 0.11798732757568359, 0.11792998504638671, 0.11840512084960937, 0.11863859558105469, 0.11815321350097656, 0.11805184173583984, 0.11871437072753906, 0.11836109161376954, 0.1180794906616211, 0.11815424346923828, 0.1181317138671875, 0.11853004455566406, 0.11842150115966797, 0.11837542724609375, 0.11854847717285157, 0.11832217407226563, 0.11862220764160156, 0.11864985656738282, 0.11943526458740235, 0.11942912292480469, 0.11993395233154297, 0.12021145629882812, 0.12013568115234376, 0.12069990539550782, 0.12175667572021484, 0.12003942108154297, 0.11987763214111329, 0.11907788848876953, 0.1194076156616211, 0.11910553741455078, 0.11873177337646484, 0.11934413146972657, 0.11961958312988281, 0.11883519744873047, 0.11877273559570313, 0.11876761627197266, 0.12034662628173828, 0.12008140563964843, 0.11964415740966797, 0.12007014465332032, 0.12065280151367187, 0.11880754852294922, 0.11892736053466797, 0.11855974578857421, 0.11855052947998047, 0.11903590393066406, 0.11845938873291016, 0.11849523162841796, 0.1184716796875, 0.11835289764404297, 0.11853619384765625, 0.11834368133544922, 0.11937382507324219, 0.11912806701660156, 0.11848806762695313, 0.11961753845214844, 0.1190113296508789, 0.11965132904052735, 0.11860281372070312, 0.11934611511230468, 0.11991145324707031]",tokens/s,8.254324575433342,,,main,False,False,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,databricks/dbrx-base,databricks/dbrx-base,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/databricks/dbrx-base/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aa53e-5dddb95c5609a1d50bf7f90c;1898565b-d7f9-4e0b-af1a-22c8efdbbbeb) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/databricks/dbrx-base/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like databricks/dbrx-base is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,huggyllama/llama-7b,huggyllama/llama-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,1979.4944,5480.382464,0.0,4833.93536,4503.282688,s,10,5.706418334960937,0.5706418334960938,0.0011422166985642673,0.5705734558105469,0.5721191223144532,0.5723321746826172,0.5725026165771484,"[0.5707071533203125, 0.5725452270507813, 0.570284912109375, 0.5698196411132812, 0.5708470458984375, 0.569266357421875, 0.5704397583007812, 0.571673828125, 0.57207177734375, 0.5687626342773437]",tokens/s,448.6176529182773,kWh,6.72348948356546e-06,3.6841831159411714e-06,3.142670724010705e-05,4.183437983961368e-05,tokens/kWh,6119368.829691346,MB,1979.4944,5480.382464,0.0,4833.93536,4688.699392,s,10,334.87464062500004,33.487464062499996,0.006284503345517177,33.486009765625,33.4966609375,33.498351953125,33.499704765625005,"[33.50004296875, 33.49628515625, 33.487125, 33.48791796875, 33.48489453125, 33.48313671875, 33.490703125, 33.48450390625, 33.48125, 33.47878125]",tokens/s,1.8813010111013093,kWh,0.0003953183352579305,0.00021666870724086642,0.0018133419907958965,0.002425329033294693,tokens/kWh,25975.85693946752,,s,629,339.4783874511717,0.5397112678079045,0.0678655066405644,0.531504150390625,0.5319763916015625,0.5321570190429687,1.1021465185546875,"[0.531127197265625, 0.5313873901367188, 0.5311314086914063, 0.5313760986328125, 0.5314212036132813, 0.5316044921875, 0.5316137084960938, 0.5314949340820313, 0.5322680053710938, 0.531794921875, 0.5314816284179688, 0.5319925537109375, 0.531863525390625, 0.5316638793945313, 0.5320120239257813, 0.5315921630859375, 0.5318819580078125, 0.5321605224609375, 0.53214208984375, 0.531884033203125, 0.5315399780273438, 0.5322833862304688, 0.532052978515625, 0.5314406127929687, 0.5317867431640625, 0.5322066040039063, 0.531673095703125, 0.53195263671875, 0.5317314453125, 0.5317376098632812, 0.5320519409179687, 0.5320724487304688, 0.531541015625, 0.5316557006835938, 0.53180517578125, 0.5318154296875, 0.5315655517578125, 0.5316137084960938, 0.5321820068359375, 0.531557373046875, 0.5312399291992187, 0.5314426879882812, 0.5312901000976562, 0.53134130859375, 0.531072998046875, 0.5317980346679687, 0.5320304565429688, 0.5319249877929687, 0.5319178466796874, 0.5318133544921875, 0.531989501953125, 0.5314345092773437, 0.531373046875, 0.531641357421875, 0.5318492431640625, 0.5319464721679688, 0.5321226196289063, 0.5319137573242188, 0.5322987670898438, 0.531788818359375, 0.53163623046875, 0.5317109985351562, 1.1043450927734375, 0.531673095703125, 0.5321574096679688, 0.5315389404296875, 0.5321922607421875, 0.531525634765625, 0.5317652587890624, 0.531631103515625, 0.5321697387695312, 0.5315164184570312, 0.5316024169921875, 0.5313341674804688, 0.5321021728515625, 0.5321441040039062, 0.5318215942382812, 0.531609619140625, 0.5315983276367188, 0.5313085327148438, 0.5319906005859375, 0.5317283325195312, 0.5320519409179687, 0.5316751098632813, 0.5315983276367188, 0.5317775268554688, 0.5322587890625, 0.5316044921875, 0.5319588012695312, 0.5312645263671875, 0.5314641723632813, 0.5313402709960937, 0.531599365234375, 0.5311528930664062, 0.5314417114257812, 0.5315245361328125, 0.5315297241210938, 0.531078125, 0.53165771484375, 0.5314437255859376, 0.5313720092773437, 0.53186767578125, 0.531873779296875, 0.5317017822265625, 0.53191064453125, 0.5315491943359375, 0.531904541015625, 0.53142529296875, 0.532937744140625, 0.5315389404296875, 0.5317601318359375, 0.5315614624023437, 0.5316034545898437, 0.5312798461914062, 0.5315686645507812, 0.53125634765625, 0.5316055297851563, 0.5312440185546875, 0.5317672729492188, 0.5314345092773437, 0.5318533325195313, 0.5322587890625, 0.5322587890625, 0.5315594482421875, 0.5315952758789062, 1.10245166015625, 0.5316915283203125, 0.5318369140625, 0.531167236328125, 0.5314037475585938, 0.5311907958984375, 0.531399658203125, 0.5313310546875, 0.5314713745117188, 0.5310955810546875, 0.5314068603515625, 0.5314180908203125, 0.5313095703125, 0.5320089721679687, 0.532147216796875, 0.5318717041015625, 0.5314703369140625, 0.5313587036132813, 0.5315277099609375, 0.5315932006835937, 0.5320427856445312, 0.5314682006835938, 0.5316864013671875, 0.5312532348632812, 0.531578857421875, 0.5313218383789062, 0.5315369262695312, 0.5313751220703125, 0.5314324340820312, 0.5314744262695312, 0.53136181640625, 0.5313638305664062, 0.5315297241210938, 0.5316331787109375, 0.5315000610351562, 0.5318041381835937, 0.5314805908203125, 0.5310812377929688, 0.5316290283203124, 0.531251220703125, 0.5314805908203125, 0.5312870483398437, 0.5313802490234375, 0.5314498291015625, 0.531378173828125, 0.5312542724609375, 0.531356689453125, 0.5313423461914063, 0.5315717163085938, 0.5312665405273438, 0.5319905395507812, 0.5315286865234375, 0.5317857055664063, 0.5316792602539062, 0.5319916381835937, 0.5315020141601563, 0.5320150756835937, 0.531610595703125, 0.5316116333007812, 0.531620849609375, 0.5315635375976563, 0.53176318359375, 0.5320653076171875, 1.1017093505859374, 0.5309368286132813, 0.5314478149414062, 0.531114990234375, 0.5318041381835937, 0.531794921875, 0.5318615112304688, 0.5316116333007812, 0.5316034545898437, 0.5310453491210938, 0.5313955688476563, 0.5313310546875, 0.5316249389648438, 0.5315194702148438, 0.5316045532226562, 0.5311835327148438, 0.5314447631835938, 0.53148876953125, 0.5315830688476563, 0.5314241943359375, 0.5317130126953125, 0.5312276611328125, 0.5314457397460938, 0.5311815795898438, 0.5318450927734375, 0.5317929077148438, 0.5315983276367188, 0.5316392822265625, 0.5314345092773437, 0.5313597412109375, 0.5314928588867187, 0.53110986328125, 0.5317550048828125, 0.531198974609375, 0.5312501831054688, 0.5313074951171874, 0.5314006958007812, 0.5310812377929688, 0.5314263305664062, 0.53121435546875, 0.5314088745117187, 0.5312266235351563, 0.5319813232421875, 0.5324287719726563, 0.53281494140625, 0.532264892578125, 0.5315143432617188, 0.531272705078125, 0.5317017822265625, 0.5313546142578125, 0.531945556640625, 0.5313545532226562, 0.53178369140625, 0.5314877319335938, 0.5314969482421875, 0.5313597412109375, 0.5315460815429688, 0.5314918212890625, 0.5317130126953125, 0.5316566772460938, 0.5317969970703125, 0.5315768432617187, 0.53257421875, 1.1031951904296875, 0.5310873413085937, 0.531431396484375, 0.5310433349609375, 0.5313966064453125, 0.5312911376953126, 0.5316239624023438, 0.5313822631835937, 0.5315317993164063, 0.53124609375, 0.5313659057617187, 0.5311549682617187, 0.5316239624023438, 0.531420166015625, 0.5314857177734374, 0.531357666015625, 0.5316966552734375, 0.5313771362304688, 0.5317969970703125, 0.5319751586914062, 0.531504150390625, 0.5317877807617187, 0.53207958984375, 0.531504150390625, 0.5317980346679687, 0.5316649169921875, 0.5317243041992188, 0.5313710327148438, 0.53178369140625, 0.5310208129882813, 0.5313812255859375, 0.5311692504882812, 0.5312440185546875, 0.530914306640625, 0.531140625, 0.5312010498046875, 0.5316300659179688, 0.5319229736328125, 0.5317314453125, 0.5311651611328125, 0.5311211547851562, 0.531178466796875, 0.5316658935546875, 0.5312122802734375, 0.5314109497070313, 0.5314529418945313, 0.5315552978515625, 0.531398681640625, 0.5312880859375, 0.5313443603515625, 0.53163623046875, 0.5312163696289063, 0.5315215454101563, 0.5311047973632812, 0.5318932495117188, 0.531894287109375, 0.5316249389648438, 0.5334948120117188, 0.5316904907226563, 0.5312880859375, 0.5315880737304688, 0.5312071533203125, 0.5317755126953125, 1.1012259521484375, 0.5312655639648437, 0.531800048828125, 0.5310771484375, 0.5316597900390625, 0.5314877319335938, 0.5313556518554687, 0.531009521484375, 0.5315706787109375, 0.5319342041015624, 0.5316925659179688, 0.5315604248046875, 0.5314334716796875, 0.5310894165039063, 0.531968994140625, 0.53129931640625, 0.5314898071289063, 0.5309706420898438, 0.5316198120117187, 0.5311743774414063, 0.53146826171875, 0.5311262817382812, 0.5313914794921875, 0.5310750732421875, 0.5313494873046875, 0.53096142578125, 0.5314263916015625, 0.5310238037109375, 0.5317867431640625, 0.5314058227539062, 0.531926025390625, 0.5319659423828125, 0.5314160766601562, 0.531178466796875, 0.5314283447265625, 0.5311897583007813, 0.5313648681640625, 0.5312819213867187, 0.53264794921875, 0.531478515625, 0.5314662475585937, 0.5313802490234375, 0.53146728515625, 0.531188720703125, 0.531399658203125, 0.5311815795898438, 0.5315440673828125, 0.5322147827148438, 0.531962890625, 0.5316351928710937, 0.5318799438476562, 0.5312614135742187, 0.5315430297851562, 0.5312973022460937, 0.531641357421875, 0.5314744262695312, 0.5315491943359375, 0.5312296752929687, 0.531863525390625, 0.5313689575195313, 0.5315000610351562, 0.5313095703125, 0.5315379028320313, 1.1023165283203125, 0.5312081909179688, 0.5320724487304688, 0.5311488037109375, 0.5318819580078125, 0.5312911376953126, 0.5314949340820313, 0.5315706787109375, 0.5317857055664063, 0.5313474731445312, 0.5318563842773437, 0.5311016845703125, 0.5315194702148438, 0.53150927734375, 0.5318553466796875, 0.5311559448242188, 0.5314652099609375, 0.5313792114257813, 0.5314171142578125, 0.5314180908203125, 0.5321830444335938, 0.5318389892578125, 0.5316331787109375, 0.5316177368164062, 0.5313392944335937, 0.5313668823242188, 0.5315061645507813, 0.5312501831054688, 0.5319915771484375, 0.5316126708984374, 0.5317601318359375, 0.5316557006835938, 0.53146826171875, 0.5313116455078125, 0.5315369262695312, 0.5313966064453125, 0.5317078857421875, 0.5314437255859376, 0.5320560913085938, 0.5326489868164063, 0.5319014282226563, 0.531631103515625, 0.5314703369140625, 0.5317478637695312, 0.5316761474609375, 0.5316065063476563, 0.531609619140625, 0.5315194702148438, 0.5318502197265625, 0.5314826049804687, 0.5319669799804687, 0.5315072021484375, 0.5315419921875, 0.5310453491210938, 0.5313034057617188, 0.5311867065429687, 0.5317181396484375, 0.531683349609375, 0.5318041381835937, 0.5316966552734375, 0.5320130615234375, 0.5314692993164063, 0.5316198120117187, 1.103932373046875, 0.5310914306640625, 0.531578857421875, 0.5308538818359375, 0.5313659057617187, 0.5314221801757812, 0.5314703369140625, 0.5309716186523438, 0.5315850830078125, 0.5310443115234375, 0.5311928100585938, 0.53098291015625, 0.5311600341796875, 0.5316239624023438, 0.531800048828125, 0.531646484375, 0.5318225708007812, 0.5314242553710937, 0.5319229736328125, 0.53176318359375, 0.5316587524414063, 0.53146826171875, 0.531483642578125, 0.5312266235351563, 0.5317498779296875, 0.5314273071289063, 0.5314631958007813, 0.5311232299804688, 0.5317816162109374, 0.5313494873046875, 0.5318584594726562, 0.5314795532226563, 0.5324308471679687, 0.5313074951171874, 0.5318523559570313, 0.5310309448242188, 0.5311488037109375, 0.5311488037109375, 0.5314488525390625, 0.530966552734375, 0.5313894653320312, 0.5310975952148438, 0.5314426879882812, 0.5313054809570312, 0.5314180908203125, 0.5313187866210938, 0.5313914794921875, 0.5315809326171875, 0.5312921752929688, 0.5312911376953126, 0.531684326171875, 0.5320560913085938, 0.5315133666992188, 0.5317980346679687, 0.5318328247070313, 0.5314559936523438, 0.5321195678710937, 0.5318225708007812, 0.5315963134765626, 0.5315829467773437, 0.5318184814453125, 0.531431396484375, 0.5319127197265625, 1.103847412109375, 0.531304443359375, 0.5315451049804687, 0.531357666015625, 0.5316792602539062, 0.5319639282226563, 0.5315829467773437, 0.5313710327148438, 0.5315205078125, 0.5312491455078125, 0.5314611206054688, 0.5316341552734375, 0.5313760986328125, 0.5311549682617187, 0.531583984375, 0.5310637817382813, 0.5312645263671875, 0.53108837890625, 0.5316741333007813, 0.531162109375, 0.5312429809570313, 0.5314140014648437, 0.531314697265625, 0.5313054809570312, 0.531968017578125, 0.531599365234375, 0.5316085815429688, 0.5312634887695312, 0.5314406127929687, 0.53096240234375, 0.5313423461914063, 0.5313720092773437, 0.5316188354492187, 0.5309050903320313, 0.5311948852539062, 0.5310873413085937, 0.5315645141601563, 0.5312706298828125, 0.5313065185546875, 0.5313433837890625, 0.53119384765625, 0.5309767456054687, 0.5315348510742187, 0.532041748046875, 0.5317161254882813, 0.5314119873046875, 0.5313494873046875, 0.5311979370117188, 0.5316075439453125, 0.5313341674804688, 0.5313423461914063, 0.531177490234375, 0.5312553100585937, 0.531019775390625, 0.5315665893554687, 0.5312440185546875, 0.5314283447265625, 0.5314006958007812, 0.5314877319335938, 0.5311262817382812, 0.5316321411132813, 0.5312686157226563, 0.5320028076171875, 1.103824951171875, 0.5310279541015624, 0.53146826171875, 0.5310341186523437, 0.5313740844726562, 0.5310853271484375, 0.5314447631835938, 0.5311447143554687, 0.5314212036132813, 0.531420166015625, 0.5311948852539062, 0.5308630981445313, 0.5313853149414063, 0.5309020385742188, 0.53150830078125, 0.5310576171875, 0.5312839965820313, 0.5317191772460937, 0.5318482055664062, 0.5312532348632812, 0.5314641723632813, 0.5309531860351563, 0.5312553100585937, 0.5309808349609375, 0.5311590576171875, 0.5310494995117188, 0.531863525390625, 0.5317703857421875, 0.5314180908203125, 0.53118359375, 0.5314396362304687, 0.5311129150390625, 0.5312420043945313, 0.5310259399414062, 0.5311918334960938, 0.5312973022460937, 0.5321564331054688, 0.5315194702148438, 0.5316976928710937, 0.5313034057617188, 0.5316218872070313, 0.531430419921875, 0.5316812744140625, 0.5315625, 0.5316239624023438, 0.5310105590820312, 0.5315389404296875, 0.5312665405273438, 0.5317191772460937, 0.5312142944335938, 0.5314641723632813, 0.53167822265625, 0.5313607788085938, 0.5311918334960938, 0.5314918212890625, 0.53161474609375, 0.5318031616210938, 0.5315338134765625, 0.5315020751953125, 0.5313065185546875, 0.531968994140625, 0.5318994140625, 0.5317734375]",tokens/s,1.8528425468336207,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neox-20b,EleutherAI/gpt-neox-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,,,MB,2108.223488,15043.395584,0.0,14396.94848,13898.251776,s,10,16.774060791015625,1.6774060791015626,0.0018297475804955295,1.6770918579101561,1.6789085327148436,1.6804502990722656,1.6816837121582031,"[1.6819920654296876, 1.67540869140625, 1.6774444580078125, 1.676206298828125, 1.6767392578125, 1.6753848876953126, 1.6765943603515625, 1.6776787109375, 1.67856591796875, 1.678046142578125]",tokens/s,152.61659248136053,kWh,1.9787035716904537e-05,1.0841899168262898e-05,9.527879844519888e-05,0.0001259077333303663,tokens/kWh,2033234.919163287,MB,2108.223488,15043.395584,0.0,14396.94848,14315.959808,s,10,982.2773984375,98.22773984375,0.006322715714689567,98.22801562500001,98.23432265625,98.236829296875,98.238834609375,"[98.233765625, 98.224484375, 98.2393359375, 98.2251640625, 98.22903125, 98.2138984375, 98.2302578125, 98.2295859375, 98.224875, 98.227]",tokens/s,0.6413666862356148,kWh,0.001159520428114467,0.0006355184711477705,0.005722492522434801,0.0075175314216970384,tokens/kWh,8380.410598374077,,s,629,995.810533813476,1.5831646006573552,0.19919555203142542,1.559141357421875,1.55962734375,1.5598268310546874,3.23526474609375,"[1.5585750732421875, 1.55860986328125, 1.5594239501953124, 1.5595242919921875, 1.558529052734375, 1.5586417236328125, 1.559109619140625, 1.5591219482421874, 1.5598243408203125, 1.5595867919921875, 1.55911572265625, 1.5588905029296876, 1.5587757568359375, 1.5587236328125, 1.559189453125, 1.558750244140625, 1.5590482177734375, 1.5591536865234374, 1.5591956787109376, 1.55924169921875, 1.558993896484375, 1.559488525390625, 1.559013427734375, 1.5594393310546875, 1.5592908935546874, 1.5589068603515626, 1.55905126953125, 1.559194580078125, 1.5587266845703125, 1.5591065673828124, 1.5589693603515624, 1.558802490234375, 1.558849609375, 1.559245849609375, 1.5592437744140626, 1.559234619140625, 1.55915771484375, 1.5592222900390624, 1.5598233642578125, 1.55966162109375, 1.559911376953125, 1.5597353515625, 1.559793701171875, 1.5596707763671875, 1.55985302734375, 1.55943115234375, 1.5597127685546874, 1.5593175048828125, 1.55945263671875, 1.559732177734375, 1.559920654296875, 1.5596063232421875, 1.5595447998046874, 1.5596646728515624, 1.559162841796875, 1.5595458984375, 1.55953662109375, 1.5590595703125, 1.559057373046875, 1.5596236572265625, 1.5592509765625, 1.559478271484375, 3.234529296875, 1.55871337890625, 1.5585946044921875, 1.5587962646484375, 1.5588270263671875, 1.558576171875, 1.558561767578125, 1.55837744140625, 1.5585587158203125, 1.5589652099609375, 1.5587061767578125, 1.55881787109375, 1.558748046875, 1.5588065185546875, 1.5585740966796875, 1.5593819580078125, 1.5591004638671875, 1.558873046875, 1.5587451171875, 1.5594066162109375, 1.559108642578125, 1.55873583984375, 1.5592396240234374, 1.559013427734375, 1.558961181640625, 1.5588013916015624, 1.5588310546875, 1.5591044921875, 1.5594649658203126, 1.558951904296875, 1.5591546630859374, 1.55907177734375, 1.5593482666015626, 1.5590062255859376, 1.559110595703125, 1.5595079345703124, 1.55951513671875, 1.5589560546875, 1.5587255859375, 1.5589560546875, 1.559404541015625, 1.5599503173828124, 1.5588074951171875, 1.5591341552734375, 1.5587113037109375, 1.55890380859375, 1.561585693359375, 1.5592509765625, 1.5593123779296876, 1.5594271240234374, 1.55922021484375, 1.5592386474609374, 1.559277587890625, 1.559350341796875, 1.5596614990234374, 1.5596533203125, 1.559194580078125, 1.5592427978515624, 1.5593604736328126, 1.5594066162109375, 1.559331787109375, 1.5592806396484375, 1.559973876953125, 3.23797607421875, 1.5602718505859374, 1.559963623046875, 1.5592427978515624, 1.5589161376953125, 1.5587706298828126, 1.5589017333984374, 1.559024658203125, 1.559089111328125, 1.5593585205078124, 1.5589334716796874, 1.5592069091796874, 1.5589990234375, 1.5592745361328124, 1.5593963623046876, 1.559119873046875, 1.5596800537109374, 1.5589478759765625, 1.5588074951171875, 1.5590042724609374, 1.558916015625, 1.55900927734375, 1.5591424560546876, 1.559677978515625, 1.5591373291015624, 1.559119873046875, 1.5591474609375, 1.5588157958984374, 1.5592601318359376, 1.5591690673828125, 1.559141357421875, 1.5591485595703125, 1.55908203125, 1.559183349609375, 1.55915576171875, 1.559400390625, 1.5592960205078126, 1.5591761474609376, 1.5593154296875, 1.5590482177734375, 1.558973388671875, 1.559626708984375, 1.559330810546875, 1.5593973388671876, 1.5593338623046875, 1.559593994140625, 1.5598807373046875, 1.5612364501953124, 1.55947412109375, 1.5594168701171875, 1.5596483154296874, 1.5595643310546874, 1.559582763671875, 1.55965234375, 1.5598623046875, 1.559635986328125, 1.5595242919921875, 1.5597813720703124, 1.5598643798828125, 1.5591075439453126, 1.5595611572265624, 1.55919873046875, 1.5593585205078124, 3.23542333984375, 1.55846044921875, 1.558307861328125, 1.5589805908203125, 1.558640625, 1.5585218505859375, 1.55879833984375, 1.558877197265625, 1.5586907958984375, 1.5587757568359375, 1.5589539794921874, 1.559329833984375, 1.559342041015625, 1.5588546142578126, 1.5590174560546874, 1.559341064453125, 1.558877197265625, 1.5588382568359376, 1.55911474609375, 1.5592764892578126, 1.5589847412109374, 1.5587337646484376, 1.55911376953125, 1.5587901611328125, 1.5588280029296875, 1.5588690185546874, 1.5593216552734375, 1.559299072265625, 1.559120849609375, 1.558898681640625, 1.5588414306640626, 1.55898876953125, 1.5588966064453125, 1.5587901611328125, 1.558739990234375, 1.5589232177734376, 1.5593912353515624, 1.55947314453125, 1.559403564453125, 1.559525390625, 1.559373779296875, 1.5593902587890625, 1.5591658935546875, 1.559456787109375, 1.5621683349609374, 1.558978515625, 1.5591126708984375, 1.55909326171875, 1.5591383056640624, 1.5590767822265625, 1.55901025390625, 1.5589765625, 1.559456787109375, 1.5590185546875, 1.55949462890625, 1.55913623046875, 1.5593287353515626, 1.559487548828125, 1.559192626953125, 1.55945263671875, 1.5594342041015625, 1.5596298828125, 1.5594813232421876, 3.236599853515625, 1.5593328857421875, 1.5589642333984375, 1.55854541015625, 1.558728759765625, 1.558466552734375, 1.55877783203125, 1.5593133544921876, 1.5591546630859374, 1.5588116455078125, 1.5587174072265626, 1.5586641845703124, 1.558667236328125, 1.558561767578125, 1.5583734130859375, 1.558703125, 1.5588648681640624, 1.55964111328125, 1.5586856689453126, 1.558662109375, 1.5585587158203125, 1.558540283203125, 1.5592540283203125, 1.558740966796875, 1.559257080078125, 1.5590430908203126, 1.5589775390625, 1.558866943359375, 1.5591044921875, 1.5596021728515626, 1.559546875, 1.55960009765625, 1.55951513671875, 1.5589847412109374, 1.5592540283203125, 1.559413818359375, 1.5594434814453124, 1.5591956787109376, 1.559342041015625, 1.5589847412109374, 1.5589805908203125, 1.5593165283203125, 1.559119873046875, 1.5594117431640624, 1.5589334716796874, 1.5593184814453125, 1.559287841796875, 1.5593902587890625, 1.5601192626953124, 1.5590482177734375, 1.559271484375, 1.559646240234375, 1.5597353515625, 1.5595601806640624, 1.5594761962890624, 1.559477294921875, 1.5593011474609375, 1.5621007080078124, 1.559258056640625, 1.5594239501953124, 1.559357421875, 1.5594639892578126, 1.559361572265625, 3.23485693359375, 1.5584061279296875, 1.5584429931640624, 1.558455322265625, 1.558666259765625, 1.5583436279296874, 1.5583365478515625, 1.55860888671875, 1.55905224609375, 1.558590576171875, 1.558578125, 1.5587183837890626, 1.55867236328125, 1.5587255859375, 1.5585269775390624, 1.5587215576171876, 1.558656982421875, 1.5586744384765625, 1.5588331298828124, 1.559310302734375, 1.558556640625, 1.5583118896484376, 1.5588935546875, 1.5596380615234375, 1.559288818359375, 1.558983642578125, 1.559357421875, 1.559225341796875, 1.558750244140625, 1.558867919921875, 1.5587225341796875, 1.5589283447265625, 1.5586417236328125, 1.5587911376953125, 1.5586907958984375, 1.559267333984375, 1.5594957275390624, 1.558765625, 1.5612979736328125, 1.5589273681640625, 1.5587225341796875, 1.558982666015625, 1.5588209228515626, 1.5590604248046875, 1.558908935546875, 1.5590697021484374, 1.5586539306640625, 1.5587706298828126, 1.5595069580078125, 1.559635986328125, 1.5590185546875, 1.55902978515625, 1.5588546142578126, 1.5590277099609375, 1.5595592041015625, 1.5593226318359374, 1.5589283447265625, 1.5589744873046876, 1.5589171142578124, 1.5588587646484375, 1.5594556884765625, 1.5590963134765625, 1.5595478515625, 3.235958740234375, 1.5587061767578125, 1.5582955322265626, 1.558918212890625, 1.559405517578125, 1.5590072021484376, 1.559066650390625, 1.559258056640625, 1.558982666015625, 1.5591373291015624, 1.5590123291015625, 1.559109619140625, 1.5587010498046876, 1.5590697021484374, 1.558677490234375, 1.558846435546875, 1.559234619140625, 1.55937890625, 1.558992919921875, 1.5595223388671875, 1.559162841796875, 1.5587706298828126, 1.559435302734375, 1.560753173828125, 1.55900830078125, 1.5587052001953126, 1.5587420654296875, 1.55875634765625, 1.5585545654296875, 1.55881884765625, 1.5590809326171875, 1.559193603515625, 1.559277587890625, 1.55932568359375, 1.55913623046875, 1.5592764892578126, 1.559235595703125, 1.5595654296875, 1.559406494140625, 1.5591424560546876, 1.558719482421875, 1.55981103515625, 1.5590491943359375, 1.5598408203125, 1.55958984375, 1.55905029296875, 1.5590296630859375, 1.5592437744140626, 1.559709716796875, 1.559160888671875, 1.5592642822265625, 1.55951513671875, 1.559525390625, 1.55964111328125, 1.559710693359375, 1.5596390380859375, 1.5594691162109375, 1.55936767578125, 1.5593133544921876, 1.5591116943359375, 1.5594761962890624, 1.5594700927734375, 1.5595284423828124, 3.23728076171875, 1.558865966796875, 1.5586630859375, 1.5593482666015626, 1.5598438720703125, 1.559294921875, 1.559099365234375, 1.5591485595703125, 1.558935546875, 1.55944970703125, 1.558636474609375, 1.5588331298828124, 1.5585986328125, 1.558814697265625, 1.55905224609375, 1.5590338134765624, 1.56031591796875, 1.5590655517578125, 1.5592652587890625, 1.5588966064453125, 1.5590400390625, 1.5587420654296875, 1.5598909912109375, 1.559056396484375, 1.5592652587890625, 1.5594056396484375, 1.5589385986328126, 1.5587860107421876, 1.55886083984375, 1.5590543212890624, 1.559300048828125, 1.5590921630859376, 1.559215087890625, 1.5589744873046876, 1.5589600830078125, 1.5591177978515625, 1.5594957275390624, 1.5592960205078126, 1.558877197265625, 1.5592017822265625, 1.5588270263671875, 1.5589222412109376, 1.5591588134765626, 1.559298095703125, 1.5588055419921876, 1.5590748291015626, 1.559099365234375, 1.5595028076171875, 1.5594915771484374, 1.5594495849609376, 1.559300048828125, 1.55938916015625, 1.5596697998046876, 1.5595755615234375, 1.5595201416015625, 1.5593564453125, 1.5596134033203124, 1.55945263671875, 1.5595294189453126, 1.5592764892578126, 1.559214111328125, 1.55919873046875, 1.5595919189453125, 3.238277099609375, 1.5593380126953125, 1.558845458984375, 1.559089111328125, 1.5596502685546876, 1.559151611328125, 1.55905126953125, 1.5589949951171875, 1.558698974609375, 1.5589119873046875, 1.559118896484375, 1.5591055908203124, 1.5588433837890625, 1.55887109375, 1.559098388671875, 1.559488525390625, 1.5590921630859376, 1.5589212646484376, 1.5586201171875, 1.5587542724609376, 1.5588372802734376, 1.5583734130859375, 1.559034912109375, 1.5590389404296876, 1.5593184814453125, 1.559047119140625, 1.559078857421875, 1.5590113525390625, 1.5589908447265626, 1.5593114013671876, 1.559214111328125, 1.559314453125, 1.5593348388671875, 1.5594014892578125, 1.5593011474609375, 1.55919873046875, 1.5591455078125, 1.559130126953125, 1.5589488525390625, 1.55922021484375, 1.5590164794921875, 1.55928369140625, 1.5593359375, 1.55940869140625, 1.5589334716796874, 1.5590072021484376, 1.5590543212890624, 1.5592960205078126, 1.5591453857421875, 1.5592960205078126, 1.55916796875, 1.55928271484375, 1.559329833984375, 1.5590318603515625, 1.5594649658203126, 1.5592314453125, 1.55926123046875, 1.5591290283203125, 1.5590205078125, 1.5590174560546874, 1.559034912109375, 1.5591658935546875, 1.559373779296875, 3.237329833984375, 1.558613037109375, 1.5583804931640626, 1.5587542724609376, 1.5585423583984375, 1.559004150390625, 1.5588382568359376, 1.55866015625, 1.5590809326171875, 1.558877197265625, 1.558772705078125, 1.558992919921875, 1.558794189453125, 1.559083984375, 1.55869287109375, 1.5590604248046875, 1.55894580078125, 1.5587255859375, 1.5587962646484375, 1.558877197265625, 1.5590615234375, 1.558513671875, 1.5592530517578125, 1.5587318115234374, 1.5593697509765625, 1.559119873046875, 1.5611954345703125, 1.559103515625, 1.5589058837890626, 1.5593779296875, 1.559484375, 1.5593184814453125, 1.5594208984375, 1.5591065673828124, 1.559189453125, 1.5593656005859375, 1.5593656005859375, 1.55962060546875, 1.5591669921875, 1.5591322021484375, 1.559208984375, 1.5593665771484375, 1.559109619140625, 1.5592222900390624, 1.5591434326171876, 1.559194580078125, 1.5592008056640625, 1.55911376953125, 1.55950390625, 1.5590687255859375, 1.559582763671875, 1.5593564453125, 1.5598284912109375, 1.5596851806640626, 1.55951513671875, 1.55947216796875, 1.5594691162109375, 1.5591895751953124, 1.5594127197265626, 1.55911572265625, 1.559299072265625, 1.5592508544921875, 1.5591474609375]",tokens/s,0.6316462606508407,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-4B,Qwen/Qwen1.5-4B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,3193.102336,5128.060928,0.0,4481.613824,4276.256768,s,10,3.2413734130859373,0.32413734130859373,0.0016825840816240875,0.3241157684326172,0.32623468627929686,0.3264099456787109,0.32655015319824215,"[0.326585205078125, 0.3253639221191406, 0.3222522888183594, 0.32284228515625, 0.32339013671875, 0.32129388427734373, 0.32326702880859376, 0.3248414001464844, 0.32534152221679685, 0.3261957397460937]",tokens/s,789.7886709580806,kWh,3.8053463575326738e-06,2.0845240635480878e-06,1.689038677618604e-05,2.27802571972668e-05,tokens/kWh,11237801.126789523,MB,3193.102336,5128.060928,0.0,4481.613824,4465.661952,s,10,189.078173828125,18.9078173828125,0.012273201694566093,18.9097021484375,18.920707812499998,18.9233919921875,18.9255393359375,"[18.91315625, 18.920111328125, 18.9024375, 18.901080078125, 18.88138671875, 18.897064453125, 18.91123828125, 18.908166015625, 18.926076171875, 18.91745703125]",tokens/s,3.331955176236681,kWh,0.00022324078258437416,0.00012235523933478362,0.0009709997090908196,0.0013165957310099774,tokens/kWh,47850.67922988926,,s,629,191.68752508544924,0.3047496424251975,0.038502073286161766,0.300015625,0.3008729064941406,0.3014152465820313,0.6232092919921874,"[0.3012270202636719, 0.30114816284179685, 0.3007774658203125, 0.30110003662109375, 0.29941351318359377, 0.29971661376953124, 0.2996531066894531, 0.2997452697753906, 0.30033203125, 0.3001241455078125, 0.30014566040039065, 0.3000565795898438, 0.3007344665527344, 0.30101809692382814, 0.2998548583984375, 0.2999808044433594, 0.3005767822265625, 0.29991937255859374, 0.3002378234863281, 0.3010672607421875, 0.30078155517578126, 0.30025625610351564, 0.29994802856445313, 0.3000330505371094, 0.3002951354980469, 0.30040884399414064, 0.30003506469726565, 0.30066790771484375, 0.30030438232421874, 0.30015179443359374, 0.30028594970703126, 0.3002818603515625, 0.30056243896484375, 0.3007068786621094, 0.3003309326171875, 0.29966232299804685, 0.2997370910644531, 0.2999521179199219, 0.29981594848632814, 0.2995947570800781, 0.29999102783203124, 0.29974835205078126, 0.299620361328125, 0.2994493408203125, 0.2994759826660156, 0.29976165771484375, 0.30030642700195315, 0.3000166320800781, 0.2999060363769531, 0.30026956176757813, 0.2995128173828125, 0.29971661376953124, 0.2999552001953125, 0.3005082092285156, 0.30011181640625, 0.2995394592285156, 0.3000186767578125, 0.30043032836914063, 0.29998388671875, 0.29997259521484376, 0.299610107421875, 0.29887283325195313, 0.6254059448242187, 0.3006320495605469, 0.3013908386230469, 0.30085733032226564, 0.3013273620605469, 0.29998489379882814, 0.2996244506835937, 0.3003135986328125, 0.299826171875, 0.2998814697265625, 0.300400634765625, 0.29994085693359374, 0.3004631042480469, 0.3001395263671875, 0.2999992370605469, 0.299926513671875, 0.300759033203125, 0.300590087890625, 0.29988455200195313, 0.29991729736328127, 0.30021221923828123, 0.30040472412109376, 0.3002593383789062, 0.30088909912109374, 0.3011358642578125, 0.2998763427734375, 0.30023577880859376, 0.30020709228515624, 0.3008174133300781, 0.3003627624511719, 0.300759033203125, 0.3007027587890625, 0.30042620849609375, 0.30063821411132813, 0.2999624328613281, 0.29996844482421875, 0.30086349487304687, 0.30146969604492185, 0.29992141723632815, 0.3000361022949219, 0.3000586242675781, 0.29992755126953125, 0.2999654541015625, 0.2998486938476562, 0.30020095825195314, 0.3000299377441406, 0.30309478759765623, 0.3002388610839844, 0.2999859313964844, 0.3003576354980469, 0.30067608642578125, 0.3001446533203125, 0.30003302001953125, 0.3000166320800781, 0.2999244689941406, 0.3001692199707031, 0.29990194702148437, 0.3005317077636719, 0.30013031005859375, 0.2998405151367188, 0.29998797607421873, 0.2999336853027344, 0.30006988525390627, 0.622824462890625, 0.30008013916015625, 0.29997259521484376, 0.2997729187011719, 0.29949029541015626, 0.300337158203125, 0.30040985107421875, 0.30018765258789065, 0.2999029846191406, 0.29995724487304687, 0.30025112915039065, 0.30034637451171875, 0.29992959594726565, 0.30029824829101565, 0.3003873291015625, 0.2998978576660156, 0.30032589721679687, 0.3002255249023438, 0.3004590454101562, 0.3001773681640625, 0.2999449462890625, 0.2996777038574219, 0.2997125244140625, 0.3017687072753906, 0.2998691711425781, 0.29983026123046874, 0.2998783874511719, 0.30037503051757813, 0.3002183532714844, 0.29920254516601563, 0.29975244140625, 0.2997073974609375, 0.29997671508789064, 0.2997862548828125, 0.30004327392578123, 0.2995845031738281, 0.29983743286132813, 0.29989376831054687, 0.3005511779785156, 0.3004989318847656, 0.3004620666503906, 0.2998620300292969, 0.2996592712402344, 0.2998609924316406, 0.29997567749023435, 0.2994964599609375, 0.3000780944824219, 0.30188134765625, 0.29991015625, 0.30026956176757813, 0.30005453491210937, 0.29941964721679687, 0.3000770568847656, 0.2994923400878906, 0.29899981689453126, 0.299219970703125, 0.29973809814453123, 0.29967666625976563, 0.30005966186523436, 0.3000115661621094, 0.3005040283203125, 0.2999951477050781, 0.30062490844726564, 0.6233589477539062, 0.29990194702148437, 0.3002234802246094, 0.2998190002441406, 0.30008734130859377, 0.30004833984375, 0.2998200378417969, 0.3000770568847656, 0.300015625, 0.3000606689453125, 0.29941146850585937, 0.29921484375, 0.299578369140625, 0.2996756591796875, 0.2996449279785156, 0.30001971435546876, 0.3000995788574219, 0.29999002075195313, 0.3003607177734375, 0.30082763671875, 0.3000340576171875, 0.300611572265625, 0.30002584838867186, 0.299694091796875, 0.29959066772460935, 0.30004019165039064, 0.2999818115234375, 0.30008218383789065, 0.29999002075195313, 0.3006935119628906, 0.2997913513183594, 0.29910015869140627, 0.2994124755859375, 0.29980160522460936, 0.2999715881347656, 0.3000504455566406, 0.3000074157714844, 0.2999378051757812, 0.2999920654296875, 0.3000719299316406, 0.2999992370605469, 0.30074981689453123, 0.30103448486328127, 0.2998681640625, 0.3038136291503906, 0.3005091857910156, 0.3002746887207031, 0.2998763427734375, 0.299767822265625, 0.29981594848632814, 0.2999541625976562, 0.29982720947265623, 0.299641845703125, 0.29993267822265623, 0.29993472290039064, 0.2999715881347656, 0.2998343811035156, 0.29893017578125, 0.29962136840820314, 0.30002072143554687, 0.2994237365722656, 0.299831298828125, 0.30021630859375, 0.623921142578125, 0.3000299377441406, 0.30005966186523436, 0.299926513671875, 0.299947021484375, 0.300115966796875, 0.29993572998046875, 0.30002072143554687, 0.30004531860351563, 0.29974118041992187, 0.30062188720703126, 0.29995513916015626, 0.29942477416992186, 0.2993663940429688, 0.29983026123046874, 0.2999603271484375, 0.2999285888671875, 0.3015209045410156, 0.29998797607421873, 0.30004327392578123, 0.30011700439453126, 0.2999449462890625, 0.30022860717773436, 0.3003965759277344, 0.30081021118164064, 0.30038836669921876, 0.30042520141601564, 0.30013336181640626, 0.3006033935546875, 0.29989376831054687, 0.29989273071289063, 0.30014874267578123, 0.3004436340332031, 0.299747314453125, 0.2998476867675781, 0.3000094604492187, 0.29924453735351564, 0.29881036376953124, 0.29879910278320315, 0.2987581481933594, 0.29875506591796874, 0.29884518432617185, 0.298787841796875, 0.2988482666015625, 0.29977191162109373, 0.3000657958984375, 0.2989967346191406, 0.2990592041015625, 0.2991626281738281, 0.2990673828125, 0.29893017578125, 0.2991349792480469, 0.299323486328125, 0.29853277587890625, 0.29891278076171873, 0.29906942749023435, 0.2987386779785156, 0.3028009033203125, 0.29910748291015626, 0.2988552551269531, 0.2989045715332031, 0.299146240234375, 0.29900595092773435, 0.621075439453125, 0.2991349792480469, 0.2998179931640625, 0.29896807861328123, 0.2991811828613281, 0.2986914672851563, 0.29862911987304686, 0.29889239501953124, 0.2989116516113281, 0.2987960205078125, 0.29937152099609377, 0.29994189453125, 0.29974630737304686, 0.30000537109375, 0.300537841796875, 0.29991015625, 0.30004837036132814, 0.300326904296875, 0.3002030029296875, 0.3000575866699219, 0.3000770568847656, 0.3004405822753906, 0.2999531555175781, 0.30005148315429686, 0.3000484008789063, 0.3000626525878906, 0.30036172485351564, 0.3001272277832031, 0.30000640869140627, 0.2996879272460938, 0.29982925415039063, 0.29986611938476565, 0.3000391540527344, 0.2998128662109375, 0.29970330810546875, 0.30041189575195315, 0.3001200561523438, 0.29997772216796875, 0.3035576171875, 0.30046923828125, 0.29996749877929685, 0.3007201232910156, 0.3004538879394531, 0.29997671508789064, 0.30013543701171874, 0.30009344482421874, 0.3003351135253906, 0.3000340576171875, 0.300626953125, 0.29982925415039063, 0.29992959594726565, 0.299978759765625, 0.3000924072265625, 0.30002484130859375, 0.29995724487304687, 0.2998896789550781, 0.299720703125, 0.2996705322265625, 0.2998804626464844, 0.2997698669433594, 0.3005307006835938, 0.29985382080078127, 0.30008731079101564, 0.6239016723632812, 0.3004989318847656, 0.3000094604492187, 0.3003781127929688, 0.3024650268554688, 0.30008941650390625, 0.30021728515625, 0.29997671508789064, 0.3000985717773437, 0.2997452697753906, 0.3001559143066406, 0.299789306640625, 0.299578369140625, 0.3000637512207031, 0.3000022888183594, 0.30041189575195315, 0.300179443359375, 0.30062490844726564, 0.3001978759765625, 0.29980978393554686, 0.30084915161132814, 0.3001692199707031, 0.30007400512695315, 0.3024322204589844, 0.3004334106445313, 0.29990194702148437, 0.30002584838867186, 0.2999787902832031, 0.3001108093261719, 0.3005962219238281, 0.3003105163574219, 0.30027059936523437, 0.299931640625, 0.2999029846191406, 0.3000309753417969, 0.30007601928710936, 0.2999808044433594, 0.299936767578125, 0.3001610107421875, 0.29992755126953125, 0.29989376831054687, 0.29997567749023435, 0.3000391540527344, 0.3000443115234375, 0.3002992248535156, 0.2999715881347656, 0.2998241577148438, 0.2998548278808594, 0.3000862731933594, 0.300653564453125, 0.3015086059570313, 0.29992242431640626, 0.299863037109375, 0.29971966552734375, 0.29996954345703125, 0.2999613342285156, 0.3001968688964844, 0.2997841796875, 0.2999797668457031, 0.29949029541015626, 0.2995916748046875, 0.29981695556640625, 0.30072625732421876, 0.6255585327148437, 0.30017230224609376, 0.299831298828125, 0.3000555419921875, 0.30022042846679686, 0.30029925537109375, 0.30140109252929687, 0.30094949340820315, 0.2998886413574219, 0.29979647827148437, 0.30036376953125, 0.29990194702148437, 0.30001458740234377, 0.300010498046875, 0.29990707397460936, 0.29992959594726565, 0.3013775329589844, 0.2998958129882813, 0.3009157104492188, 0.3004344177246094, 0.2997698974609375, 0.299400146484375, 0.3000780944824219, 0.300410888671875, 0.3009054870605469, 0.3005409240722656, 0.301384765625, 0.30065350341796876, 0.30062591552734375, 0.30096075439453124, 0.3004375, 0.30160894775390623, 0.3008819274902344, 0.3004241943359375, 0.29964389038085937, 0.2996091003417969, 0.2996643981933594, 0.29964901733398436, 0.2998384704589844, 0.29929779052734373, 0.29952517700195314, 0.30001556396484375, 0.29967666625976563, 0.29992242431640626, 0.2999613342285156, 0.299926513671875, 0.2998753356933594, 0.2995947570800781, 0.2999183349609375, 0.2996172790527344, 0.29993267822265623, 0.30100991821289064, 0.30017843627929686, 0.30001254272460937, 0.29992141723632815, 0.29997055053710936, 0.300147705078125, 0.3002347412109375, 0.29967462158203123, 0.299504638671875, 0.299514892578125, 0.29878680419921877, 0.2995189819335938, 0.6255062866210938, 0.30063821411132813, 0.30005966186523436, 0.30045697021484374, 0.3000965270996094, 0.2999285888671875, 0.300015625, 0.300263427734375, 0.299894775390625, 0.29981695556640625, 0.30063821411132813, 0.300000244140625, 0.29996954345703125, 0.29998797607421873, 0.30009344482421874, 0.2999808044433594, 0.3002378234863281, 0.30191104125976564, 0.3011993713378906, 0.30066586303710935, 0.3013355407714844, 0.30207794189453124, 0.302529541015625, 0.30110720825195314, 0.3007958984375, 0.3010785217285156, 0.30087783813476565, 0.2998056945800781, 0.30067608642578125, 0.3011614685058594, 0.3007979431152344, 0.30112152099609374, 0.3007201232910156, 0.3007068176269531, 0.30179122924804686, 0.3001978759765625, 0.3004272766113281, 0.2999080810546875, 0.3004764099121094, 0.30043954467773437, 0.3001835632324219, 0.3006484375, 0.3004712829589844, 0.3014246826171875, 0.3003248291015625, 0.30087884521484376, 0.3002808227539063, 0.299840576171875, 0.3000575256347656, 0.30002383422851564, 0.2996745910644531, 0.29950567626953123, 0.2997145690917969, 0.29997567749023435, 0.30001458740234377, 0.30037503051757813, 0.30008114624023435, 0.2989066162109375, 0.30013543701171874, 0.2999111633300781, 0.29975347900390625, 0.2997350463867188, 0.30005453491210937, 0.6268078002929688, 0.30009548950195314, 0.29986407470703125, 0.30065869140625, 0.3011307373046875, 0.30074368286132813, 0.3008604125976562, 0.30035250854492185, 0.30087167358398437, 0.29988760375976564, 0.30013644409179685, 0.3006300048828125, 0.3005399169921875, 0.30181991577148437, 0.30090756225585935, 0.29990191650390624, 0.30007601928710936, 0.30002688598632815, 0.2999869384765625, 0.3001077880859375, 0.30047845458984374, 0.29997567749023435, 0.3000227966308594, 0.3018362731933594, 0.30386279296875, 0.30291455078125, 0.303236083984375, 0.3023329162597656, 0.30000436401367186, 0.2989014892578125, 0.2988851318359375, 0.29931622314453127, 0.2989588623046875, 0.29878271484375, 0.2992056274414063, 0.30051431274414064, 0.3006761169433594, 0.29950460815429686, 0.29902847290039064, 0.2990940246582031, 0.29902545166015626, 0.29914212036132815, 0.29988455200195313, 0.299652099609375, 0.30047232055664064, 0.2996326293945312, 0.30089727783203124, 0.29991729736328127, 0.3000965270996094, 0.30000234985351565, 0.29993365478515627, 0.3000115051269531, 0.2999603881835938, 0.29999508666992186, 0.29998284912109374, 0.2999869384765625, 0.30106521606445313, 0.3007928466796875, 0.300732421875, 0.29975653076171876, 0.29992755126953125, 0.30012313842773436, 0.29981491088867185]",tokens/s,3.281382029005844,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-1.8B,Qwen/Qwen1.5-1.8B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,1866.50624,3328.704512,0.0,2682.257408,2578.238464,s,10,1.4175592651367188,0.14175592651367186,0.0022402045159939975,0.14135794067382812,0.14292337188720702,0.14542544631958007,0.1474271058654785,"[0.14792752075195312, 0.14161701965332033, 0.1391898193359375, 0.1415839385986328, 0.14236735534667969, 0.14101449584960937, 0.13986236572265626, 0.1410111083984375, 0.14185369873046874, 0.14113194274902344]",tokens/s,1805.9209677932563,kWh,1.6547061226986076e-06,9.066846465884736e-07,6.786722250360737e-06,9.348113019647817e-06,tokens/kWh,27385205.919305906,MB,1866.50624,3328.704512,0.0,2682.257408,2667.0976,s,10,83.8020732421875,8.38020732421875,0.04772557025936409,8.37658544921875,8.433103320312501,8.4430912109375,8.4510815234375,"[8.4288447265625, 8.39970703125, 8.4144833984375, 8.4530791015625, 8.4308837890625, 8.3407841796875, 8.3356474609375, 8.31612109375, 8.32905859375, 8.3534638671875]",tokens/s,7.51771377038971,kWh,9.806880771561905e-05,5.3749043913185225e-05,0.00039216448579924195,0.0005439823374280462,tokens/kWh,115812.5837281126,,s,629,84.93340989685055,0.1350292685164556,0.016826908436194207,0.13322035217285155,0.13436354370117187,0.1346594787597656,0.27325439575195315,"[0.1383321533203125, 0.13691903686523438, 0.13555609130859375, 0.1336924133300781, 0.1340948486328125, 0.13436006164550782, 0.1343426513671875, 0.1342003173828125, 0.13426687622070313, 0.13448089599609375, 0.13400473022460938, 0.13385011291503907, 0.13449215698242187, 0.13411634826660157, 0.13182975769042968, 0.13270323181152344, 0.13469491577148437, 0.13397605895996093, 0.13359616088867188, 0.13410610961914063, 0.13234483337402345, 0.13145497131347655, 0.13187686157226564, 0.1315246124267578, 0.13338418579101563, 0.13427098083496095, 0.13404261779785157, 0.13425152587890626, 0.13394534301757813, 0.13425050354003906, 0.1328660430908203, 0.1342433319091797, 0.13415116882324218, 0.13333401489257812, 0.1342013397216797, 0.13399346923828126, 0.1340968933105469, 0.134508544921875, 0.13404876708984376, 0.1326940155029297, 0.1316812744140625, 0.13177658081054688, 0.13164845275878906, 0.13489971923828126, 0.13331251525878907, 0.1342248992919922, 0.13452799987792968, 0.13407334899902343, 0.13159321594238282, 0.13165260314941407, 0.13322035217285155, 0.13399244689941406, 0.13447474670410156, 0.13405696105957032, 0.13407948303222655, 0.13395046997070312, 0.13439283752441405, 0.1342300109863281, 0.13573222351074218, 0.1321625671386719, 0.13170687866210937, 0.13169970703125, 0.2737838134765625, 0.13152255249023437, 0.13275955200195313, 0.13433351135253907, 0.1345811767578125, 0.13403852844238281, 0.13368319702148437, 0.13302169799804686, 0.13411737060546874, 0.13280256652832031, 0.13474508666992188, 0.13335859680175782, 0.1340590057373047, 0.13169459533691405, 0.13173248291015624, 0.13171405029296876, 0.13407743835449218, 0.1342689208984375, 0.13365248107910158, 0.13403340148925783, 0.1339299774169922, 0.13434367370605468, 0.1315061798095703, 0.13148466491699218, 0.13182156372070314, 0.1337518005371094, 0.1333534698486328, 0.1340999755859375, 0.13400575256347655, 0.13407852172851562, 0.1340497283935547, 0.13159117126464845, 0.13308108520507814, 0.13406515502929686, 0.13437747192382812, 0.1340518341064453, 0.1342433319091797, 0.13421466064453125, 0.13415834045410155, 0.1342986297607422, 0.13406617736816406, 0.13197926330566406, 0.1316177978515625, 0.13170687866210937, 0.13170687866210937, 0.13363917541503906, 0.13486598205566405, 0.13352549743652345, 0.13383468627929687, 0.13421055603027343, 0.13174989318847657, 0.13169973754882813, 0.13152662658691405, 0.13283226013183594, 0.13399142456054688, 0.13478297424316407, 0.13450035095214843, 0.1341460418701172, 0.1339115447998047, 0.13337496948242186, 0.134002685546875, 0.13234687805175782, 0.13289779663085938, 0.2724710388183594, 0.13155020141601562, 0.13163827514648438, 0.131810302734375, 0.1316618194580078, 0.13126144409179688, 0.13352549743652345, 0.13428326416015626, 0.13351628112792968, 0.13398326110839845, 0.1340497589111328, 0.1338419189453125, 0.13386444091796876, 0.13370777893066407, 0.13431910705566405, 0.1338470458984375, 0.13402931213378907, 0.13414399719238282, 0.13386752319335937, 0.13391973876953125, 0.13366067504882811, 0.13545677185058594, 0.13440921020507812, 0.13534104919433593, 0.13244415283203126, 0.134466552734375, 0.13180108642578126, 0.131599365234375, 0.13178880310058594, 0.13154098510742188, 0.13177754211425782, 0.13156147766113283, 0.13368832397460936, 0.134044677734375, 0.13403347778320313, 0.13381727600097656, 0.13421772766113282, 0.1335265350341797, 0.13435289001464842, 0.133959716796875, 0.13362889099121095, 0.1339842529296875, 0.13384602355957032, 0.13403443908691406, 0.13420748901367188, 0.13402316284179688, 0.1340590057373047, 0.13474713134765626, 0.13409791564941406, 0.13352960205078124, 0.13388697814941405, 0.1340088348388672, 0.13404978942871093, 0.13377023315429687, 0.13412864685058593, 0.13397708129882813, 0.13398016357421874, 0.13395968627929689, 0.1340712890625, 0.13423004150390624, 0.13353570556640626, 0.13389414978027345, 0.13325209045410155, 0.27623629760742185, 0.134403076171875, 0.13430271911621094, 0.13406105041503907, 0.1344286651611328, 0.1346447296142578, 0.13446553039550782, 0.13406207275390625, 0.13451161193847655, 0.13428941345214843, 0.1343856658935547, 0.13401190185546874, 0.134466552734375, 0.13412249755859376, 0.13459762573242187, 0.1339463653564453, 0.13446553039550782, 0.13476864624023438, 0.1342044219970703, 0.13403443908691406, 0.13416653442382812, 0.134276123046875, 0.13387055969238282, 0.13414707946777343, 0.13453106689453126, 0.13373338317871095, 0.13383680725097657, 0.13395046997070312, 0.133791748046875, 0.13404570007324218, 0.1341265869140625, 0.13388697814941405, 0.1346693115234375, 0.1340518341064453, 0.1335562286376953, 0.1343969268798828, 0.13446348571777345, 0.1351065673828125, 0.1340641326904297, 0.13445733642578125, 0.13435903930664062, 0.1340631103515625, 0.1340518341064453, 0.13444816589355468, 0.13536051940917967, 0.13402621459960937, 0.13327052307128906, 0.13432012939453125, 0.1344593963623047, 0.13412249755859376, 0.1338419189453125, 0.13387469482421874, 0.13338726806640624, 0.13416653442382812, 0.13386854553222657, 0.13441127014160156, 0.13412351989746094, 0.13347328186035157, 0.13382246398925782, 0.13420338439941407, 0.1338275909423828, 0.13340467834472655, 0.1342750701904297, 0.27683428955078127, 0.13397196960449217, 0.1343129577636719, 0.1338665008544922, 0.13388082885742186, 0.1340958709716797, 0.1340712890625, 0.13391871643066405, 0.13356646728515625, 0.1333104705810547, 0.13409178161621094, 0.133396484375, 0.1351403503417969, 0.13372825622558593, 0.13417575073242188, 0.13519052124023437, 0.1343938903808594, 0.1338654327392578, 0.133359619140625, 0.1340712890625, 0.13396890258789063, 0.1339894104003906, 0.1338838653564453, 0.13374668884277344, 0.133538818359375, 0.13404876708984376, 0.13366169738769532, 0.1339351043701172, 0.13478195190429687, 0.13399859619140625, 0.13399449157714843, 0.13394432067871093, 0.13413682556152343, 0.1340262451171875, 0.13399449157714843, 0.13403546142578124, 0.13435289001464842, 0.1340712890625, 0.13406112670898437, 0.1335233917236328, 0.13358387756347656, 0.13382553100585937, 0.13504620361328126, 0.13462828063964843, 0.13439897155761718, 0.13393516540527345, 0.13389511108398439, 0.1334097900390625, 0.13408767700195312, 0.13458636474609376, 0.13346713256835938, 0.13414297485351562, 0.13160243225097656, 0.1317795867919922, 0.13214413452148438, 0.13376716613769532, 0.133718017578125, 0.13606605529785157, 0.13217485046386718, 0.13167718505859374, 0.13171302795410156, 0.13281280517578126, 0.13389926147460937, 0.27561163330078126, 0.1342044219970703, 0.13399655151367187, 0.13381837463378907, 0.13696818542480468, 0.13187277221679689, 0.13215437316894532, 0.13239193725585938, 0.1329521026611328, 0.13212156677246092, 0.13276876831054688, 0.132463623046875, 0.13214002990722656, 0.13234072875976563, 0.13270118713378906, 0.13262847900390626, 0.1321871337890625, 0.1320509490966797, 0.13238067626953126, 0.13276876831054688, 0.1321564178466797, 0.1327431640625, 0.1328158721923828, 0.13263360595703125, 0.1317580871582031, 0.1315215301513672, 0.13196083068847655, 0.1318225860595703, 0.13159117126464845, 0.13150003051757814, 0.13156454467773437, 0.13174887084960937, 0.1325496368408203, 0.1326755828857422, 0.13302169799804686, 0.13294898986816406, 0.1324881896972656, 0.13196800231933595, 0.13392588806152345, 0.1337159729003906, 0.13316915893554687, 0.1326755828857422, 0.13233255004882813, 0.13182054138183594, 0.13182566833496093, 0.13186253356933594, 0.13180519104003907, 0.1316822967529297, 0.13177548217773438, 0.13161984252929687, 0.1317611541748047, 0.131852294921875, 0.13162086486816407, 0.1321553955078125, 0.13262950134277343, 0.13191372680664062, 0.13178265380859375, 0.13150515747070313, 0.13170176696777344, 0.1315205078125, 0.13172735595703125, 0.13178778076171874, 0.13167103576660155, 0.2728120422363281, 0.13346815490722655, 0.1335142364501953, 0.1333534698486328, 0.13239910888671874, 0.13167616271972657, 0.13235302734375, 0.1338050537109375, 0.13347225952148437, 0.1334886474609375, 0.1320120391845703, 0.1320816650390625, 0.13306573486328124, 0.13341183471679688, 0.13176934814453126, 0.13158195495605468, 0.13284147644042968, 0.13212979125976562, 0.1316290588378906, 0.1317058563232422, 0.13171098327636718, 0.13157887268066407, 0.13160960388183593, 0.13298486328125, 0.13236732482910157, 0.13162188720703125, 0.13165260314941407, 0.13161062622070313, 0.13162495422363282, 0.13157376098632811, 0.13175910949707031, 0.13201100158691406, 0.13279539489746095, 0.13196493530273437, 0.13224140930175782, 0.13323365783691407, 0.13173248291015624, 0.13189529418945312, 0.133254150390625, 0.13235711669921876, 0.1316505584716797, 0.13167922973632812, 0.13156556701660158, 0.13257522583007814, 0.1328158721923828, 0.13173350524902344, 0.13212261962890626, 0.13244415283203126, 0.13331968688964843, 0.1319720916748047, 0.13162188720703125, 0.1338275909423828, 0.13188710021972655, 0.13169049072265626, 0.1329100799560547, 0.13289677429199218, 0.13157273864746094, 0.13191987609863282, 0.1330063934326172, 0.13177952575683594, 0.1324830780029297, 0.1317232666015625, 0.132210693359375, 0.2737438659667969, 0.13398220825195312, 0.13256195068359375, 0.13281686401367188, 0.13316409301757812, 0.13165664672851562, 0.131557373046875, 0.1315246124267578, 0.1316116485595703, 0.13375692749023438, 0.13275033569335937, 0.13160755920410155, 0.13167514038085937, 0.1316546630859375, 0.13151744079589844, 0.1325260772705078, 0.13415525817871093, 0.13211955261230468, 0.13158604431152343, 0.13158706665039063, 0.131599365234375, 0.13191885375976561, 0.13164959716796876, 0.13163615417480468, 0.131557373046875, 0.1315635223388672, 0.13156454467773437, 0.13150822448730468, 0.13155328369140626, 0.131778564453125, 0.13244825744628907, 0.13156761169433595, 0.13152870178222656, 0.13237452697753907, 0.13321420288085936, 0.1316188507080078, 0.13167100524902345, 0.13171507263183593, 0.13173043823242186, 0.13229055786132812, 0.13326028442382812, 0.13208883666992188, 0.13159730529785157, 0.13276570129394533, 0.13298074340820312, 0.13150210571289062, 0.13231517028808593, 0.13168019104003906, 0.1316177978515625, 0.13166490173339843, 0.13228031921386718, 0.131704833984375, 0.13154815673828124, 0.1315246124267578, 0.13159117126464845, 0.13162803649902344, 0.13144985961914063, 0.13153074645996093, 0.13159423828125, 0.13154304504394532, 0.1314959411621094, 0.13150822448730468, 0.13215335083007812, 0.2743265380859375, 0.13167922973632812, 0.13160243225097656, 0.13170278930664062, 0.13161068725585937, 0.13170375061035156, 0.13167820739746094, 0.13153996276855467, 0.13159628295898437, 0.13154917907714844, 0.13170994567871094, 0.13149491882324219, 0.1311682586669922, 0.1336678466796875, 0.13168333435058593, 0.1315020751953125, 0.13163827514648438, 0.1326510009765625, 0.13166490173339843, 0.131599365234375, 0.1316054992675781, 0.1315574035644531, 0.1316433563232422, 0.13159321594238282, 0.13151744079589844, 0.13169664001464843, 0.13158604431152343, 0.13171609497070313, 0.13180723571777345, 0.13306982421875, 0.1319505920410156, 0.13152665710449218, 0.1316188201904297, 0.13160755920410155, 0.13153485107421875, 0.1316321258544922, 0.13152255249023437, 0.13150413513183593, 0.13162393188476562, 0.13158604431152343, 0.13178880310058594, 0.13175398254394532, 0.13282815551757812, 0.13279335021972657, 0.1332623291015625, 0.13260493469238283, 0.1335029754638672, 0.1333217315673828, 0.13349990844726561, 0.13284454345703126, 0.13330943298339842, 0.13328793334960937, 0.13312818908691407, 0.13345382690429688, 0.13288447570800782, 0.13335142517089843, 0.1332725830078125, 0.1326561279296875, 0.13411943054199219, 0.1333104705810547, 0.13202330017089844, 0.1328404541015625, 0.13351219177246093, 0.27342642211914064, 0.1317181396484375, 0.131778564453125, 0.1316300811767578, 0.13164851379394532, 0.13157376098632811, 0.13182464599609375, 0.1316505584716797, 0.1314580535888672, 0.1316853790283203, 0.13220352172851563, 0.13342617797851564, 0.13168435668945314, 0.1316259765625, 0.13177754211425782, 0.13159423828125, 0.13173356628417968, 0.1318133087158203, 0.1316444091796875, 0.1315082550048828, 0.13163209533691406, 0.13170994567871094, 0.13161062622070313, 0.13163929748535155, 0.13251788330078124, 0.13330738830566408, 0.13480653381347657, 0.13196902465820312, 0.1324451904296875, 0.13310054016113282, 0.13354290771484376, 0.13340882873535156, 0.1336657257080078, 0.13366886901855468, 0.133396484375, 0.13281893920898438, 0.13362892150878905, 0.13327769470214842, 0.1322239990234375, 0.13191474914550783, 0.13346099853515625, 0.1332162628173828, 0.13334835815429688, 0.13336679077148436, 0.13285171508789062, 0.13309132385253905, 0.13161369323730468, 0.13310464477539063, 0.13307391357421874, 0.13365248107910158, 0.13310464477539063, 0.13244313049316406, 0.1326376953125, 0.13354908752441405, 0.13300936889648438, 0.13332179260253907, 0.13388691711425782, 0.13313023376464844, 0.1333780517578125, 0.13311077880859376, 0.13335551452636718, 0.13207347106933592, 0.1331865539550781]",tokens/s,7.405801801245284,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-14B,Qwen/Qwen-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 550, in from_pretrained + model_class = get_class_from_dynamic_module( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 498, in get_class_from_dynamic_module + final_module = get_cached_module_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 323, in get_cached_module_file + modules_needed = check_imports(resolved_module_file) + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 181, in check_imports + raise ImportError( +ImportError: This modeling file requires the following packages that were not found in your environment: transformers_stream_generator. Run `pip install transformers_stream_generator` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-7.5B,facebook/xglm-7.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1469, in _autoset_attn_implementation + cls._check_and_enable_flash_attn_2( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1560, in _check_and_enable_flash_attn_2 + raise ValueError( +ValueError: XGLMForCausalLM does not support Flash Attention 2.0 yet. Please request to add support where the model is hosted, on its model hub page: https://huggingface.co//tmp/tmp2h78w9mi/no_weights_model/discussions/new or in the Transformers GitHub repo: https://github.com/huggingface/transformers/issues/new + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-12b,EleutherAI/pythia-12b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,1531.768832,9676.783616,0.0,9030.336512,8583.572992,s,10,9.471071899414062,0.9471071899414062,0.0007865321197485112,0.9470617980957031,0.9481417907714844,0.9482434417724609,0.9483247625732422,"[0.9481192016601563, 0.9483450927734375, 0.9464827270507813, 0.9463027954101563, 0.9464989624023438, 0.9459486694335938, 0.9466808471679687, 0.9474427490234375, 0.9474686889648437, 0.9477821655273437]",tokens/s,270.2967549172948,kWh,1.1175276083175583e-05,6.1235719211254035e-06,5.30962293456365e-05,7.039507734993748e-05,tokens/kWh,3636617.9232591945,MB,1531.768832,9676.783616,0.0,9030.336512,8872.966144,s,10,564.2868125,56.42868125,0.005902619636999738,56.430064453125,56.434701171875,56.4354833984375,56.4361091796875,"[56.42066796875, 56.42100390625, 56.427828125, 56.43380078125, 56.436265625, 56.4306328125, 56.43452734375, 56.42949609375, 56.41955859375, 56.43303125]",tokens/s,1.116453523357858,kWh,0.0006659982151861745,0.00036502579988969956,0.003165337650954964,0.004196361666030838,tokens/kWh,15013.005316004865,,s,629,571.914466552734,0.9092439849804999,0.11250283527003516,0.8956805419921875,0.8961118286132812,0.8962535400390624,1.84229025390625,"[0.8955299682617187, 0.8956344604492188, 0.8952893676757813, 0.8954654541015625, 0.8952473754882813, 0.8954511108398437, 0.8954111938476562, 0.895388671875, 0.8953507690429687, 0.8954296264648437, 0.8954921264648438, 0.8958034057617188, 0.8957071533203125, 0.895331298828125, 0.89537841796875, 0.8954317016601563, 0.895688720703125, 0.8951869506835938, 0.89523095703125, 0.895341552734375, 0.8953385009765625, 0.89526171875, 0.8952105712890625, 0.8952554931640625, 0.8955637817382812, 0.8952135620117188, 0.895278076171875, 0.89569384765625, 0.8954439697265625, 0.8954368286132812, 0.8953395385742188, 0.895320068359375, 0.895224853515625, 0.8956580200195312, 0.8956251220703125, 0.895593505859375, 0.8955872802734375, 0.895393798828125, 0.89556591796875, 0.8955975341796875, 0.8953139038085938, 0.8952350463867188, 0.8954644775390626, 0.8952453002929688, 0.8950845336914063, 0.89560888671875, 0.8954244384765625, 0.8954317016601563, 0.8957972412109375, 0.8959805297851563, 0.8959723510742188, 0.896089111328125, 0.8958925170898437, 0.8958883666992188, 0.8961648559570312, 0.8958853149414062, 0.896374755859375, 0.896205810546875, 0.8961065063476562, 0.89630517578125, 0.89601025390625, 0.8961126098632812, 1.844611083984375, 0.8954337158203125, 0.89558837890625, 0.8955914306640625, 0.8959385375976563, 0.8956160278320312, 0.8950343627929688, 0.8949083862304688, 0.8952064208984375, 0.895373291015625, 0.8951787719726563, 0.895162353515625, 0.8951439208984375, 0.895182861328125, 0.895109130859375, 0.8953907470703125, 0.8957255859375, 0.8957091674804688, 0.8956385498046875, 0.8958136596679688, 0.8958320922851563, 0.8952658081054687, 0.8952637329101563, 0.8951807861328125, 0.8952391967773438, 0.8951900024414062, 0.8952842407226562, 0.8955750122070313, 0.89556787109375, 0.8954613647460937, 0.895562744140625, 0.8955679321289063, 0.8955903930664062, 0.8953548583984375, 0.8952647705078125, 0.8954501342773438, 0.8954869995117187, 0.89586181640625, 0.8957409057617187, 0.8956713256835938, 0.8955576171875, 0.8954132690429687, 0.8955822143554687, 0.89568359375, 0.8954531860351562, 0.8953220825195313, 0.89735986328125, 0.8953026733398437, 0.8954552612304687, 0.8961781616210938, 0.8958126220703125, 0.8957644653320312, 0.8956436767578125, 0.8958248901367187, 0.896047119140625, 0.8958515014648437, 0.8958351440429687, 0.8958453979492188, 0.895657958984375, 0.8957132568359375, 0.8961444091796875, 0.8957071533203125, 0.8954951782226562, 1.842377685546875, 0.895404052734375, 0.8950650634765625, 0.8951746826171875, 0.8952135620117188, 0.895152099609375, 0.8950364379882813, 0.8954869995117187, 0.8951756591796876, 0.8951070556640625, 0.89539794921875, 0.8955360717773437, 0.8956303100585937, 0.895278076171875, 0.8953108520507812, 0.895466552734375, 0.8955247802734375, 0.8954808349609376, 0.89577880859375, 0.8958197631835938, 0.8954214477539062, 0.8957296752929688, 0.8957849731445312, 0.8969154663085938, 0.8955699462890625, 0.8955740356445312, 0.8955402221679688, 0.8957010498046875, 0.8956046752929687, 0.8959457397460937, 0.89588427734375, 0.895657958984375, 0.8959969482421875, 0.8955279541015625, 0.89569384765625, 0.8956375122070312, 0.8958167114257812, 0.89558837890625, 0.895529052734375, 0.89546337890625, 0.8958197631835938, 0.8958023681640624, 0.895668212890625, 0.895762451171875, 0.8955596923828125, 0.89588427734375, 0.8957982788085938, 0.896742431640625, 0.8959180297851562, 0.895889404296875, 0.8958699340820313, 0.8959774780273437, 0.8961423950195313, 0.8960029907226562, 0.895941650390625, 0.8961116333007813, 0.8960184326171875, 0.8960419921875, 0.8958904418945313, 0.8957849731445312, 0.8956774291992188, 0.8956559448242187, 0.8954009399414062, 1.84203369140625, 0.8953282470703126, 0.8955074462890625, 0.8958883666992188, 0.895805419921875, 0.895224853515625, 0.8955054321289062, 0.89569482421875, 0.8954593505859375, 0.895592529296875, 0.89609619140625, 0.89571533203125, 0.89652734375, 0.8958136596679688, 0.8957890625, 0.8957183837890625, 0.8957440185546875, 0.8959334106445312, 0.895752197265625, 0.8961761474609375, 0.8955248413085938, 0.896421875, 0.895963134765625, 0.895330322265625, 0.8955340576171875, 0.8957603759765626, 0.8954306640625, 0.8958146362304688, 0.895805419921875, 0.8954982299804688, 0.8955299682617187, 0.8953681640625, 0.8958136596679688, 0.8957706298828125, 0.89615869140625, 0.8956036987304687, 0.8954429321289062, 0.895578125, 0.8957962036132813, 0.8954183959960937, 0.8954849243164062, 0.89569384765625, 0.8957583618164062, 0.8959283447265625, 0.8975103759765625, 0.8956825561523437, 0.8957440185546875, 0.895594482421875, 0.8960829467773438, 0.8958392333984375, 0.8954531860351562, 0.8957625122070313, 0.8960316772460938, 0.8954849243164062, 0.8954644775390626, 0.8962181396484376, 0.895805419921875, 0.8961618041992188, 0.8958760986328125, 0.8960430297851563, 0.89603173828125, 0.8957348022460937, 0.8958709716796875, 1.8425006103515624, 0.8954132690429687, 0.8952770385742187, 0.8951552124023437, 0.8951265258789063, 0.8949452514648437, 0.8951756591796876, 0.8958238525390625, 0.895236083984375, 0.895140869140625, 0.8951900024414062, 0.8952207641601563, 0.895283203125, 0.8956876831054688, 0.8953272094726562, 0.89558837890625, 0.8958331298828125, 0.896849853515625, 0.8957357788085938, 0.8959354858398437, 0.8956190795898438, 0.8957183837890625, 0.89544189453125, 0.8957081298828125, 0.8961085205078125, 0.8960030517578125, 0.895741943359375, 0.8960706787109375, 0.8960758056640625, 0.8956846313476563, 0.8959539794921875, 0.8957930908203126, 0.8956805419921875, 0.8956405639648437, 0.8957777709960938, 0.895805419921875, 0.8960061645507813, 0.8956876831054688, 0.8957982788085938, 0.8960235595703125, 0.896294921875, 0.8963491821289062, 0.8959744262695313, 0.8960379028320312, 0.8958555908203125, 0.8957880249023438, 0.8960327758789063, 0.8959774780273437, 0.8963082275390625, 0.8961085205078125, 0.8957849731445312, 0.8961433715820313, 0.896015380859375, 0.8959569702148438, 0.89592626953125, 0.895857666015625, 0.89588427734375, 0.8977633056640625, 0.8960972900390625, 0.8961024169921875, 0.8960040893554687, 0.8959642333984374, 0.895754150390625, 1.8420654296875, 0.895541259765625, 0.8955330810546875, 0.895551513671875, 0.8954542846679687, 0.89588525390625, 0.8957655639648437, 0.8955913696289063, 0.8954900512695313, 0.895973388671875, 0.8957081298828125, 0.895446044921875, 0.8953692016601562, 0.8954695434570312, 0.8955350952148438, 0.8955064086914063, 0.8957685546875, 0.895510498046875, 0.8955617065429687, 0.89569384765625, 0.8958258666992187, 0.8956436767578125, 0.8955709228515625, 0.8953487548828125, 0.895425537109375, 0.89541015625, 0.8953661499023438, 0.8959344482421875, 0.8957112426757813, 0.8955719604492187, 0.8957173461914063, 0.8952852783203125, 0.895425537109375, 0.8956323852539062, 0.8953405151367188, 0.8953170166015625, 0.8956170043945313, 0.89562109375, 0.8976466064453125, 0.8961249389648438, 0.8960225830078125, 0.8955401611328125, 0.8958443603515625, 0.8955822143554687, 0.8957726440429687, 0.8958177490234375, 0.8958822631835938, 0.8957255859375, 0.89603173828125, 0.8960040893554687, 0.8958167114257812, 0.8959232177734375, 0.8961556396484375, 0.895921142578125, 0.8957132568359375, 0.8960726928710937, 0.8958699340820313, 0.8956088256835938, 0.8956508178710938, 0.8958750610351562, 0.8959140014648438, 0.896184326171875, 0.8957747192382812, 1.8425538330078124, 0.8953139038085938, 0.8954296264648437, 0.8954317016601563, 0.8954019775390625, 0.8954296264648437, 0.895804443359375, 0.89565185546875, 0.8956016845703125, 0.8954736938476563, 0.8951910400390625, 0.8951807861328125, 0.8956160278320312, 0.8953865966796875, 0.895224853515625, 0.8951572265625, 0.895447021484375, 0.8951286010742188, 0.8952534790039063, 0.8959088745117187, 0.8953425903320312, 0.8954378051757812, 0.8953170166015625, 0.8966533203125, 0.8956589965820313, 0.8957174072265625, 0.8958781127929687, 0.8958924560546875, 0.89592626953125, 0.895657958984375, 0.8963993530273437, 0.8960133056640625, 0.8955350952148438, 0.895636474609375, 0.8958023681640624, 0.8959241943359375, 0.8960481567382812, 0.8959610595703125, 0.8962498779296875, 0.8960297241210937, 0.8959395751953125, 0.8961136474609375, 0.896173095703125, 0.8961300048828125, 0.8959856567382812, 0.8956856079101563, 0.8959078369140625, 0.8958648071289063, 0.8959365234375, 0.8959959106445312, 0.89580029296875, 0.8960338134765625, 0.8962744140625, 0.8961085205078125, 0.896195556640625, 0.8962867431640625, 0.8956589965820313, 0.896268310546875, 0.8962437133789063, 0.896083984375, 0.895952880859375, 0.8960859985351562, 0.8958607177734375, 1.8435286865234375, 0.8953579711914063, 0.8959723510742188, 0.895552490234375, 0.8962406616210937, 0.895710205078125, 0.8959600830078125, 0.895594482421875, 0.8955453491210937, 0.8953630981445313, 0.8957020263671875, 0.8955637817382812, 0.895478759765625, 0.8956026611328125, 0.89554736328125, 0.8955330810546875, 0.8961474609375, 0.895688720703125, 0.8956405639648437, 0.89556787109375, 0.8954163208007813, 0.8957962036132813, 0.8957061157226562, 0.8955453491210937, 0.895515625, 0.8955586547851563, 0.8955166625976563, 0.8954368286132812, 0.8956876831054688, 0.8957532348632813, 0.8957552490234375, 0.89554638671875, 0.8954439697265625, 0.8958248901367187, 0.8957378540039063, 0.8956109008789063, 0.895552490234375, 0.8955668334960938, 0.8956426391601563, 0.8957552490234375, 0.8958330688476562, 0.8957849731445312, 0.8957276000976563, 0.895825927734375, 0.8963133544921875, 0.8962559814453125, 0.89609423828125, 0.8956958618164063, 0.8961679077148438, 0.8956907348632812, 0.8954450073242187, 0.8954439697265625, 0.8957511596679687, 0.8955791625976562, 0.8954685668945312, 0.8961618041992188, 0.8959201049804687, 0.8955299682617187, 0.8956661987304687, 0.8956989135742187, 0.8956763916015625, 0.895662109375, 0.8958924560546875, 1.8434007568359374, 0.8952268676757813, 0.8956057739257812, 0.8958668823242187, 0.8953005981445312, 0.8950661010742188, 0.8951449584960938, 0.8954685668945312, 0.8951981811523437, 0.8952791137695313, 0.8951910400390625, 0.895224853515625, 0.8951367797851563, 0.8953661499023438, 0.8954818725585938, 0.8955637817382812, 0.8956846313476563, 0.8963553466796875, 0.8955187377929688, 0.89537841796875, 0.8953743286132813, 0.895267822265625, 0.8952647705078125, 0.8953159790039062, 0.895520751953125, 0.8957470703125, 0.8953702392578125, 0.8953262329101562, 0.8954111938476562, 0.8956661987304687, 0.8953436279296875, 0.8954030151367187, 0.895283203125, 0.8952268676757813, 0.89514599609375, 0.8952166137695312, 0.8958832397460937, 0.89552587890625, 0.8954111938476562, 0.895561767578125, 0.8955135498046874, 0.895520751953125, 0.8956928100585938, 0.8956426391601563, 0.8954716186523437, 0.8955125732421875, 0.8955330810546875, 0.8961566772460937, 0.8961351928710938, 0.8959539184570312, 0.8957091674804688, 0.89584228515625, 0.8958197631835938, 0.8957470703125, 0.8957593383789062, 0.8957214965820313, 0.8955811767578125, 0.8955985717773437, 0.8961331176757813, 0.8957828979492187, 0.8960563354492187, 0.8956763916015625, 0.8957511596679687, 1.8439761962890624, 0.8954634399414062, 0.89548388671875, 0.8955084838867188, 0.89558837890625, 0.895562744140625, 0.8955801391601562, 0.895515625, 0.8953907470703125, 0.8954337158203125, 0.8955177001953125, 0.8955903930664062, 0.8959283447265625, 0.8955166625976563, 0.89547265625, 0.8954685668945312, 0.895752197265625, 0.895425537109375, 0.8954317016601563, 0.8954531860351562, 0.8953671875, 0.89560986328125, 0.8953211059570313, 0.8955269165039063, 0.8957849731445312, 0.8958924560546875, 0.8968284301757813, 0.8958668823242187, 0.8959989624023438, 0.8959027099609375, 0.8958955688476562, 0.8957982788085938, 0.8957511596679687, 0.895730712890625, 0.8959959106445312, 0.8958034057617188, 0.895724609375, 0.8958135375976563, 0.8960409545898438, 0.8961474609375, 0.8957860107421876, 0.8955709228515625, 0.8958484497070313, 0.89590576171875, 0.895847412109375, 0.895763427734375, 0.8960266723632813, 0.8961084594726563, 0.8961710205078125, 0.89615869140625, 0.8962139892578125, 0.895963134765625, 0.8957614135742188, 0.8957449951171875, 0.8958822631835938, 0.8960726928710937, 0.8957849731445312, 0.8957849731445312, 0.895599609375, 0.895626220703125, 0.8956661987304687, 0.8960829467773438, 0.8956692504882813]",tokens/s,1.09981480935664,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-7B,Qwen/Qwen1.5-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,2832.146432,8389.132288,0.0,7742.685184,7007.0144,s,10,5.830655212402344,0.5830655212402344,0.0011864011955084602,0.5831084289550781,0.5846742309570313,0.5847881042480468,0.5848792028808594,"[0.5832706909179688, 0.5849019775390625, 0.5821006469726563, 0.5817232055664062, 0.582320068359375, 0.5811737670898437, 0.5829461669921875, 0.5834404907226562, 0.5841292724609375, 0.58464892578125]",tokens/s,439.05871754423805,kWh,6.867315049892591e-06,3.7630052303029514e-06,3.309072091699999e-05,4.372104119719553e-05,tokens/kWh,5855304.288051151,MB,2832.146432,8389.132288,0.0,7742.685184,7283.984384,s,10,342.646671875,34.264667187499995,0.008129823876006116,34.26347265625,34.275117968749996,34.278303125,34.28085125,"[34.26221484375, 34.28148828125, 34.27441015625, 34.256734375, 34.26859765625, 34.26473046875, 34.2578125, 34.267296875, 34.25966015625, 34.2537265625]",tokens/s,1.8386286857904421,kWh,0.0004044825890237167,0.00022169153025011835,0.0019209350367467965,0.002547109156020632,tokens/kWh,24733.92231781122,,s,629,347.34827374267576,0.5522230107196754,0.06935218607240677,0.5438034057617187,0.5445187377929688,0.5449166748046874,1.1266134716796876,"[0.54485400390625, 0.5438034057617187, 0.543246337890625, 0.5442498779296875, 0.5438187255859375, 0.544405517578125, 0.5433927612304688, 0.5439907836914063, 0.5433046875, 0.5437378540039063, 0.5438269653320312, 0.5440072021484375, 0.543795166015625, 0.544226318359375, 0.5433211059570312, 0.5441300659179688, 0.54451708984375, 0.5443307495117188, 0.5436416015625, 0.5442396240234375, 0.5436170043945312, 0.545089599609375, 0.543943603515625, 0.5440245971679688, 0.5435012817382813, 0.54389453125, 0.543447021484375, 0.544269287109375, 0.5434480590820312, 0.5442416381835937, 0.5436282958984375, 0.5436856079101563, 0.5433425903320312, 0.5441137084960938, 0.5437623901367188, 0.5438228759765625, 0.54314599609375, 0.543266845703125, 0.5434480590820312, 0.54394677734375, 0.543478759765625, 0.544342041015625, 0.5433477172851563, 0.5437071533203125, 0.5434163208007813, 0.5436990356445313, 0.5434121704101562, 0.5436589965820312, 0.5431746826171875, 0.5436907348632812, 0.5431572265625, 0.5437921142578125, 0.5443922119140625, 0.5438034057617187, 0.5433538818359375, 0.5437992553710937, 0.54312548828125, 0.5436221313476562, 0.5442550048828125, 0.5447546997070313, 0.5435003051757813, 0.544110595703125, 1.129606201171875, 0.5438555908203125, 0.5446000366210938, 0.5434962158203125, 0.5441279907226563, 0.5435760498046875, 0.5443614501953125, 0.5440512084960938, 0.5445314331054687, 0.544, 0.544395263671875, 0.5442437133789062, 0.5449338989257813, 0.5445150756835937, 0.5455985107421875, 0.5442651977539062, 0.5440481567382812, 0.5437183837890625, 0.5441484985351562, 0.5433558959960938, 0.544912353515625, 0.5442252807617187, 0.544775146484375, 0.5442785034179688, 0.5442467651367188, 0.5443768310546875, 0.544701416015625, 0.5438463745117188, 0.5441720581054688, 0.543763427734375, 0.5439447021484375, 0.5434183959960938, 0.5442539672851563, 0.543873046875, 0.5442447509765626, 0.5435484008789062, 0.544189453125, 0.5435852661132813, 0.5441228637695312, 0.5439774780273438, 0.5444413452148438, 0.5440932006835938, 0.5443020629882812, 0.5437276000976563, 0.5443696899414062, 0.5436436767578126, 0.5473484497070312, 0.5439129638671875, 0.5444976806640625, 0.54388427734375, 0.5441033935546875, 0.5436795043945313, 0.5439283447265625, 0.5437163696289062, 0.5439866943359375, 0.5433651123046875, 0.5438566284179688, 0.5434593505859375, 0.5443276977539062, 0.5436395263671875, 0.54405224609375, 0.5434828491210938, 0.5446829833984375, 1.1269549560546874, 0.5436589965820312, 0.5440983276367187, 0.5438801879882813, 0.5444874267578125, 0.544152587890625, 0.5444679565429688, 0.5439887084960937, 0.5441515502929688, 0.5438320922851563, 0.54458984375, 0.5438555908203125, 0.544068603515625, 0.5435648193359375, 0.5439713134765625, 0.5431132202148438, 0.5440512084960938, 0.54389453125, 0.5443778686523437, 0.543446044921875, 0.5438597412109375, 0.5438648071289063, 0.5439907836914063, 0.5458063354492187, 0.5444915161132813, 0.543572998046875, 0.5443184814453125, 0.5437357788085937, 0.543878173828125, 0.5435125732421875, 0.5439928588867188, 0.5434962158203125, 0.544068603515625, 0.5433712768554687, 0.5437890625, 0.5446963500976563, 0.5459650268554688, 0.5440255737304688, 0.5438883666992187, 0.5443931884765625, 0.5446287231445313, 0.5444362182617187, 0.5449840698242188, 0.5441444091796875, 0.5444034423828125, 0.543510498046875, 0.5440491333007812, 0.5454448852539062, 0.5443717041015625, 0.5432412109375, 0.5437112426757813, 0.5432013549804687, 0.5434971313476562, 0.5437726440429688, 0.5439866943359375, 0.5433477172851563, 0.5437194213867188, 0.5431173095703125, 0.5434132690429687, 0.5431746826171875, 0.544648193359375, 0.5442304077148438, 0.5442119750976563, 1.1263170166015626, 0.5434583129882813, 0.5442969360351563, 0.543373291015625, 0.5437562866210938, 0.543373291015625, 0.5437644653320313, 0.5435750122070313, 0.5440972900390625, 0.5432647705078125, 0.5439580078125, 0.5433252563476563, 0.5438104858398437, 0.5432924194335937, 0.5439324340820313, 0.5438750610351563, 0.5440696411132813, 0.543283203125, 0.543783935546875, 0.54326171875, 0.5440430297851563, 0.543267822265625, 0.5438822631835938, 0.54329443359375, 0.5437296752929688, 0.5433179931640625, 0.5443358764648437, 0.5437020263671875, 0.5440706787109375, 0.5436016845703125, 0.5435985717773437, 0.54325244140625, 0.5440829467773437, 0.5445253295898438, 0.5435402221679687, 0.54327294921875, 0.5437296752929688, 0.5436436767578126, 0.54380029296875, 0.5434234619140625, 0.5437747192382812, 0.5434429321289063, 0.5436068115234375, 0.543056884765625, 0.5473935546875, 0.54410546875, 0.5440604248046875, 0.5434931030273438, 0.5439365234375, 0.5433773803710937, 0.5438289794921874, 0.5436354370117188, 0.5443809204101563, 0.5432913818359375, 0.5437869873046876, 0.5433190307617187, 0.5438975830078125, 0.54358837890625, 0.543973388671875, 0.5433487548828125, 0.5437777709960937, 0.5435279541015625, 0.5438392333984375, 1.126728759765625, 0.5433446655273437, 0.544162841796875, 0.5432760620117187, 0.5441720581054688, 0.5443881225585937, 0.54490625, 0.54394677734375, 0.5442590942382812, 0.5439252319335938, 0.544701416015625, 0.5439826049804688, 0.5446226196289062, 0.5433385009765626, 0.5438064575195313, 0.5434009399414063, 0.5440983276367187, 0.5450567626953124, 0.5439652099609374, 0.5432319946289063, 0.5438095092773437, 0.5432586059570312, 0.5437614135742187, 0.543435791015625, 0.544484375, 0.5434317016601562, 0.5437009887695312, 0.5431490478515625, 0.5438167114257813, 0.543298583984375, 0.5435801391601562, 0.5435494995117187, 0.5441719970703125, 0.5433681640625, 0.5436928100585937, 0.5437245483398437, 0.5445037841796875, 0.5434193725585937, 0.5443307495117188, 0.543531005859375, 0.5440133056640625, 0.5436139526367187, 0.5439487915039063, 0.54411572265625, 0.5449932861328125, 0.5437962036132813, 0.5439794921875, 0.543466552734375, 0.5439456787109375, 0.543562744140625, 0.5440255737304688, 0.5437542114257813, 0.5439508666992188, 0.5435565795898437, 0.5450424194335938, 0.5437081298828125, 0.5441198120117188, 0.5460613403320312, 0.5440328369140625, 0.5436876220703125, 0.5439970092773437, 0.5443685913085937, 0.5439518432617187, 1.1256268310546875, 0.5433876342773437, 0.5439027099609375, 0.5437491455078125, 0.5440143432617187, 0.5435873413085938, 0.5440634765625, 0.5435217895507812, 0.544837646484375, 0.5441290283203125, 0.5442662353515625, 0.5436077880859375, 0.5440040893554687, 0.5432811279296875, 0.5437716674804688, 0.5439754028320313, 0.543921142578125, 0.5436795043945313, 0.5448601684570312, 0.543963134765625, 0.5444270629882813, 0.5435064086914062, 0.5439723510742187, 0.5439713134765625, 0.5442775268554687, 0.5437449951171875, 0.544595947265625, 0.5438966064453125, 0.5437860107421875, 0.5432279052734375, 0.5435350952148438, 0.5433651123046875, 0.5438668823242188, 0.5433426513671875, 0.5449727783203125, 0.5435340576171875, 0.5439334106445313, 0.54326171875, 0.5467739868164062, 0.5434767456054688, 0.5440921630859376, 0.5435054931640625, 0.544133056640625, 0.5435484008789062, 0.5440020751953125, 0.5437808837890625, 0.5440993041992187, 0.5436477661132812, 0.5438197631835937, 0.5434757690429688, 0.54377978515625, 0.5431531372070313, 0.5442908325195313, 0.5436088256835937, 0.5439539184570312, 0.54348291015625, 0.54430517578125, 0.5432504272460937, 0.5439170532226563, 0.5432053833007813, 0.5438597412109375, 0.5436006469726562, 0.543825927734375, 1.1284039306640625, 0.5435811767578125, 0.54392626953125, 0.5432975463867188, 0.5440030517578125, 0.5434276123046875, 0.5444700317382812, 0.5436016845703125, 0.5438228759765625, 0.5433507690429688, 0.5437460327148438, 0.5433599853515625, 0.5435975952148437, 0.5432709350585937, 0.54407373046875, 0.5433487548828125, 0.5436477661132812, 0.5439201049804687, 0.544205810546875, 0.5434214477539062, 0.543847412109375, 0.5432391967773438, 0.5437501220703125, 0.54501171875, 0.543856689453125, 0.5445160522460938, 0.544321533203125, 0.5439661865234375, 0.5444495849609375, 0.5435463256835937, 0.543562744140625, 0.5435699462890625, 0.5438617553710937, 0.5433108520507812, 0.5436497802734375, 0.5432432861328125, 0.5446655883789062, 0.5437214965820313, 0.5437655029296875, 0.543151123046875, 0.5436876831054688, 0.54325146484375, 0.5437511596679687, 0.5447864379882813, 0.5438218994140624, 0.5431879272460938, 0.543825927734375, 0.543182861328125, 0.5435903930664062, 0.54308349609375, 0.5436334228515625, 0.5431613159179688, 0.5436016845703125, 0.5431839599609375, 0.5440931396484375, 0.544869384765625, 0.543978515625, 0.5436641235351563, 0.5438402709960938, 0.54329443359375, 0.5435975952148437, 0.5432872924804687, 0.5449031982421875, 1.12785302734375, 0.5432197265625, 0.5443031005859374, 0.543805419921875, 0.5440696411132813, 0.543562744140625, 0.5437880249023438, 0.5434265747070313, 0.54437890625, 0.5443410034179688, 0.5438085327148438, 0.5433548583984374, 0.5437552490234375, 0.5441167602539062, 0.5438873901367187, 0.5431286010742188, 0.5460684814453125, 0.5433446655273437, 0.54401123046875, 0.5432525024414062, 0.5443113403320312, 0.5433211059570312, 0.5440449829101562, 0.54329443359375, 0.5437286376953125, 0.5433364868164062, 0.5437962036132813, 0.5443225708007813, 0.544879638671875, 0.543446044921875, 0.543447021484375, 0.5436170043945312, 0.5437798461914063, 0.543077392578125, 0.5447864379882813, 0.5435914306640625, 0.5437706298828126, 0.5436293334960938, 0.5449195556640625, 0.5433282470703125, 0.544373779296875, 0.543984619140625, 0.54413720703125, 0.5432801513671875, 0.5436323852539062, 0.5439119262695312, 0.5450363159179688, 0.543909912109375, 0.5446492309570312, 0.5441402587890625, 0.5445682983398438, 0.5445365600585937, 0.5437798461914063, 0.5442201538085938, 0.5439365234375, 0.5434225463867187, 0.5440900268554687, 0.5439365844726562, 0.5439845581054688, 0.5436405639648437, 0.5438587036132813, 0.5438146362304688, 0.5440071411132813, 1.1298836669921875, 0.5435586547851563, 0.5440675659179688, 0.5432913818359375, 0.5438289794921874, 0.5433005981445312, 0.5437767944335937, 0.544300048828125, 0.54409521484375, 0.543405029296875, 0.54519091796875, 0.5432658081054688, 0.5438023681640625, 0.5436221313476562, 0.5439180908203125, 0.5432565307617188, 0.5435023193359375, 0.5447024536132813, 0.54401953125, 0.5435442504882813, 0.5441484985351562, 0.5433333740234375, 0.5439180908203125, 0.5433262329101562, 0.5437921142578125, 0.5436846313476562, 0.5451581420898437, 0.54331494140625, 0.5439129638671875, 0.5431747436523438, 0.5436364135742188, 0.543467529296875, 0.5444925537109375, 0.5434685668945313, 0.5439539184570312, 0.543331298828125, 0.5441863403320313, 0.5441474609375, 0.5438177490234375, 0.5437880249023438, 0.5436282958984375, 0.5433865966796875, 0.5437327270507812, 0.5433026733398437, 0.5443440551757812, 0.5432484130859375, 0.5437440185546875, 0.5434849243164063, 0.5438463745117188, 0.5431654663085937, 0.5436897583007813, 0.5432340698242187, 0.5437265625, 0.5432975463867188, 0.543541259765625, 0.5436201171875, 0.5445703735351562, 0.543752197265625, 0.5440389404296875, 0.5439641723632812, 0.5437122802734375, 0.543457275390625, 0.5449390258789063, 1.1290306396484375, 0.5434531860351562, 0.5440993041992187, 0.5435391845703125, 0.5443635864257812, 0.54331591796875, 0.5436958618164063, 0.54315625, 0.54371533203125, 0.5437439575195312, 0.544100341796875, 0.5435381469726562, 0.5437388916015625, 0.5432217407226563, 0.5436754150390625, 0.54318896484375, 0.5440829467773437, 0.544564208984375, 0.543515625, 0.5431910400390625, 0.5438494873046875, 0.54326171875, 0.5435248413085938, 0.5433456420898437, 0.5437112426757813, 0.5435238647460937, 0.5456578369140626, 0.5435504760742188, 0.5442672729492187, 0.543446044921875, 0.54443212890625, 0.5432156372070313, 0.5440870971679688, 0.5433077392578125, 0.5435750122070313, 0.5440040893554687, 0.5438587036132813, 0.5430794067382813, 0.5436282958984375, 0.5436846313476562, 0.5436190795898438, 0.5430947875976563, 0.5436282958984375, 0.5432340698242187, 0.5435166625976563, 0.5432954711914062, 0.5439907836914063, 0.5439354858398437, 0.5442140502929688, 0.543595458984375, 0.544216064453125, 0.5438013305664062, 0.5442723999023438, 0.5444034423828125, 0.5439794921875, 0.54333642578125, 0.5436118774414063, 0.5431787719726563, 0.5439519653320313, 0.5432032470703125, 0.5435668334960938, 0.5430866088867188, 0.5438269653320312]",tokens/s,1.810862605483909,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-34B,,cuda,0,42,,,,,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,,MB,1743.122432,22129.672192,0.0,21483.225088,20799.036928,s,10,27.909502197265624,2.7909502197265623,0.0030225550178342967,2.7914050292968753,2.79439267578125,2.794928149414062,2.7953565283203123,"[2.794273681640625, 2.795463623046875, 2.7873525390625, 2.789448486328125, 2.787631103515625, 2.78628466796875, 2.790373291015625, 2.792436767578125, 2.79304931640625, 2.793188720703125]",tokens/s,91.72503263962948,kWh,3.290233454770512e-05,1.803171549396211e-05,0.0001583880155991868,0.00020932206564085407,tokens/kWh,1222995.765956342,MB,1743.720448,22129.672192,0.0,21483.225088,20902.142976,s,10,1661.799328125,166.1799328125,0.024199823300055656,166.1772734375,166.2117125,166.21745,166.22204,"[166.17496875, 166.167359375, 166.14703125, 166.2104375, 166.2231875, 166.181296875, 166.179578125, 166.204078125, 166.159546875, 166.15184375]",tokens/s,0.37910714569299164,kWh,0.0019616428730719624,0.001075153456676562,0.009382404450362024,0.012419200780110549,tokens/kWh,5072.790199261051,,s,629,1684.2762666015624,2.677704716377683,0.33156351074934576,2.63752294921875,2.6395720703125,2.6405790039062498,5.428127246093751,"[2.63648046875, 2.637477783203125, 2.636610595703125, 2.63773388671875, 2.63674365234375, 2.6374482421875, 2.63684716796875, 2.637740966796875, 2.636241943359375, 2.6376162109375, 2.636966796875, 2.6365615234375, 2.639180908203125, 2.637003662109375, 2.637414306640625, 2.638074951171875, 2.638488525390625, 2.637603759765625, 2.6381865234375, 2.6367314453125, 2.63737646484375, 2.636360595703125, 2.63697314453125, 2.6366044921875, 2.6370693359375, 2.636568603515625, 2.636980224609375, 2.636353515625, 2.63706103515625, 2.636314697265625, 2.637646728515625, 2.63864208984375, 2.637822998046875, 2.63769091796875, 2.636580810546875, 2.636631103515625, 2.6366474609375, 2.636380126953125, 2.635778076171875, 2.639795166015625, 2.637285400390625, 2.636735595703125, 2.63680419921875, 2.636505126953125, 2.636515380859375, 2.637959228515625, 2.6366884765625, 2.638927001953125, 2.6396015625, 2.6393056640625, 2.640069580078125, 2.638898193359375, 2.6396181640625, 2.639129638671875, 2.6404248046875, 2.63950439453125, 2.6398291015625, 2.639203369140625, 2.639447021484375, 2.638824462890625, 2.638972900390625, 2.637701171875, 5.43399951171875, 2.63682861328125, 2.637526123046875, 2.637077392578125, 2.636991455078125, 2.636851318359375, 2.63690966796875, 2.636982177734375, 2.638088134765625, 2.63716455078125, 2.63853662109375, 2.638309326171875, 2.63889111328125, 2.64060107421875, 2.637055908203125, 2.63606884765625, 2.63802880859375, 2.637928466796875, 2.6390927734375, 2.6384384765625, 2.637506591796875, 2.63756298828125, 2.63665869140625, 2.636583984375, 2.63687060546875, 2.6370908203125, 2.637446044921875, 2.63802685546875, 2.6375556640625, 2.637516845703125, 2.63621728515625, 2.637552734375, 2.637285400390625, 2.637663330078125, 2.6371767578125, 2.637763671875, 2.63699560546875, 2.63773583984375, 2.637076416015625, 2.637444091796875, 2.636642333984375, 2.636317626953125, 2.63705078125, 2.636695556640625, 2.637413330078125, 2.63615478515625, 2.6434365234375, 2.637322265625, 2.638057373046875, 2.63786181640625, 2.638676025390625, 2.637177734375, 2.63716455078125, 2.637470703125, 2.63781689453125, 2.636496826171875, 2.63722802734375, 2.637602783203125, 2.638507080078125, 2.637663330078125, 2.636621826171875, 2.63727001953125, 2.6375107421875, 5.42925830078125, 2.638592041015625, 2.6379560546875, 2.637390869140625, 2.636768310546875, 2.63889501953125, 2.638235595703125, 2.638288818359375, 2.637874267578125, 2.638148681640625, 2.637728759765625, 2.637327392578125, 2.63690869140625, 2.637347900390625, 2.6369453125, 2.637433837890625, 2.63707958984375, 2.636822509765625, 2.637656982421875, 2.63617529296875, 2.636454833984375, 2.6358466796875, 2.636423095703125, 2.63856640625, 2.63604833984375, 2.63657470703125, 2.637000732421875, 2.637095947265625, 2.636664794921875, 2.6381435546875, 2.640372802734375, 2.638454833984375, 2.636282958984375, 2.636675048828125, 2.635629638671875, 2.636507080078125, 2.63680615234375, 2.63701611328125, 2.636884033203125, 2.63752294921875, 2.6366669921875, 2.63699462890625, 2.63780859375, 2.6374072265625, 2.636739501953125, 2.6374296875, 2.637625244140625, 2.639678466796875, 2.63699658203125, 2.63729052734375, 2.637656982421875, 2.636547119140625, 2.6364794921875, 2.636135498046875, 2.636613525390625, 2.636599365234375, 2.63747265625, 2.6357861328125, 2.63600439453125, 2.6359716796875, 2.636907470703125, 2.636198974609375, 2.639195068359375, 5.42974560546875, 2.637918212890625, 2.638065673828125, 2.6382080078125, 2.639459228515625, 2.638834716796875, 2.640337890625, 2.639097900390625, 2.638284912109375, 2.639035400390625, 2.636672119140625, 2.63703955078125, 2.63772265625, 2.639283203125, 2.6380830078125, 2.639520751953125, 2.63948291015625, 2.637526123046875, 2.638147705078125, 2.637602783203125, 2.63681640625, 2.63762841796875, 2.638529541015625, 2.63696484375, 2.63714599609375, 2.637178955078125, 2.6376796875, 2.63855810546875, 2.641314697265625, 2.637876220703125, 2.63853369140625, 2.63912548828125, 2.637345703125, 2.636560302734375, 2.636971923828125, 2.636669921875, 2.636619873046875, 2.6392392578125, 2.63902099609375, 2.638465087890625, 2.63775341796875, 2.638636962890625, 2.637673583984375, 2.6372373046875, 2.64167431640625, 2.6387353515625, 2.637962158203125, 2.639416259765625, 2.637681640625, 2.639564697265625, 2.6377236328125, 2.638668701171875, 2.638095458984375, 2.63796728515625, 2.637655029296875, 2.637210693359375, 2.63780859375, 2.638158935546875, 2.637846435546875, 2.63988232421875, 2.641334228515625, 2.636745849609375, 2.63619482421875, 5.42609521484375, 2.63847412109375, 2.64098095703125, 2.6405458984375, 2.64087451171875, 2.64148388671875, 2.64074755859375, 2.64026318359375, 2.640604248046875, 2.640962646484375, 2.638622802734375, 2.63895458984375, 2.636958740234375, 2.636030029296875, 2.636072021484375, 2.637117431640625, 2.636577880859375, 2.642241455078125, 2.64078857421875, 2.641033203125, 2.641383544921875, 2.64137841796875, 2.639701904296875, 2.640848876953125, 2.640034912109375, 2.6397451171875, 2.640384033203125, 2.637106201171875, 2.636432373046875, 2.636267578125, 2.637424560546875, 2.636958740234375, 2.63680712890625, 2.637274169921875, 2.638011474609375, 2.63804931640625, 2.63707861328125, 2.636590087890625, 2.63739794921875, 2.63689013671875, 2.63980029296875, 2.639002685546875, 2.638836669921875, 2.63720849609375, 2.637477783203125, 2.637814697265625, 2.637075439453125, 2.637104248046875, 2.638521240234375, 2.63874755859375, 2.63686962890625, 2.63682763671875, 2.637802490234375, 2.63747998046875, 2.6362255859375, 2.63657568359375, 2.636940185546875, 2.642130859375, 2.6372998046875, 2.63684912109375, 2.63803271484375, 2.63796533203125, 2.637073486328125, 5.4244033203125, 2.636745849609375, 2.6364814453125, 2.6371123046875, 2.637408203125, 2.636239990234375, 2.63697607421875, 2.638215087890625, 2.639066162109375, 2.63918994140625, 2.639048583984375, 2.63975634765625, 2.639510498046875, 2.63817626953125, 2.637822021484375, 2.6385029296875, 2.637655029296875, 2.637551513671875, 2.636876708984375, 2.63750048828125, 2.63714599609375, 2.63798779296875, 2.63727099609375, 2.637442138671875, 2.636712890625, 2.637854736328125, 2.638200927734375, 2.6394736328125, 2.63754541015625, 2.636971923828125, 2.637844482421875, 2.63727197265625, 2.6361865234375, 2.636801025390625, 2.637619140625, 2.637947998046875, 2.63809228515625, 2.637560791015625, 2.6417490234375, 2.63729052734375, 2.63712158203125, 2.63689111328125, 2.63716259765625, 2.635864990234375, 2.6373251953125, 2.63663623046875, 2.63742578125, 2.637765625, 2.63775341796875, 2.638487548828125, 2.638257080078125, 2.63924951171875, 2.639295654296875, 2.639500244140625, 2.63644970703125, 2.637189208984375, 2.637624267578125, 2.639413330078125, 2.639233154296875, 2.63777587890625, 2.637401123046875, 2.637765625, 2.63680712890625, 5.42891748046875, 2.637751220703125, 2.6369423828125, 2.637287353515625, 2.6384208984375, 2.6376162109375, 2.637705078125, 2.6376396484375, 2.636984375, 2.636735595703125, 2.637454345703125, 2.63906005859375, 2.641005615234375, 2.639869873046875, 2.638488525390625, 2.638159912109375, 2.638043212890625, 2.6367958984375, 2.6363330078125, 2.637013916015625, 2.63684716796875, 2.63693115234375, 2.636872802734375, 2.64043212890625, 2.636421142578125, 2.63861962890625, 2.63849169921875, 2.638710693359375, 2.637824951171875, 2.6379111328125, 2.636610595703125, 2.6375966796875, 2.637263916015625, 2.63708984375, 2.637641845703125, 2.637759521484375, 2.6372197265625, 2.63781591796875, 2.63773583984375, 2.638950439453125, 2.637347900390625, 2.63666796875, 2.638234619140625, 2.638180419921875, 2.637719482421875, 2.63872216796875, 2.639690673828125, 2.63703662109375, 2.636780517578125, 2.636748779296875, 2.637224853515625, 2.636233642578125, 2.636378173828125, 2.636863525390625, 2.63923193359375, 2.638593994140625, 2.638180419921875, 2.63895556640625, 2.638899169921875, 2.637263916015625, 2.63600537109375, 2.637727783203125, 2.63825, 5.43155078125, 2.637918212890625, 2.63883056640625, 2.63889208984375, 2.63790185546875, 2.637695068359375, 2.636559326171875, 2.636966796875, 2.63714208984375, 2.639033447265625, 2.6404404296875, 2.640530517578125, 2.64198046875, 2.640649169921875, 2.637720458984375, 2.637758544921875, 2.639655029296875, 2.63736328125, 2.63708251953125, 2.636771240234375, 2.636508056640625, 2.63598583984375, 2.6370068359375, 2.636777587890625, 2.636590087890625, 2.63606884765625, 2.63714111328125, 2.63913671875, 2.637537353515625, 2.63705712890625, 2.636992431640625, 2.638615478515625, 2.636080078125, 2.637097900390625, 2.636854248046875, 2.640509033203125, 2.63925146484375, 2.6389052734375, 2.6383544921875, 2.63872705078125, 2.638085205078125, 2.6382275390625, 2.637551513671875, 2.637571044921875, 2.637200439453125, 2.63806982421875, 2.637642822265625, 2.637718505859375, 2.6373642578125, 2.6366044921875, 2.638297119140625, 2.63918798828125, 2.63943896484375, 2.6392646484375, 2.639478759765625, 2.637856689453125, 2.638085205078125, 2.638904296875, 2.6383564453125, 2.639803466796875, 2.640150634765625, 2.637856689453125, 2.640256103515625, 5.4306796875, 2.637127685546875, 2.6382705078125, 2.640280517578125, 2.639151123046875, 2.639075439453125, 2.63817529296875, 2.638035888671875, 2.638784423828125, 2.639287353515625, 2.637106201171875, 2.637173828125, 2.6368388671875, 2.6376591796875, 2.635720703125, 2.63617333984375, 2.636442626953125, 2.63904150390625, 2.6368818359375, 2.635509765625, 2.636674072265625, 2.6366484375, 2.637701171875, 2.638306396484375, 2.64076806640625, 2.63686865234375, 2.6378896484375, 2.639107177734375, 2.6401259765625, 2.6394111328125, 2.63695166015625, 2.63650927734375, 2.636439453125, 2.63674267578125, 2.637043701171875, 2.637275146484375, 2.637382568359375, 2.63693115234375, 2.636669921875, 2.637486083984375, 2.636729248046875, 2.636916748046875, 2.6372802734375, 2.638066650390625, 2.637075439453125, 2.637048828125, 2.636240966796875, 2.637787109375, 2.636506103515625, 2.636660888671875, 2.63712548828125, 2.637421630859375, 2.637157470703125, 2.636517333984375, 2.637552734375, 2.636777587890625, 2.63666064453125, 2.6360966796875, 2.637487060546875, 2.636655517578125, 2.6372158203125, 2.637021240234375, 2.63826025390625, 5.43210595703125, 2.637992919921875, 2.6383369140625, 2.636921875, 2.637557861328125, 2.637305908203125, 2.63912646484375, 2.63752099609375, 2.6370693359375, 2.638551025390625, 2.6377646484375, 2.638067626953125, 2.637068359375, 2.63798876953125, 2.6374892578125, 2.63644775390625, 2.636535888671875, 2.637727783203125, 2.63798876953125, 2.637382568359375, 2.637177734375, 2.637806640625, 2.63628173828125, 2.6376171875, 2.63735302734375, 2.63743896484375, 2.6397236328125, 2.637740966796875, 2.637970458984375, 2.6363720703125, 2.636199951171875, 2.6363935546875, 2.638331787109375, 2.636755859375, 2.636908447265625, 2.636675048828125, 2.63809033203125, 2.636525634765625, 2.63657373046875, 2.63596435546875, 2.63809130859375, 2.6373671875, 2.636396484375, 2.637010009765625, 2.637177734375, 2.636538818359375, 2.637276123046875, 2.638035888671875, 2.636796875, 2.636739501953125, 2.636539794921875, 2.63828076171875, 2.63714404296875, 2.637462646484375, 2.637360107421875, 2.638277587890625, 2.636430419921875, 2.6370283203125, 2.6365234375, 2.638316650390625, 2.636391357421875, 2.636579833984375, 2.636466064453125]",tokens/s,0.3734541728532224,,,main,False,False,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/gemma-2b,google/gemma-2b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/gemma-2b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aa9ca-1d834c9f323b0b817d31ef84;ee948139-9732-450c-a061-6b0b9ca9c9c7) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/gemma-2b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/gemma-2b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/polyglot-ko-12.8b,EleutherAI/polyglot-ko-12.8b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,,,MB,1723.056128,9941.024768,0.0,9294.577664,8910.495232,s,10,10.523069458007814,1.0523069458007812,0.0008028232056597039,1.0523904418945311,1.0530852783203124,1.05345859375,1.05375724609375,"[1.0527669677734375, 1.0538319091796875, 1.05145703125, 1.0517464599609374, 1.052252197265625, 1.0511856689453125, 1.0525286865234376, 1.0530023193359375, 1.0528663330078125, 1.051431884765625]",tokens/s,243.27502638043498,kWh,1.2425229681862726e-05,6.805541535795782e-06,6.044174279779546e-05,7.967251401545396e-05,tokens/kWh,3213153.283331113,MB,1723.056128,9941.024768,0.0,9294.577664,9220.876288,s,10,626.3534921875,62.63534921875,0.004239119732727424,62.635099609375004,62.63956796875,62.641526171875,62.643092734375,"[62.6319921875, 62.6329375, 62.6391328125, 62.643484375, 62.63139453125, 62.6282421875, 62.63577734375, 62.634421875, 62.63843359375, 62.63767578125]",tokens/s,1.0058218048721415,kWh,0.0007394197557369868,0.0004052695339850652,0.003591932262432407,0.004736621552154459,tokens/kWh,13300.619292952455,,s,629,634.8295839233394,1.009268018956025,0.1250286498200473,0.9942149047851563,0.9945757568359375,0.99473427734375,2.0461909814453128,"[0.9937213745117187, 0.9941636962890625, 0.9939281616210938, 0.9936691284179687, 0.9940469970703125, 0.9936640014648438, 0.99386572265625, 0.9940029296875, 0.9938134765625, 0.9938718872070312, 0.9940316162109375, 0.9940879516601563, 0.9947422485351562, 0.9942005615234375, 0.994029541015625, 0.993860595703125, 0.9938984985351562, 0.994188232421875, 0.9941381225585938, 0.9940950927734375, 0.9939527587890625, 0.9939947509765625, 0.99481396484375, 0.994408447265625, 0.994361328125, 0.9940614013671875, 0.99426806640625, 0.9940326538085937, 0.9939323120117187, 0.993944580078125, 0.9940070190429687, 0.994107421875, 0.994229248046875, 0.9940459594726563, 0.9943674926757813, 0.9942251586914063, 0.99422412109375, 0.9942691650390625, 0.9943838500976563, 0.994155517578125, 0.994197509765625, 0.9944873046875, 0.9943777465820313, 0.994418701171875, 0.994387939453125, 0.9942783813476562, 0.9941801147460938, 0.9943582763671875, 0.9945180053710938, 0.9940408325195312, 0.9938134765625, 0.9940326538085937, 0.9944760131835938, 0.9945057373046875, 0.9942804565429687, 0.99428662109375, 0.9941749877929688, 0.994018310546875, 0.9942005615234375, 0.994018310546875, 0.9942138671875, 0.99439306640625, 2.048400390625, 0.9937254638671875, 0.9939384155273437, 0.993649658203125, 0.993681396484375, 0.993491943359375, 0.9935380249023438, 0.993723388671875, 0.9939885864257813, 0.993955810546875, 0.9941115112304687, 0.9941227416992188, 0.9936732177734375, 0.9938882446289062, 0.993818603515625, 0.9939773559570313, 0.9937152099609375, 0.9938954467773438, 0.994029541015625, 0.9940746459960937, 0.9940223999023438, 0.9943040161132812, 0.9939876098632813, 0.9940018920898438, 0.993870849609375, 0.9941780395507812, 0.9940029296875, 0.9939568481445312, 0.9940223999023438, 0.9943408813476563, 0.9943121948242187, 0.9944913940429687, 0.994107421875, 0.9944105224609375, 0.9942425537109375, 0.9945681762695312, 0.9944022827148438, 0.9939404907226562, 0.994323486328125, 0.9944729614257812, 0.9945559692382813, 0.9941841430664062, 0.9939691772460938, 0.9944022827148438, 0.9942015991210937, 0.9941749877929688, 0.996316162109375, 0.994234375, 0.9943941040039063, 0.9945528564453125, 0.9943756713867188, 0.9942384643554687, 0.99406640625, 0.9946685180664062, 0.9946071166992188, 0.9947156372070313, 0.9943460083007812, 0.9942159423828125, 0.9944033203125, 0.994250732421875, 0.994207763671875, 0.9942732543945313, 0.9941258544921875, 2.0457523193359375, 0.9937326049804688, 0.993923095703125, 0.9941954345703125, 0.9940490112304687, 0.9940633544921875, 0.9942056884765625, 0.9942149047851563, 0.9938370361328125, 0.9937274780273437, 0.9940162353515625, 0.9940469970703125, 0.99371826171875, 0.9940367431640625, 0.9938534545898438, 0.9942271728515625, 0.9941575927734375, 0.9941626586914063, 0.9939773559570313, 0.9939947509765625, 0.993728515625, 0.993818603515625, 0.99376025390625, 0.9947381591796876, 0.9940807495117188, 0.9942988891601563, 0.9943746337890625, 0.9940838623046875, 0.9939465942382812, 0.9941636962890625, 0.9940090942382812, 0.99420263671875, 0.9942149047851563, 0.9942138671875, 0.9943306274414062, 0.9944053955078125, 0.9946675415039062, 0.9946388549804688, 0.9943070678710938, 0.9942528076171875, 0.9945692138671876, 0.9944391479492187, 0.9944893188476562, 0.9945906982421875, 0.9944791259765625, 0.99445556640625, 0.9944688720703125, 0.9956966552734375, 0.9946471557617188, 0.9943448486328125, 0.99428662109375, 0.9943275756835938, 0.9942988891601563, 0.9942845458984375, 0.99435107421875, 0.9946736450195313, 0.9949030151367187, 0.99470849609375, 0.9946101684570312, 0.99445556640625, 0.9944708862304688, 0.9945303344726563, 0.9944258422851563, 2.046361572265625, 0.9943121948242187, 0.993902587890625, 0.9938494262695312, 0.9939803466796875, 0.993817626953125, 0.99403369140625, 0.9938892822265625, 0.993890380859375, 0.9941790161132813, 0.9943521118164063, 0.994566162109375, 0.9950842895507812, 0.9945149536132812, 0.9943418579101563, 0.994255859375, 0.9942322998046875, 0.9944514770507813, 0.9944678344726563, 0.9941493530273438, 0.99386572265625, 0.994466796875, 0.994135009765625, 0.9944791259765625, 0.993997802734375, 0.9939465942382812, 0.993850341796875, 0.994050048828125, 0.994281494140625, 0.9941104736328125, 0.9941370849609374, 0.99445556640625, 0.9942169799804688, 0.9945046997070313, 0.9944360961914063, 0.994355224609375, 0.9945057373046875, 0.9943777465820313, 0.9944063720703125, 0.9943849487304688, 0.9944248046875, 0.9944248046875, 0.9944043579101562, 0.9945538330078125, 0.9957181396484375, 0.9942650756835938, 0.9944627075195313, 0.9945119018554688, 0.9939578857421875, 0.9942630615234375, 0.9940695190429687, 0.994323486328125, 0.9945487060546875, 0.9944248046875, 0.9942774047851562, 0.9945712890625, 0.9945681762695312, 0.994572265625, 0.9946071166992188, 0.9948231811523438, 0.994545654296875, 0.9945897216796875, 0.994703369140625, 2.0469268798828124, 0.9937100830078125, 0.99370703125, 0.9939507446289062, 0.9940214233398438, 0.993982421875, 0.9939998779296875, 0.994187255859375, 0.9942363891601562, 0.994165771484375, 0.9937623291015625, 0.9936098022460937, 0.9937017822265625, 0.99392919921875, 0.9935421142578125, 0.9941094360351562, 0.9936434936523437, 0.9947597045898438, 0.9939476318359375, 0.9940572509765625, 0.993876953125, 0.9940459594726563, 0.9939384155273437, 0.9936783447265625, 0.9934899291992187, 0.9937745971679688, 0.9935821533203125, 0.9939537353515625, 0.993924072265625, 0.9939793701171875, 0.9942067260742188, 0.9939005737304687, 0.9939332885742187, 0.9940510864257812, 0.9938903198242187, 0.9940438842773438, 0.9939251098632812, 0.9943121948242187, 0.9945211181640625, 0.9941473388671875, 0.9944022827148438, 0.9940101318359374, 0.9942968139648437, 0.9944330444335937, 0.994639892578125, 0.9945938110351562, 0.9943070678710938, 0.99437158203125, 0.9943593139648438, 0.9942528076171875, 0.9945589599609375, 0.994482177734375, 0.9939876098632813, 0.9945620727539063, 0.9944330444335937, 0.9943951416015625, 0.9943367919921875, 0.9962434692382812, 0.9943162841796875, 0.994186279296875, 0.9945906372070312, 0.9944248657226562, 0.9944125366210937, 2.04558642578125, 0.993912841796875, 0.9936803588867188, 0.9937315673828125, 0.9937705688476562, 0.9936362915039062, 0.9941083984375, 0.9939169311523437, 0.993712158203125, 0.9938135375976562, 0.9937837524414063, 0.9940541381835938, 0.9938093872070313, 0.9940275268554688, 0.9939824829101562, 0.9938954467773438, 0.9938401489257812, 0.9937479858398437, 0.9936363525390625, 0.9938779907226563, 0.9936895751953125, 0.9938565063476562, 0.9942282104492187, 0.9945733032226562, 0.9943101196289063, 0.994271240234375, 0.9942517700195312, 0.9937531127929687, 0.9937540893554687, 0.9940654296875, 0.9936906127929688, 0.9937561645507812, 0.9941299438476563, 0.994255859375, 0.9943173217773438, 0.99435107421875, 0.9943480224609375, 0.9938595581054688, 0.9955552978515625, 0.9940582275390625, 0.994145263671875, 0.9937991943359376, 0.9941483764648438, 0.994471923828125, 0.9942374267578125, 0.9941565551757813, 0.994735107421875, 0.99420263671875, 0.9940562133789063, 0.994302978515625, 0.9941094360351562, 0.9940899658203125, 0.9943009033203125, 0.9945149536132812, 0.9943971557617187, 0.9942517700195312, 0.9943009033203125, 0.9943582763671875, 0.9941473388671875, 0.9944699096679688, 0.9942067260742188, 0.9939999389648437, 0.994462646484375, 2.0474276123046873, 0.9942875366210937, 0.9945169677734375, 0.9941104736328125, 0.99388623046875, 0.99411865234375, 0.9940193481445313, 0.9938565063476562, 0.9939681396484376, 0.9939568481445312, 0.9942251586914063, 0.9939885864257813, 0.9937736206054687, 0.9940674438476562, 0.9934888916015625, 0.9936240844726563, 0.9936025390625, 0.9939957885742188, 0.9941309204101563, 0.9939783935546875, 0.9941585693359375, 0.9935872192382813, 0.9937633056640625, 0.99496142578125, 0.9941268310546875, 0.9944862670898438, 0.993723388671875, 0.9938739013671875, 0.9938882446289062, 0.9944063720703125, 0.9942282104492187, 0.9941043090820313, 0.9945272216796875, 0.9945446166992188, 0.9939815063476563, 0.9940366821289063, 0.9942149047851563, 0.9942435913085937, 0.9943531494140625, 0.9939906616210937, 0.994450439453125, 0.9943357543945313, 0.9943040161132812, 0.9945149536132812, 0.9941370849609374, 0.9942384643554687, 0.99449853515625, 0.9946624145507813, 0.9942548217773437, 0.9945487060546875, 0.9945855712890626, 0.9943418579101563, 0.9944760131835938, 0.9945589599609375, 0.9944545288085938, 0.9943593139648438, 0.9944330444335937, 0.9942958374023437, 0.9943746337890625, 0.9943889770507812, 0.9947330322265625, 0.9947310180664063, 0.9944688720703125, 2.047277099609375, 0.9937192993164062, 0.9936015625, 0.993828857421875, 0.9938534545898438, 0.9940787353515625, 0.9938462524414062, 0.9938677978515625, 0.9937264404296875, 0.9940316162109375, 0.9937715454101562, 0.9938134765625, 0.9937797241210937, 0.9938841552734375, 0.9937469482421875, 0.9938042602539062, 0.9948897094726562, 0.99428759765625, 0.99430810546875, 0.9943807983398437, 0.9940869140625, 0.994150390625, 0.9942937622070313, 0.994207763671875, 0.9942947998046875, 0.9942835083007813, 0.9943101196289063, 0.994366455078125, 0.9940070190429687, 0.9944422607421874, 0.9940172729492187, 0.9939844970703124, 0.9939251098632812, 0.9939476318359375, 0.9940101318359374, 0.9939158935546875, 0.9941954345703125, 0.9944647827148437, 0.994376708984375, 0.9946132202148438, 0.9943787231445312, 0.9942466430664062, 0.994313232421875, 0.9944729614257812, 0.9945261840820312, 0.9942916870117188, 0.9944391479492187, 0.9942671508789063, 0.994545654296875, 0.994798583984375, 0.9944873046875, 0.9943193359375, 0.994164794921875, 0.9943551635742187, 0.9943255004882813, 0.99406640625, 0.9942916870117188, 0.994587646484375, 0.9942916870117188, 0.9949767456054688, 0.9941903076171875, 0.9943889770507812, 0.9943162841796875, 2.047072265625, 0.9936138305664063, 0.9937510375976563, 0.9939783935546875, 0.9938709106445313, 0.9944656982421874, 0.993997802734375, 0.99403369140625, 0.993713134765625, 0.9936732177734375, 0.9935554809570313, 0.9939323120117187, 0.9941073608398437, 0.994260986328125, 0.9937418212890625, 0.9940961303710938, 0.9939844970703124, 0.9946102294921875, 0.993966064453125, 0.9940264892578125, 0.9938206787109375, 0.9939199829101563, 0.9943121948242187, 0.9942937622070313, 0.9941575927734375, 0.9946552124023438, 0.9939906616210937, 0.9938114624023437, 0.9936793823242187, 0.9941647338867188, 0.9941442260742187, 0.994471923828125, 0.9945323486328125, 0.99468798828125, 0.9943142700195312, 0.994466796875, 0.9942467041015625, 0.9940725708007813, 0.9942282104492187, 0.994255859375, 0.9944166259765626, 0.9944391479492187, 0.9945436401367187, 0.9948764038085938, 0.99462451171875, 0.9944658203125, 0.994460693359375, 0.994155517578125, 0.9942097778320312, 0.9942763671875, 0.9942702026367187, 0.994260986328125, 0.9944309692382812, 0.9945098266601563, 0.9942968139648437, 0.9946491088867188, 0.9945579223632812, 0.99460302734375, 0.994281494140625, 0.9945108642578125, 0.99460302734375, 0.9943797607421875, 0.9946961669921875, 2.04813623046875, 0.99420263671875, 0.99418115234375, 0.9938411254882813, 0.9938524169921875, 0.9939671020507812, 0.9941248168945312, 0.994044921875, 0.9940910034179687, 0.9938984985351562, 0.9937213745117187, 0.9939066772460937, 0.9943142700195312, 0.9939824829101562, 0.9937520751953125, 0.9941227416992188, 0.9939323120117187, 0.9941923828125, 0.9938759765625, 0.9943316650390625, 0.994208740234375, 0.99416064453125, 0.9944412231445312, 0.99399169921875, 0.9939773559570313, 0.9942149658203125, 0.9955389404296875, 0.9943828735351562, 0.9939456176757813, 0.9944596557617188, 0.9941442260742187, 0.9940377807617188, 0.9945589599609375, 0.994135009765625, 0.9938985595703125, 0.9942609252929687, 0.994212890625, 0.9942732543945313, 0.9943726196289062, 0.994460693359375, 0.9944248657226562, 0.9942844848632812, 0.9949808349609375, 0.9943121948242187, 0.9940910034179687, 0.994150390625, 0.9942640380859376, 0.9941688232421875, 0.9940387573242188, 0.9946736450195313, 0.994753662109375, 0.99428857421875, 0.994555908203125, 0.9943367919921875, 0.9942702026367187, 0.994302978515625, 0.994234375, 0.9944371337890625, 0.9942466430664062, 0.9943602905273438, 0.9947944946289062, 0.99437158203125, 0.9947586669921875]",tokens/s,0.9908170884423626,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-125m,,cuda,0,42,,,,,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,,MB,1273.409536,921.174016,0.0,274.726912,220.646912,s,10,0.326115104675293,0.032611510467529296,0.0012484582231843272,0.03218409538269043,0.032784641265869136,0.034564048767089836,0.035987574768066403,"[0.036343456268310544, 0.03226867294311524, 0.03213779067993164, 0.032141822814941406, 0.03203395080566406, 0.03214742279052735, 0.032220767974853515, 0.03238921737670898, 0.03208832168579102, 0.032343681335449216]",tokens/s,7849.989047728858,kWh,3.834838485734976e-07,2.1012961289151069e-07,8.25654520552228e-07,1.419267982017236e-06,tokens/kWh,180374674.2994524,MB,1273.409536,921.174016,0.0,274.726912,250.722304,s,10,20.085640136718748,2.008564013671875,0.006065788108395492,2.006720520019531,2.0181213745117184,2.0185785705566404,2.018944327392578,"[2.018019775390625, 2.006443115234375, 2.0069979248046876, 2.0103148193359375, 2.0040799560546874, 2.00180322265625, 2.00428125, 2.0133179931640623, 2.0013463134765623, 2.0190357666015624]",tokens/s,31.36569189290069,kWh,2.3731767884623614e-05,1.3005590929520364e-05,4.874369069043466e-05,8.548104950457861e-05,tokens/kWh,737005.457526882,,s,629,20.34066634368894,0.03233810229521298,0.0038819833531807996,0.03179007911682129,0.0323544822692871,0.032856268310546874,0.06427897552490235,"[0.03209625625610352, 0.031882240295410154, 0.031893503189086916, 0.032045055389404296, 0.03220787048339844, 0.031922176361083986, 0.031958015441894534, 0.03196518325805664, 0.031784959793090824, 0.03223756790161133, 0.031974399566650394, 0.031955968856811526, 0.03202969741821289, 0.03302809524536133, 0.03192835235595703, 0.031831008911132816, 0.03254272079467774, 0.033226783752441404, 0.032270305633544924, 0.032140289306640625, 0.031940607070922854, 0.032451583862304685, 0.03203788757324219, 0.03188531112670898, 0.032231422424316404, 0.031936511993408204, 0.03298099136352539, 0.03213516616821289, 0.03167948722839355, 0.031526912689208986, 0.031693824768066405, 0.03172966384887695, 0.03170918464660644, 0.03180953598022461, 0.03179929542541504, 0.031628288269042966, 0.031508480072021484, 0.03155660820007324, 0.03158937644958496, 0.03198259162902832, 0.03151974487304687, 0.03146342468261719, 0.031752191543579104, 0.03203481674194336, 0.03182796859741211, 0.03228160095214844, 0.03201331329345703, 0.03220377731323242, 0.03181056022644043, 0.031865856170654294, 0.031473663330078124, 0.031682559967041016, 0.031628288269042966, 0.03277414321899414, 0.03190790367126465, 0.0320777587890625, 0.03253350448608398, 0.032473087310791016, 0.032159744262695314, 0.03166924858093262, 0.03159654426574707, 0.031492095947265625, 0.06451507568359376, 0.032, 0.03342745590209961, 0.032282623291015625, 0.03176243209838867, 0.0318525447845459, 0.03180646324157715, 0.03143065643310547, 0.03154841613769531, 0.03168870353698731, 0.031514623641967776, 0.03150643157958984, 0.03214438247680664, 0.033160190582275394, 0.03213312149047851, 0.03156684875488281, 0.031958015441894534, 0.03179110336303711, 0.03184332847595215, 0.03142963218688965, 0.03191705513000488, 0.031768575668334964, 0.03164057540893555, 0.0317573127746582, 0.03179827117919922, 0.031731712341308595, 0.03301375961303711, 0.031920127868652344, 0.03172352027893066, 0.03289395141601562, 0.031927295684814457, 0.03162521553039551, 0.03168358421325684, 0.03179110336303711, 0.031666175842285156, 0.03177369689941406, 0.031678464889526366, 0.03179315185546875, 0.031765504837036135, 0.031699968338012696, 0.03142860794067383, 0.03162419128417969, 0.031615999221801756, 0.03140505599975586, 0.03147776031494141, 0.031458303451538085, 0.03143987274169922, 0.03174399948120117, 0.0317255687713623, 0.03142758369445801, 0.03156991958618164, 0.031783935546875, 0.03183616065979004, 0.0317890567779541, 0.03189248085021973, 0.03191910362243652, 0.031932416915893554, 0.032277503967285154, 0.032382976531982424, 0.0317706241607666, 0.03181260871887207, 0.031749120712280275, 0.03174502372741699, 0.06404198455810547, 0.03170611190795898, 0.031704063415527346, 0.031628288269042966, 0.03138764762878418, 0.031698944091796875, 0.03154022407531738, 0.03150540733337402, 0.03196416091918945, 0.03177881622314453, 0.032020481109619144, 0.03214233779907227, 0.03199692726135254, 0.031780864715576174, 0.03226009750366211, 0.032115711212158206, 0.03199180793762207, 0.03156787109375, 0.031453184127807614, 0.03144908714294434, 0.0314839038848877, 0.03146649551391602, 0.031458303451538085, 0.03172147178649903, 0.031893503189086916, 0.0328007698059082, 0.03156787109375, 0.031731712341308595, 0.031849472045898435, 0.03165388870239258, 0.03178700828552246, 0.03168259239196777, 0.03179311943054199, 0.031783935546875, 0.03212083053588867, 0.03178188705444336, 0.03179724884033203, 0.033549312591552735, 0.03284889602661133, 0.031834112167358396, 0.03212492752075195, 0.03198259162902832, 0.03174297523498535, 0.031731712341308595, 0.031753215789794925, 0.03182387161254883, 0.031473663330078124, 0.03165388870239258, 0.03157196807861328, 0.031453184127807614, 0.031734783172607424, 0.03165695953369141, 0.03180953598022461, 0.03184537506103516, 0.03207372665405273, 0.032285694122314454, 0.032556095123291016, 0.0321484146118164, 0.0324956169128418, 0.03182592010498047, 0.03148287963867188, 0.03179110336303711, 0.03181056022644043, 0.06443007659912109, 0.03183001518249512, 0.03196928024291992, 0.03182694435119629, 0.03170099258422852, 0.03181056022644043, 0.0315043830871582, 0.03177881622314453, 0.03177574348449707, 0.03191910362243652, 0.0321710090637207, 0.03204403305053711, 0.03181875228881836, 0.031703039169311525, 0.03174604797363281, 0.03175526428222656, 0.031764480590820314, 0.031731712341308595, 0.03183923149108887, 0.03340390396118164, 0.032922622680664065, 0.032161792755126956, 0.03223756790161133, 0.031734783172607424, 0.03291340637207031, 0.032074752807617186, 0.03202150344848633, 0.031716352462768556, 0.03155046463012695, 0.031692800521850584, 0.03226828765869141, 0.03171737670898438, 0.03140505599975586, 0.03134976005554199, 0.03197952079772949, 0.03200307083129883, 0.031696895599365234, 0.03169484710693359, 0.03176755142211914, 0.0321710090637207, 0.03153919982910156, 0.03143987274169922, 0.03146444892883301, 0.03160371208190918, 0.03245568084716797, 0.0314337272644043, 0.03163545608520508, 0.03231846237182617, 0.03222323226928711, 0.031942655563354495, 0.03174399948120117, 0.03162112045288086, 0.03139788818359375, 0.032302078247070314, 0.03218841552734375, 0.03275161743164062, 0.0322529296875, 0.03198361587524414, 0.03178803253173828, 0.031665151596069335, 0.03129241561889649, 0.03186073684692383, 0.031955968856811526, 0.06424063873291015, 0.03157811164855957, 0.03165695953369141, 0.03165081596374512, 0.03146854400634766, 0.031693824768066405, 0.03204095840454101, 0.03246182250976563, 0.03183103942871094, 0.03218636703491211, 0.03204095840454101, 0.031784959793090824, 0.03218431854248047, 0.03177779197692871, 0.031698944091796875, 0.03156684875488281, 0.031731712341308595, 0.03239424133300781, 0.03142758369445801, 0.03172966384887695, 0.03203379058837891, 0.03207167816162109, 0.03187302398681641, 0.033006591796875, 0.03175526428222656, 0.0316180477142334, 0.031959039688110355, 0.031514623641967776, 0.031884288787841795, 0.03161497688293457, 0.031440895080566404, 0.031442943572998046, 0.03174092864990234, 0.03172966384887695, 0.03197747230529785, 0.03176755142211914, 0.03205222320556641, 0.03165695953369141, 0.03194777679443359, 0.03171327972412109, 0.031562751770019534, 0.03168972778320313, 0.031731712341308595, 0.03172966384887695, 0.03186483192443847, 0.032143360137939454, 0.03146240043640137, 0.03147776031494141, 0.03138355255126953, 0.0313753604888916, 0.03180646324157715, 0.0317255687713623, 0.03186380767822266, 0.03160678482055664, 0.032494590759277346, 0.03222528076171875, 0.032045055389404296, 0.03183206367492676, 0.03183206367492676, 0.03180851173400879, 0.03134464073181152, 0.031470592498779294, 0.031731712341308595, 0.06429388427734375, 0.03167436790466309, 0.031923200607299806, 0.03159756851196289, 0.03151667213439941, 0.031647743225097655, 0.031802400588989255, 0.03235427093505859, 0.03198259162902832, 0.03172147178649903, 0.031888383865356446, 0.03179417610168457, 0.031507455825805664, 0.03172761535644531, 0.031579135894775394, 0.031692800521850584, 0.0315156478881836, 0.031440895080566404, 0.03173990440368652, 0.03155356788635254, 0.031624160766601565, 0.03153510475158691, 0.033051647186279294, 0.03177369689941406, 0.031783935546875, 0.03135897636413574, 0.03179007911682129, 0.031714303970336914, 0.031848447799682614, 0.031476736068725586, 0.03139993667602539, 0.03133440017700195, 0.03155763244628906, 0.03146956825256347, 0.03141024017333984, 0.031344575881958006, 0.03177574348449707, 0.03194675254821777, 0.03219046401977539, 0.03170918464660644, 0.031678464889526366, 0.03172249603271484, 0.03172147178649903, 0.031513599395751955, 0.03161497688293457, 0.03177779197692871, 0.03177574348449707, 0.03137228775024414, 0.03147369575500488, 0.031712223052978515, 0.03170918464660644, 0.03205222320556641, 0.03178803253173828, 0.03221299362182617, 0.03220889663696289, 0.03131391906738281, 0.03146956825256347, 0.03166720008850098, 0.031661056518554685, 0.03207376098632812, 0.03320111846923828, 0.032717823028564456, 0.032519168853759765, 0.06434508514404297, 0.031736831665039066, 0.031752191543579104, 0.031851520538330076, 0.03142963218688965, 0.03157196807861328, 0.032121856689453124, 0.03311513519287109, 0.03222534561157227, 0.03178079986572266, 0.03179827117919922, 0.03183206367492676, 0.03181465530395508, 0.03183206367492676, 0.03177267265319824, 0.03184025573730469, 0.031542272567749025, 0.03150540733337402, 0.03178598403930664, 0.03185663986206055, 0.03176038360595703, 0.03312639999389649, 0.03152179145812988, 0.032102401733398435, 0.03188019180297851, 0.03193036842346191, 0.031324159622192385, 0.031851520538330076, 0.03158016014099121, 0.031389696121215824, 0.03189555168151856, 0.03170099258422852, 0.03173785591125488, 0.031285247802734374, 0.03177068710327149, 0.031356864929199216, 0.03158527946472168, 0.031526912689208986, 0.03160985565185547, 0.03185868835449219, 0.031848447799682614, 0.031437824249267575, 0.0317388801574707, 0.03174297523498535, 0.0319866886138916, 0.03175014305114746, 0.031922176361083986, 0.031817728042602536, 0.03180544090270996, 0.03139072036743164, 0.03181260871887207, 0.03197952079772949, 0.03224166488647461, 0.03209011077880859, 0.03164057540893555, 0.03165184020996094, 0.031438848495483396, 0.03207372665405273, 0.03194675254821777, 0.03155148887634277, 0.03187302398681641, 0.03263488006591797, 0.03184025573730469, 0.06429801940917969, 0.03187299156188965, 0.03203276824951172, 0.032094207763671875, 0.03197849655151367, 0.031958015441894534, 0.031926271438598636, 0.031834112167358396, 0.03189452743530274, 0.03218841552734375, 0.032353279113769534, 0.0318791675567627, 0.03179315185546875, 0.031887359619140625, 0.03185868835449219, 0.03156480026245117, 0.03182796859741211, 0.0322426872253418, 0.032043006896972655, 0.03181670379638672, 0.03299020767211914, 0.032031742095947266, 0.03186892890930176, 0.032304126739501955, 0.03237171173095703, 0.0319866886138916, 0.031454208374023435, 0.031525888442993165, 0.03151155281066895, 0.031492095947265625, 0.03167334365844727, 0.031730688095092774, 0.03897753524780274, 0.032933887481689454, 0.031889408111572266, 0.03211980819702148, 0.031719423294067385, 0.031352832794189454, 0.03140505599975586, 0.031687711715698244, 0.031402975082397463, 0.03169177627563476, 0.031749120712280275, 0.03167948722839355, 0.03165695953369141, 0.0316753921508789, 0.03172659111022949, 0.03139072036743164, 0.03140812873840332, 0.03175526428222656, 0.03156684875488281, 0.03263590240478516, 0.03202252960205078, 0.03183923149108887, 0.0319682559967041, 0.031649791717529296, 0.031784959793090824, 0.03172352027893066, 0.03167129516601563, 0.0314337272644043, 0.03163443183898926, 0.03186483192443847, 0.03146137619018555, 0.06431948852539063, 0.03137843132019043, 0.03158016014099121, 0.031425535202026365, 0.031645696640014646, 0.031730688095092774, 0.03186486434936524, 0.031797216415405276, 0.0318156795501709, 0.03175628852844238, 0.031699968338012696, 0.031410175323486327, 0.03138355255126953, 0.03278540802001953, 0.03300454330444336, 0.032178176879882815, 0.0318156795501709, 0.031855615615844726, 0.031870975494384765, 0.03283967971801758, 0.0314204158782959, 0.0314081916809082, 0.031652799606323244, 0.03211775970458984, 0.031666175842285156, 0.03154022407531738, 0.031373311996459964, 0.03173785591125488, 0.031546367645263675, 0.0316753921508789, 0.0315043830871582, 0.031529983520507815, 0.031764480590820314, 0.03139481544494629, 0.03133440017700195, 0.03138559913635254, 0.03137843132019043, 0.031735807418823245, 0.031543296813964845, 0.03143065643310547, 0.03159859275817871, 0.031936511993408204, 0.031735807418823245, 0.03193548774719238, 0.031458303451538085, 0.031456256866455076, 0.031716352462768556, 0.0317388801574707, 0.032037952423095706, 0.031414207458496095, 0.0326901741027832, 0.03216793441772461, 0.031585344314575194, 0.03156883239746094, 0.03191398429870605, 0.031631359100341795, 0.031903743743896484, 0.03156076812744141, 0.03184736061096191, 0.03181363105773926, 0.03224371337890625, 0.03225702285766602, 0.031955968856811526, 0.06444134521484375, 0.03154431915283203, 0.031665151596069335, 0.03288780975341797, 0.031886335372924804, 0.03204403305053711, 0.03172659111022949, 0.031509504318237305, 0.03173785591125488, 0.032, 0.032320510864257815, 0.03182387161254883, 0.03230003356933594, 0.03187404823303223, 0.032198688507080075, 0.03247203063964844, 0.03235532760620117, 0.032702465057373044, 0.03311718368530273, 0.03155148887634277, 0.03170201683044434, 0.03225094223022461, 0.03220576095581055, 0.03237580871582031, 0.032230400085449216, 0.03218431854248047, 0.03240447998046875, 0.032449535369873043, 0.03225600051879883, 0.03243622589111328, 0.0319498233795166, 0.03165081596374512, 0.031528959274291994, 0.03161702346801758, 0.0321003532409668, 0.03190784072875977, 0.03203788757324219, 0.032328704833984374, 0.032467967987060545, 0.03184435272216797, 0.031665151596069335, 0.03177369689941406, 0.03174604797363281, 0.03233894348144531, 0.031704063415527346, 0.03172454452514648, 0.032039936065673826, 0.03196211242675781, 0.031926271438598636, 0.0328611831665039, 0.03221200180053711, 0.03217712020874024, 0.03181056022644043, 0.03184332847595215, 0.032064510345458985, 0.03207372665405273, 0.03184339141845703, 0.03191596794128418, 0.031817760467529294, 0.031796192169189455, 0.03153408050537109, 0.032161792755126956, 0.03232460784912109]",tokens/s,30.92327406447812,,,main,False,False,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-13b,facebook/opt-13b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,3081.781248,9521.594368,0.0,8875.147264,8264.141824,s,10,10.63840625,1.0638406249999999,0.0011581823391520636,1.0637650756835937,1.0652460571289062,1.0654749572753905,1.0656580773925781,"[1.0649481201171875, 1.065703857421875, 1.0634771728515624, 1.0627923583984376, 1.063096435546875, 1.0621585693359374, 1.064052978515625, 1.0644849853515626, 1.0651951904296875, 1.06249658203125]",tokens/s,240.63754850497463,kWh,1.2549749844604067e-05,6.874677577252441e-06,5.877399146358986e-05,7.819841888544636e-05,tokens/kWh,3273723.479946787,MB,3081.781248,9521.594368,0.0,8875.147264,8556.643328,s,10,631.86266796875,63.186266796874996,0.007118498406389631,63.1850546875,63.19375234375,63.198333203125,63.201997890625,"[63.192734375, 63.2029140625, 63.18565625, 63.17730078125, 63.18101953125, 63.1805546875, 63.18579296875, 63.18146875, 63.1907734375, 63.184453125]",tokens/s,0.9970520999844192,kWh,0.0007460630791551536,0.0004089103833046283,0.003496975492022614,0.004651948954482396,tokens/kWh,13542.710940389017,,s,629,640.4335947265614,1.0181774160994632,0.12635276340148643,1.0029219970703125,1.0034655395507812,1.0036467651367187,2.0656750390625,"[1.00284619140625, 1.002883056640625, 1.0025042114257812, 1.0028574829101562, 1.0028175659179688, 1.002925048828125, 1.0030203247070313, 1.0030990600585938, 1.0027622680664063, 1.0032742309570313, 1.0026455078125, 1.003062255859375, 1.0025420532226563, 1.0029680786132813, 1.0027857666015625, 1.00307763671875, 1.0029732055664062, 1.0030858154296876, 1.002630126953125, 1.0030980834960936, 1.0028267822265624, 1.003293701171875, 1.0026495971679688, 1.0032005004882814, 1.0028339233398438, 1.0035537719726562, 1.0033336181640624, 1.0031492919921876, 1.0029854736328125, 1.0029179077148438, 1.0027100219726564, 1.0030469360351562, 1.0027202758789062, 1.0029025268554688, 1.002692626953125, 1.003177978515625, 1.0032864990234376, 1.0032291870117187, 1.0031237182617188, 1.0033694458007814, 1.0031052856445313, 1.0034606323242188, 1.0029014892578125, 1.0031063232421875, 1.002977294921875, 1.0030796508789062, 1.0032691040039063, 1.003052001953125, 1.003093994140625, 1.0031728515625, 1.0026219482421874, 1.003230224609375, 1.0031892700195313, 1.003472900390625, 1.0031318969726561, 1.0034933471679688, 1.0038251342773437, 1.0035701904296874, 1.0035916748046876, 1.0035220336914064, 1.0029598999023437, 1.0031124267578124, 2.068518798828125, 1.0033325805664062, 1.0031472778320312, 1.0031912841796875, 1.0029844360351563, 1.0031206665039063, 1.0034401245117188, 1.0029373168945312, 1.0032005004882814, 1.00276123046875, 1.003198486328125, 1.0033233642578125, 1.0032588500976563, 1.0031349487304688, 1.0031708374023438, 1.0031646728515624, 1.0034074096679688, 1.0032445068359375, 1.0034769897460938, 1.0028963623046876, 1.0031943969726562, 1.0029393920898437, 1.0032639770507812, 1.0037217407226562, 1.0034319458007812, 1.0026045532226562, 1.003093017578125, 1.0026946411132813, 1.0035282592773438, 1.0035670166015624, 1.0035159301757812, 1.0033766479492188, 1.0034708251953126, 1.0032630004882812, 1.0034298706054687, 1.002693603515625, 1.0028206176757812, 1.0027888793945312, 1.0029660034179688, 1.0028257446289062, 1.0030294799804687, 1.00284619140625, 1.003236328125, 1.0038517456054687, 1.0034739379882813, 1.0031769409179687, 1.003452392578125, 1.0032343139648439, 1.0034002075195312, 1.0033592529296875, 1.0034892578125, 1.003430908203125, 1.0036060180664061, 1.0037442626953126, 1.0036817626953125, 1.003325439453125, 1.0032271118164062, 1.0029475708007813, 1.0032077026367188, 1.0028257446289062, 1.0033950805664062, 1.0032373657226563, 1.0033449096679687, 2.06620166015625, 1.003240478515625, 1.0030346069335938, 1.0028810424804688, 1.0028892211914062, 1.0025840454101562, 1.0026762084960938, 1.0028451538085938, 1.0030079956054687, 1.0030458984375, 1.0029598999023437, 1.0025256958007813, 1.002809326171875, 1.0025021362304687, 1.002545166015625, 1.002588134765625, 1.0025830688476562, 1.0028103637695311, 1.002876953125, 1.0027571411132812, 1.0027847900390625, 1.0027694091796875, 1.0029219970703125, 1.0027468872070313, 1.0030151977539064, 1.002692626953125, 1.0029998168945313, 1.0028185424804688, 1.002977294921875, 1.0030653686523439, 1.0028308715820313, 1.0027254028320312, 1.0027396850585937, 1.002481689453125, 1.0026741943359374, 1.002397705078125, 1.0026148071289063, 1.0031001586914063, 1.003051025390625, 1.0032445678710937, 1.0031226806640625, 1.00276123046875, 1.0029619140625, 1.0027110595703126, 1.00299365234375, 1.0028328857421875, 1.0028206176757812, 1.0037903442382812, 1.0038609619140626, 1.0040064086914062, 1.0035978393554688, 1.0031810302734374, 1.0034503784179687, 1.0034176025390624, 1.0035722045898436, 1.0029711303710938, 1.0032793579101562, 1.003167724609375, 1.0029445190429687, 1.0031759643554687, 1.0028656616210938, 1.002893310546875, 1.0028052368164062, 2.0654345703125, 1.0026424560546876, 1.0026322021484375, 1.0025789184570313, 1.0027899169921874, 1.0024918823242188, 1.0032875366210938, 1.002587158203125, 1.002587158203125, 1.0025154418945312, 1.0026045532226562, 1.0023638916015625, 1.002629150390625, 1.0024959716796875, 1.0028257446289062, 1.0028626098632814, 1.0027550659179687, 1.0028287963867188, 1.0026536865234374, 1.0024857788085937, 1.0027151489257813, 1.0024734497070313, 1.0031738891601563, 1.0028011474609375, 1.003345947265625, 1.002503173828125, 1.0029957275390624, 1.0027591552734374, 1.003062255859375, 1.003240478515625, 1.0029967651367186, 1.002982421875, 1.0035916748046876, 1.0024970092773438, 1.0031472778320312, 1.0028124389648438, 1.0027591552734374, 1.0026690673828125, 1.0027110595703126, 1.0026762084960938, 1.0028493041992188, 1.0025471801757813, 1.0032220458984376, 1.0027745361328124, 1.0030745849609375, 1.0030172119140626, 1.0028216552734375, 1.002661865234375, 1.0027734985351562, 1.00288818359375, 1.0029906005859375, 1.00273046875, 1.0033080444335938, 1.0029434814453124, 1.0028789672851564, 1.0029700927734375, 1.0029117431640624, 1.0027509765625, 1.002767333984375, 1.0027171630859375, 1.0029578247070312, 1.002545166015625, 1.003062255859375, 2.0657685546875, 1.0028359375, 1.0028124389648438, 1.0024847412109374, 1.0028124389648438, 1.0025441284179688, 1.0027601928710939, 1.0023495483398437, 1.0028533935546875, 1.0025604858398438, 1.0030325927734376, 1.0025574340820314, 1.0029946899414062, 1.0025062255859376, 1.0027683715820312, 1.002640380859375, 1.0027284545898438, 1.0027683715820312, 1.0030612182617187, 1.002598388671875, 1.0030366821289063, 1.0028635864257813, 1.0032056274414063, 1.0027100219726564, 1.0028585205078124, 1.0025328369140625, 1.0029660034179688, 1.0023075561523438, 1.0032752685546875, 1.002946533203125, 1.0032271118164062, 1.0027387084960937, 1.0030264282226562, 1.0025748291015626, 1.0029188842773438, 1.0027284545898438, 1.002767333984375, 1.0028052368164062, 1.0030530395507813, 1.0026967163085938, 1.0031349487304688, 1.0030233764648437, 1.0033059692382813, 1.002514404296875, 1.0027438354492189, 1.0027816772460938, 1.0031032104492188, 1.0027970581054688, 1.0033837890625, 1.0027315063476563, 1.0034063110351563, 1.0029496459960938, 1.0030530395507813, 1.0026383056640624, 1.00305615234375, 1.0027919311523437, 1.0027427978515624, 1.0032557983398438, 1.0037903442382812, 1.0031646728515624, 1.003399169921875, 1.0027448120117188, 1.0029649658203126, 2.064649169921875, 1.0027888793945312, 1.0026015014648437, 1.0022328491210937, 1.0030059814453125, 1.0026045532226562, 1.0029281005859374, 1.0032117919921875, 1.0038660888671875, 1.0030458984375, 1.0028523559570313, 1.00279296875, 1.0029219970703125, 1.00295068359375, 1.0030786743164062, 1.00269775390625, 1.0029230346679687, 1.0029240112304687, 1.0026373291015624, 1.0023751220703125, 1.0028328857421875, 1.0023956298828125, 1.002513427734375, 1.002144775390625, 1.003087890625, 1.0024324951171875, 1.0031769409179687, 1.0026843872070312, 1.0027018432617187, 1.0026710815429687, 1.0026076049804689, 1.0026281127929688, 1.00263525390625, 1.0025758666992188, 1.00335107421875, 1.0025952758789063, 1.0027919311523437, 1.0030172119140626, 1.0029312133789063, 1.0026270751953126, 1.0026127319335938, 1.0024099731445313, 1.0028635864257813, 1.0025287475585938, 1.003124755859375, 1.0027868041992187, 1.0029752197265625, 1.0032691040039063, 1.0028994750976563, 1.0029813842773438, 1.0031022338867188, 1.0025799560546875, 1.0030069580078125, 1.0030786743164062, 1.0034902954101563, 1.00322509765625, 1.0034298706054687, 1.0037340087890625, 1.0030960693359374, 1.00282470703125, 1.0031349487304688, 1.0028359375, 1.0032772827148437, 2.06746826171875, 1.0032711791992188, 1.00282470703125, 1.0029578247070312, 1.0026793212890626, 1.0024990844726562, 1.0027325439453125, 1.0022778930664062, 1.0023464965820312, 1.0023444213867188, 1.0030346069335938, 1.0042449951171875, 1.0037616577148438, 1.0032691040039063, 1.0031185913085938, 1.0024017944335937, 1.0026875, 1.0032916259765625, 1.0031452026367187, 1.0032855224609376, 1.0035599365234376, 1.0035138549804687, 1.0038927612304687, 1.0032435302734375, 1.002841064453125, 1.0024775390625, 1.0027438354492189, 1.0022512817382812, 1.002450927734375, 1.00248779296875, 1.002692626953125, 1.0029946899414062, 1.0030489501953126, 1.0032496337890624, 1.0034002075195312, 1.003052001953125, 1.0032271118164062, 1.0028707885742187, 1.0030796508789062, 1.0030028686523438, 1.0040872802734375, 1.00299365234375, 1.0028626098632814, 1.00265576171875, 1.002756103515625, 1.002808349609375, 1.0027335815429688, 1.0026813354492188, 1.00265673828125, 1.002919921875, 1.0032476196289062, 1.0029752197265625, 1.0029025268554688, 1.00297216796875, 1.0027991333007813, 1.0026178588867187, 1.002767333984375, 1.0026639404296875, 1.0027540283203125, 1.0026741943359374, 1.0029127807617189, 1.002988525390625, 1.0027018432617187, 2.066872314453125, 1.0026639404296875, 1.00259228515625, 1.0022307739257812, 1.0025277709960938, 1.0030377197265625, 1.0026045532226562, 1.002482666015625, 1.0029168701171876, 1.002771484375, 1.0028124389648438, 1.0024642333984375, 1.0027807006835938, 1.0025420532226563, 1.0025379638671874, 1.002556396484375, 1.002534912109375, 1.0028687133789063, 1.002840087890625, 1.0028328857421875, 1.0028451538085938, 1.0025236206054688, 1.0027632446289063, 1.0025379638671874, 1.0027888793945312, 1.0026751708984376, 1.003261962890625, 1.0027161865234375, 1.0028973999023438, 1.0030325927734376, 1.00279296875, 1.0027479248046876, 1.0025471801757813, 1.0023690185546874, 1.0029291381835939, 1.0029486083984376, 1.0027970581054688, 1.0028840942382813, 1.0029475708007813, 1.0030294799804687, 1.0031657104492187, 1.0029014892578125, 1.003304931640625, 1.0028472290039063, 1.0032578735351563, 1.0034647216796875, 1.0028277587890626, 1.0030632934570312, 1.0030980834960936, 1.00314111328125, 1.0028748779296874, 1.002660888671875, 1.0028431396484374, 1.0028472290039063, 1.0029691162109375, 1.0030172119140626, 1.002925048828125, 1.0033325805664062, 1.0035189819335937, 1.0036541137695312, 1.004179443359375, 1.0036951293945313, 1.002919921875, 2.06862841796875, 1.0036357421875, 1.0038538208007812, 1.0032230224609375, 1.0033796997070312, 1.002618896484375, 1.002708984375, 1.0022072143554688, 1.0026813354492188, 1.0022891235351563, 1.002555419921875, 1.0025963745117188, 1.00295166015625, 1.0026690673828125, 1.002914794921875, 1.003019287109375, 1.0031646728515624, 1.002956787109375, 1.0028861694335938, 1.0026751708984376, 1.003303955078125, 1.0028124389648438, 1.0030386962890625, 1.0028052368164062, 1.00309912109375, 1.0027315063476563, 1.0034688110351562, 1.0020556640625, 1.003325439453125, 1.0033530883789064, 1.003283447265625, 1.0034749145507813, 1.0035108032226563, 1.0035435791015626, 1.0037903442382812, 1.0029168701171876, 1.0031943969726562, 1.0024591064453126, 1.0026526489257812, 1.0026332397460938, 1.0027837524414063, 1.0034381103515626, 1.0032527465820313, 1.0027008056640625, 1.0030745849609375, 1.002945556640625, 1.0031943969726562, 1.0027479248046876, 1.0030663452148438, 1.002672119140625, 1.0029179077148438, 1.0026751708984376, 1.002841064453125, 1.0025738525390624, 1.0032977905273437, 1.0027786254882813, 1.0031134643554687, 1.00257177734375, 1.0031124267578124, 1.0026741943359374, 1.0032947387695312, 1.0030386962890625, 1.0032332763671874, 2.068487060546875, 1.0029629516601561, 1.0030684204101563, 1.0028482666015626, 1.0030899047851562, 1.0025728149414062, 1.0029260864257812, 1.0024949951171875, 1.0026229858398437, 1.0027960205078126, 1.003514892578125, 1.0023598022460938, 1.0027949829101563, 1.0025420532226563, 1.0026260375976563, 1.00259228515625, 1.0028103637695311, 1.002603515625, 1.00250830078125, 1.002608642578125, 1.0032783203125, 1.0025861206054687, 1.0032824096679687, 1.0027950439453126, 1.0030663452148438, 1.0023987426757812, 1.0027417602539062, 1.0023075561523438, 1.0025850830078125, 1.0028851318359375, 1.003109375, 1.0031339721679688, 1.003430908203125, 1.0028308715820313, 1.00331005859375, 1.0026096801757813, 1.0028687133789063, 1.0031749267578125, 1.003378662109375, 1.0034810791015625, 1.0036776733398438, 1.0027991333007813, 1.0026577758789061, 1.0023618774414063, 1.002555419921875, 1.0025308227539063, 1.00274072265625, 1.0025973510742188, 1.0027908935546874, 1.003040771484375, 1.0038589477539062, 1.002471435546875, 1.003072509765625, 1.0029865112304688, 1.0029404296875, 1.0026116943359376, 1.003293701171875, 1.00347802734375, 1.0035588989257813, 1.00347802734375, 1.0037841796875, 1.0034053344726563, 1.0032077026367188]",tokens/s,0.9821471034300691,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/recurrentgemma-7b,google/recurrentgemma-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 404 Client Error: Not Found for url: https://huggingface.co/google/recurrentgemma-7b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1823, in _raise_on_head_call_error + raise head_call_error + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 352, in hf_raise_for_status + raise RepositoryNotFoundError(message, response) from e +huggingface_hub.utils._errors.RepositoryNotFoundError: 404 Client Error. (Request ID: Root=1-664aaac7-230defcf3ca338ce0b586efa;26c9392d-28c6-4de7-aa1c-ac69f7615fa0) + +Repository Not Found for url: https://huggingface.co/google/recurrentgemma-7b/resolve/main/config.json. +Please make sure you specified the correct `repo_id` and `repo_type`. +If you are trying to access a private or gated repo, make sure you are authenticated. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 422, in cached_file + raise EnvironmentError( +OSError: google/recurrentgemma-7b is not a local folder and is not a valid model identifier listed on 'https://huggingface.co/models' +If this is a private repository, make sure to pass a token having permission to this repo either by logging in with `huggingface-cli login` or by passing `token=` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-410m,EleutherAI/pythia-410m,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,1378.24256,1340.604416,0.0,694.157312,598.498816,s,10,0.5756522903442383,0.05756522903442383,0.0016369727383896268,0.0568012638092041,0.05918429183959961,0.06064101028442383,0.061806385040283206,"[0.06209772872924805, 0.05672822570800781, 0.056688095092773436, 0.05679859161376953, 0.05886057662963867, 0.057347423553466795, 0.05689571380615235, 0.05671388626098633, 0.05680393600463867, 0.05671811294555664]",tokens/s,4447.129009890898,kWh,6.782567046937489e-07,3.716537143845926e-07,1.9805285685486103e-06,3.0304389876269517e-06,tokens/kWh,84476209.89738722,MB,1378.24256,1340.604416,0.0,694.157312,659.031552,s,10,35.66919848632813,3.566919848632813,0.03950000205433062,3.552089111328125,3.624692236328125,3.64104033203125,3.65411880859375,"[3.57450146484375, 3.532454345703125, 3.56766015625, 3.657388427734375, 3.621059326171875, 3.5402509765625, 3.56031982421875, 3.532387939453125, 3.539317626953125, 3.5438583984375]",tokens/s,17.662297633109883,kWh,4.217196008799568e-05,2.3111615147015983e-05,0.00011743306696224983,0.00018271664219726155,tokens/kWh,344796.1786205822,,s,629,36.1255710258484,0.05743334026366993,0.0068485805293125165,0.05611315155029297,0.058175487518310545,0.05834158096313477,0.11305373931884766,"[0.05799008178710938, 0.05708492660522461, 0.05782732772827148, 0.05744332885742188, 0.057527294158935545, 0.05809561538696289, 0.0565401611328125, 0.05911040115356445, 0.05911040115356445, 0.058807296752929686, 0.058791934967041014, 0.05813555145263672, 0.05813248062133789, 0.05811199951171875, 0.058054656982421876, 0.05821952056884765, 0.057041919708251954, 0.056956928253173826, 0.057338878631591796, 0.05677363204956055, 0.05590323257446289, 0.05592268753051758, 0.05590425491333008, 0.05611008071899414, 0.05601484680175781, 0.05587148666381836, 0.055932926177978515, 0.056509441375732425, 0.05619200134277344, 0.056089599609375, 0.05595647811889649, 0.055981056213378906, 0.056079360961914064, 0.05589503860473633, 0.0559554557800293, 0.056065025329589846, 0.0564029426574707, 0.05732863998413086, 0.057398273468017576, 0.05806694412231445, 0.05649203109741211, 0.05601792144775391, 0.05590425491333008, 0.05598822402954102, 0.05598310470581055, 0.05660671997070312, 0.057234432220458986, 0.05611315155029297, 0.05587251281738281, 0.055923713684082034, 0.05598822402954102, 0.056018943786621096, 0.05615820693969727, 0.05591244888305664, 0.05669478225708008, 0.05584281539916992, 0.05593395233154297, 0.05587148666381836, 0.056016895294189455, 0.056000511169433595, 0.05590016174316406, 0.05584896087646484, 0.11391897583007812, 0.056886272430419924, 0.05697945785522461, 0.05592268753051758, 0.0564951057434082, 0.05716787338256836, 0.056594432830810545, 0.05600153732299805, 0.055695358276367186, 0.05583564758300781, 0.05596057510375976, 0.055913471221923826, 0.05584896087646484, 0.05577318572998047, 0.05587148666381836, 0.056035327911376956, 0.055943168640136716, 0.05580799865722656, 0.05776486587524414, 0.057622528076171874, 0.05631180953979492, 0.055818241119384764, 0.055777278900146485, 0.05587251281738281, 0.0560118408203125, 0.05598918533325195, 0.05593395233154297, 0.05613772964477539, 0.056908798217773435, 0.05586841583251953, 0.055809024810791016, 0.05591961669921875, 0.05592063903808594, 0.05590425491333008, 0.056174591064453126, 0.05596364974975586, 0.05545369720458984, 0.05593804931640625, 0.055806976318359375, 0.05584896087646484, 0.05586431884765625, 0.055890945434570315, 0.055839744567871094, 0.05590630340576172, 0.055741439819335936, 0.05607321548461914, 0.055858177185058595, 0.05598515319824219, 0.05599641418457031, 0.056035327911376956, 0.056120319366455076, 0.05593907165527344, 0.05596364974975586, 0.055932926177978515, 0.055947265625, 0.05584076690673828, 0.05575884628295898, 0.055823360443115234, 0.05576704025268555, 0.05586636734008789, 0.055923713684082034, 0.05593088150024414, 0.0557916145324707, 0.11292671966552735, 0.05593600082397461, 0.056169471740722655, 0.055927806854248044, 0.05587558364868164, 0.05590835189819336, 0.055769088745117185, 0.056676353454589844, 0.05695078277587891, 0.05639987182617188, 0.056005630493164066, 0.05686995315551758, 0.05640902328491211, 0.05631488037109375, 0.05584588623046875, 0.057027584075927736, 0.05787955093383789, 0.05715967941284179, 0.05641113662719727, 0.05674803161621094, 0.057038848876953124, 0.05642956924438477, 0.05662105560302735, 0.05600460815429688, 0.05647359848022461, 0.056395774841308595, 0.05592473602294922, 0.05653299331665039, 0.05728460693359375, 0.056796192169189456, 0.056970207214355466, 0.05783552169799805, 0.056842239379882815, 0.05652377700805664, 0.05740031814575195, 0.056120319366455076, 0.05686272048950195, 0.05625446319580078, 0.05662105560302735, 0.056174591064453126, 0.05706854248046875, 0.056346622467041016, 0.05597798538208008, 0.05667020797729492, 0.05655551910400391, 0.05687705612182617, 0.057256961822509764, 0.05748121643066406, 0.057133056640625, 0.05615923309326172, 0.055962623596191405, 0.05652070236206055, 0.056753150939941405, 0.056771583557128906, 0.056970241546630856, 0.05616844940185547, 0.057388031005859375, 0.05657702255249023, 0.057078784942626956, 0.05769420623779297, 0.05860966491699219, 0.056180736541748044, 0.05587148666381836, 0.11307008361816406, 0.05602406311035156, 0.05604044723510742, 0.056018943786621096, 0.058791934967041014, 0.0583741455078125, 0.05816524887084961, 0.05799321746826172, 0.058156032562255856, 0.058518527984619144, 0.058423297882080075, 0.05813350296020508, 0.058103809356689455, 0.05830451202392578, 0.058180606842041016, 0.058041343688964846, 0.05809254455566406, 0.05821952056884765, 0.05822873687744141, 0.05809151840209961, 0.058120193481445315, 0.05805158233642578, 0.05811609649658203, 0.05830348968505859, 0.058006526947021485, 0.05869670486450195, 0.058369022369384765, 0.058347518920898435, 0.058028030395507815, 0.058055679321289064, 0.058210304260253906, 0.0581058578491211, 0.058054656982421876, 0.058288158416748045, 0.058100704193115235, 0.05831987380981445, 0.058275840759277345, 0.05816524887084961, 0.05805977630615235, 0.05837823867797851, 0.05834444808959961, 0.05804032135009766, 0.05805158233642578, 0.05809971237182617, 0.058052608489990234, 0.05806489562988281, 0.05809254455566406, 0.05809664154052734, 0.05816729736328125, 0.0582042236328125, 0.05809862518310547, 0.05814169692993164, 0.05818675231933594, 0.057608192443847656, 0.05821440124511719, 0.058124320983886715, 0.058272735595703125, 0.058308609008789064, 0.058169345855712894, 0.058071041107177736, 0.058175487518310545, 0.05823897552490234, 0.0583372802734375, 0.11740262603759766, 0.058298366546630856, 0.0585544319152832, 0.05803107070922851, 0.05821948623657226, 0.05826047897338867, 0.05810793685913086, 0.058076126098632816, 0.0581662712097168, 0.05814988708496094, 0.05825843048095703, 0.05809356689453125, 0.05850931167602539, 0.058336254119873046, 0.058426368713378904, 0.05825331115722656, 0.058418174743652344, 0.0581396484375, 0.05809254455566406, 0.05811097717285156, 0.05806796646118164, 0.05830144119262695, 0.05805977630615235, 0.05824512100219727, 0.058036224365234375, 0.058142719268798826, 0.05810892868041992, 0.05846227264404297, 0.058175487518310545, 0.058110912322998046, 0.05813248062133789, 0.0584089584350586, 0.05830451202392578, 0.0581396484375, 0.058211326599121094, 0.05804851150512695, 0.0581662712097168, 0.05817350387573242, 0.05801055908203125, 0.0581662712097168, 0.05825331115722656, 0.05611929702758789, 0.05594521713256836, 0.055949310302734374, 0.05611008071899414, 0.056019966125488284, 0.055982078552246094, 0.055962623596191405, 0.05608448028564453, 0.05570457458496094, 0.056204288482666016, 0.05575372695922851, 0.055787521362304686, 0.05749350357055664, 0.05674803161621094, 0.055940097808837894, 0.05600460815429688, 0.056019966125488284, 0.05588787078857422, 0.05589913558959961, 0.05583052825927735, 0.055907329559326174, 0.05695283126831055, 0.11374079895019532, 0.055927806854248044, 0.05590323257446289, 0.05608652877807617, 0.05595340728759766, 0.056174591064453126, 0.05607321548461914, 0.05605686569213867, 0.056255455017089846, 0.05570969772338867, 0.05589606475830078, 0.05596063995361328, 0.056186817169189454, 0.05588991928100586, 0.055947265625, 0.055981056213378906, 0.056204288482666016, 0.05590016174316406, 0.055982078552246094, 0.05592575836181641, 0.05591862487792969, 0.055920608520507814, 0.05575475311279297, 0.055944190979003904, 0.05595852661132812, 0.055911422729492184, 0.05676134490966797, 0.05655039978027344, 0.055940097808837894, 0.055934974670410156, 0.05602918243408203, 0.056057857513427733, 0.055940097808837894, 0.05590425491333008, 0.056027137756347656, 0.05611724853515625, 0.055994369506835937, 0.05601484680175781, 0.05608448028564453, 0.055962623596191405, 0.05571177673339844, 0.055808990478515626, 0.05793484878540039, 0.057431041717529295, 0.056551422119140625, 0.05656576156616211, 0.05593600082397461, 0.05665689468383789, 0.05718732833862305, 0.055906368255615235, 0.058727359771728514, 0.05800755310058594, 0.05647257614135742, 0.056513534545898435, 0.05671116638183594, 0.05588172912597656, 0.056030208587646485, 0.05598003387451172, 0.05590528106689453, 0.05593600082397461, 0.05535129547119141, 0.05544755172729492, 0.05648691177368164, 0.11342233276367188, 0.05600358581542969, 0.05600153732299805, 0.05581107330322266, 0.055964672088623046, 0.05609062576293945, 0.05611724853515625, 0.05599846267700195, 0.056147968292236325, 0.05704908752441406, 0.05751091384887695, 0.05676031875610352, 0.05597798538208008, 0.05692620849609375, 0.05621247863769531, 0.05710335922241211, 0.05665280151367187, 0.055989246368408206, 0.0569989128112793, 0.056965118408203126, 0.056476673126220706, 0.05757952117919922, 0.05608243179321289, 0.057240577697753904, 0.05665280151367187, 0.056497150421142575, 0.057215999603271485, 0.0573573112487793, 0.05771366500854492, 0.056035327911376956, 0.05668044662475586, 0.05652479934692383, 0.056048641204833986, 0.056766464233398435, 0.05603737640380859, 0.05681568145751953, 0.05561439895629883, 0.055951358795166016, 0.05671014404296875, 0.05602099227905273, 0.05684531021118164, 0.05712998580932617, 0.05622579193115235, 0.05718425750732422, 0.05587046432495117, 0.05815193557739258, 0.05669580841064453, 0.056288257598876956, 0.05639168167114258, 0.05692416000366211, 0.05890764617919922, 0.05709619140625, 0.055943168640136716, 0.05594521713256836, 0.05592473602294922, 0.055702529907226565, 0.05582233428955078, 0.05554380798339844, 0.05590425491333008, 0.0567193603515625, 0.05655756759643555, 0.05586636734008789, 0.05595238494873047, 0.11324825286865234, 0.056097793579101565, 0.05586841583251953, 0.0565401611328125, 0.05603737640380859, 0.05586841583251953, 0.05586841583251953, 0.055787582397460934, 0.05582944107055664, 0.05609062576293945, 0.05594524765014648, 0.05589398574829101, 0.05601484680175781, 0.05601587295532227, 0.05612851333618164, 0.056167423248291014, 0.05636508941650391, 0.05600048065185547, 0.05591449737548828, 0.05615820693969727, 0.05590425491333008, 0.056697856903076174, 0.056491008758544924, 0.05597491073608398, 0.0558837776184082, 0.055894016265869144, 0.05606195068359375, 0.056013824462890625, 0.055877632141113284, 0.0558131217956543, 0.0560445442199707, 0.05591756820678711, 0.05575478363037109, 0.05573014450073242, 0.055923713684082034, 0.05588889694213867, 0.05579673767089844, 0.05603123092651367, 0.05595340728759766, 0.05692416000366211, 0.05709721755981445, 0.0572149772644043, 0.05634969711303711, 0.0558919677734375, 0.05590528106689453, 0.0558766098022461, 0.055841793060302736, 0.05595033645629883, 0.05608448028564453, 0.05609267044067383, 0.05588479995727539, 0.05583769607543945, 0.05596160125732422, 0.0559554557800293, 0.05604556655883789, 0.05690572738647461, 0.05594214248657227, 0.055856128692626954, 0.05609267044067383, 0.05597183990478516, 0.0560076789855957, 0.05595340728759766, 0.055993408203125, 0.11313145446777344, 0.05605068969726563, 0.05714432144165039, 0.05607219314575195, 0.055894016265869144, 0.05605990219116211, 0.055856128692626954, 0.055787521362304686, 0.0557916145324707, 0.05595647811889649, 0.05597183990478516, 0.055567359924316405, 0.05594521713256836, 0.05598720169067383, 0.05587251281738281, 0.055973888397216794, 0.05577318572998047, 0.055757823944091796, 0.055972862243652347, 0.05598310470581055, 0.05603839874267578, 0.056036350250244144, 0.05586438369750977, 0.055879615783691404, 0.05582950210571289, 0.056062976837158204, 0.055747615814208985, 0.05714838409423828, 0.05821952056884765, 0.0560711669921875, 0.05586739349365234, 0.05611520004272461, 0.05608857727050781, 0.05595340728759766, 0.0560076789855957, 0.055930912017822264, 0.05865881729125977, 0.05698665618896484, 0.0559431037902832, 0.05590835189819336, 0.05589811325073242, 0.05590323257446289, 0.05572403335571289, 0.05595852661132812, 0.055923713684082034, 0.05552844619750977, 0.05647052764892578, 0.05722623825073242, 0.05685452651977539, 0.05584384155273438, 0.05696409606933594, 0.05613260650634765, 0.05573222351074219, 0.05603123092651367, 0.05594214248657227, 0.05622579193115235, 0.05601587295532227, 0.055932926177978515, 0.05593907165527344, 0.05829119873046875, 0.05608038330078125, 0.056581119537353515, 0.056043521881103515, 0.11301171112060547, 0.05596160125732422, 0.05600460815429688, 0.05594828796386719, 0.0573306884765625, 0.05730099105834961, 0.056136703491210936, 0.05597183990478516, 0.055757823944091796, 0.05651865768432617, 0.05613772964477539, 0.05588684844970703, 0.05589811325073242, 0.055934974670410156, 0.055951358795166016, 0.05582438278198242, 0.055932926177978515, 0.05587251281738281, 0.05707059097290039, 0.057232383728027345, 0.057132030487060545, 0.056750080108642575, 0.05586227035522461, 0.055964672088623046, 0.05600972747802734, 0.05593088150024414, 0.05628518295288086, 0.05651148986816406, 0.056172542572021485, 0.05634764862060547, 0.055940097808837894, 0.0572149772644043, 0.05617049789428711, 0.05602201461791992, 0.05591654586791992, 0.05619609451293945, 0.0580239372253418, 0.056460289001464846, 0.055949310302734374, 0.056018943786621096, 0.056033279418945314, 0.05574860763549805, 0.0558960952758789, 0.056247264862060546, 0.05595443344116211, 0.05688217544555664, 0.05635276794433594, 0.055982078552246094, 0.055987232208251955, 0.0560035514831543, 0.05590937423706055, 0.05664051055908203, 0.05720883178710937, 0.056853504180908204, 0.056360958099365234, 0.0558919677734375, 0.05595852661132812, 0.05596160125732422, 0.05597491073608398, 0.05603228759765625, 0.0562564811706543, 0.055975936889648435, 0.055949310302734374]",tokens/s,17.41148948344487,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-32B,Qwen/Qwen1.5-32B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,4397.162496,24111.480832,0.0,23465.033728,21690.932224,s,10,26.23831787109375,2.623831787109375,0.003031033246657479,2.6228023681640624,2.627721069335937,2.628540100097656,2.629195324707031,"[2.629359130859375, 2.621765380859375, 2.6223037109375, 2.621092041015625, 2.623301025390625, 2.620621826171875, 2.62457861328125, 2.6205380859375, 2.627218994140625, 2.6275390625]",tokens/s,97.56723020801203,kWh,3.09448529779911e-05,1.6958929474258183e-05,0.00014797261837800056,0.00019587640083024984,tokens/kWh,1306946.619985398,MB,4397.162496,24111.480832,0.0,23465.033728,21890.217984,s,10,1555.758796875,155.5758796875,0.01362531366110068,155.577890625,155.5897375,155.594603125,155.598495625,"[155.55940625, 155.57425, 155.583734375, 155.59946875, 155.58153125, 155.58503125, 155.568875, 155.58865625, 155.564453125, 155.553390625]",tokens/s,0.4049470915835152,kWh,0.0018369476799004608,0.0010068097738273171,0.008895798588855384,0.011739556042583163,tokens/kWh,5366.4721026484,,s,629,1576.8986267089822,2.50699304723209,0.31148250553470674,2.46929931640625,2.47089541015625,2.471475439453125,5.08945892578125,"[2.468798583984375, 2.46915185546875, 2.46936669921875, 2.469295166015625, 2.469130126953125, 2.468601806640625, 2.46879638671875, 2.46974560546875, 2.468950927734375, 2.468516845703125, 2.468890625, 2.469065673828125, 2.468599853515625, 2.46870947265625, 2.468912109375, 2.469274658203125, 2.468938720703125, 2.46896337890625, 2.4693310546875, 2.46970068359375, 2.469392333984375, 2.46949072265625, 2.47088330078125, 2.4697119140625, 2.469168212890625, 2.468552734375, 2.4692490234375, 2.4690791015625, 2.47006005859375, 2.46929931640625, 2.46951123046875, 2.46925, 2.469129150390625, 2.468915283203125, 2.46999560546875, 2.4695, 2.46917724609375, 2.4696279296875, 2.469708740234375, 2.469084228515625, 2.46938427734375, 2.468663330078125, 2.46932275390625, 2.468631591796875, 2.468556884765625, 2.46858447265625, 2.46909326171875, 2.46888232421875, 2.46854248046875, 2.469202880859375, 2.469517333984375, 2.468483154296875, 2.468864013671875, 2.46881494140625, 2.468998046875, 2.469455810546875, 2.469496826171875, 2.469419921875, 2.4694599609375, 2.469094482421875, 2.47101025390625, 2.468877197265625, 5.09138330078125, 2.468682861328125, 2.468284423828125, 2.471290771484375, 2.471075927734375, 2.47086279296875, 2.470519775390625, 2.471779296875, 2.4712724609375, 2.470220703125, 2.469166015625, 2.47142919921875, 2.469087158203125, 2.470717529296875, 2.469718994140625, 2.4686796875, 2.468135986328125, 2.469396484375, 2.47069482421875, 2.469185546875, 2.46943359375, 2.4694208984375, 2.469435302734375, 2.472340576171875, 2.46964404296875, 2.4692294921875, 2.4692080078125, 2.4687861328125, 2.4688701171875, 2.468831298828125, 2.468864013671875, 2.468359130859375, 2.468737060546875, 2.4681728515625, 2.46879345703125, 2.47047265625, 2.471103515625, 2.4709150390625, 2.469086181640625, 2.467948486328125, 2.46850244140625, 2.46847998046875, 2.468877197265625, 2.47016845703125, 2.472048583984375, 2.471004150390625, 2.46881689453125, 2.46902783203125, 2.469129150390625, 2.468443115234375, 2.468276123046875, 2.46843603515625, 2.468770751953125, 2.46885888671875, 2.468476806640625, 2.46837255859375, 2.46976318359375, 2.4682578125, 2.468187255859375, 2.467946533203125, 2.4700517578125, 2.468588623046875, 2.468370361328125, 5.09084375, 2.46865625, 2.4686162109375, 2.471500732421875, 2.47128173828125, 2.47047265625, 2.470948974609375, 2.46904638671875, 2.46915283203125, 2.468949951171875, 2.469103515625, 2.469283935546875, 2.46955615234375, 2.46936669921875, 2.46929931640625, 2.469792724609375, 2.46934521484375, 2.468339599609375, 2.469074951171875, 2.46917724609375, 2.46955322265625, 2.469060546875, 2.471446533203125, 2.471505859375, 2.471520263671875, 2.46911376953125, 2.470220703125, 2.46964111328125, 2.46999658203125, 2.469699462890625, 2.469473388671875, 2.4692724609375, 2.469899169921875, 2.469782470703125, 2.47012158203125, 2.469866455078125, 2.4691630859375, 2.46879345703125, 2.469590087890625, 2.47096533203125, 2.471996337890625, 2.46934326171875, 2.468957275390625, 2.468894775390625, 2.468242431640625, 2.470201416015625, 2.46946923828125, 2.469783447265625, 2.469652587890625, 2.469385009765625, 2.4691455078125, 2.469442626953125, 2.46881494140625, 2.470005859375, 2.469909423828125, 2.46930224609375, 2.469780517578125, 2.46929296875, 2.468728759765625, 2.46875146484375, 2.468708251953125, 2.469340087890625, 2.468431884765625, 5.08925537109375, 2.469267333984375, 2.4689326171875, 2.468461669921875, 2.470912841796875, 2.47052294921875, 2.47040625, 2.4696259765625, 2.470639404296875, 2.470507568359375, 2.469961669921875, 2.4698583984375, 2.470289306640625, 2.470327392578125, 2.470380615234375, 2.470289306640625, 2.47063037109375, 2.4716962890625, 2.4711904296875, 2.469096435546875, 2.46898583984375, 2.469234619140625, 2.46925927734375, 2.469075927734375, 2.468787109375, 2.470892578125, 2.468845458984375, 2.470190185546875, 2.470021240234375, 2.46980908203125, 2.469390380859375, 2.469427001953125, 2.469474365234375, 2.46984912109375, 2.468821044921875, 2.46932470703125, 2.46972412109375, 2.470828125, 2.470804443359375, 2.47084228515625, 2.471381103515625, 2.471739501953125, 2.4696513671875, 2.469551025390625, 2.46862744140625, 2.468588623046875, 2.468927490234375, 2.469970947265625, 2.469509033203125, 2.468726806640625, 2.470922119140625, 2.47033642578125, 2.470152099609375, 2.47086181640625, 2.469655517578125, 2.469350341796875, 2.469897216796875, 2.4691435546875, 2.468494384765625, 2.469423095703125, 2.470116455078125, 2.470024169921875, 2.4697353515625, 5.09418994140625, 2.47079833984375, 2.471482421875, 2.470703125, 2.4698798828125, 2.469234619140625, 2.469583984375, 2.469041259765625, 2.469078125, 2.468781982421875, 2.469856201171875, 2.470139892578125, 2.4699013671875, 2.46999853515625, 2.469337158203125, 2.4686396484375, 2.46900732421875, 2.46865625, 2.4688505859375, 2.46949365234375, 2.468842529296875, 2.469308349609375, 2.469699462890625, 2.469474365234375, 2.46980810546875, 2.470485107421875, 2.470642578125, 2.470287353515625, 2.46951220703125, 2.46904931640625, 2.47022900390625, 2.470436767578125, 2.46926025390625, 2.469675048828125, 2.47014306640625, 2.46932275390625, 2.468830322265625, 2.469003173828125, 2.469866455078125, 2.46951123046875, 2.468826171875, 2.469062744140625, 2.469338134765625, 2.469421142578125, 2.468864013671875, 2.4689970703125, 2.469686279296875, 2.468821044921875, 2.468958251953125, 2.46921533203125, 2.4694580078125, 2.468634521484375, 2.46913232421875, 2.4693955078125, 2.469814208984375, 2.468588623046875, 2.469433349609375, 2.470032470703125, 2.470792236328125, 2.469285888671875, 2.469370849609375, 2.469603271484375, 2.46991357421875, 5.0895380859375, 2.472522705078125, 2.472004638671875, 2.469856201171875, 2.468981689453125, 2.469172119140625, 2.46925, 2.4692265625, 2.469295166015625, 2.46936376953125, 2.47034765625, 2.468842529296875, 2.470485107421875, 2.471478271484375, 2.469750732421875, 2.468664306640625, 2.471371826171875, 2.468883544921875, 2.471439453125, 2.469444580078125, 2.46917626953125, 2.469329833984375, 2.469687255859375, 2.46869091796875, 2.469650390625, 2.468767822265625, 2.46934619140625, 2.46850048828125, 2.46826806640625, 2.46904443359375, 2.4693974609375, 2.469318603515625, 2.47453076171875, 2.469267333984375, 2.470331298828125, 2.4707685546875, 2.469550048828125, 2.46943017578125, 2.469969970703125, 2.469962646484375, 2.46923486328125, 2.468137939453125, 2.468229248046875, 2.46898779296875, 2.468890625, 2.468874267578125, 2.469168212890625, 2.470299560546875, 2.47147119140625, 2.469370849609375, 2.469350341796875, 2.470095947265625, 2.469555419921875, 2.46888232421875, 2.4688681640625, 2.469992431640625, 2.468869140625, 2.469044189453125, 2.469062744140625, 2.469525390625, 2.468809814453125, 2.46928076171875, 2.468957275390625, 5.09382470703125, 2.469845947265625, 2.4695244140625, 2.469098388671875, 2.46944677734375, 2.469396484375, 2.470770751953125, 2.469032958984375, 2.469140380859375, 2.468855712890625, 2.469170166015625, 2.469623779296875, 2.469095458984375, 2.4704697265625, 2.47050244140625, 2.4702392578125, 2.469874755859375, 2.4695224609375, 2.469214111328125, 2.469747802734375, 2.46868896484375, 2.469425048828125, 2.468915283203125, 2.4693955078125, 2.468820068359375, 2.46910986328125, 2.468548583984375, 2.4704716796875, 2.469411865234375, 2.4691220703125, 2.468193359375, 2.468788330078125, 2.4683017578125, 2.46884033203125, 2.470220703125, 2.4702197265625, 2.468990966796875, 2.468908935546875, 2.4688486328125, 2.4704501953125, 2.470517822265625, 2.470140869140625, 2.469308349609375, 2.470096923828125, 2.469972900390625, 2.4689111328125, 2.468103271484375, 2.468663330078125, 2.468252685546875, 2.468338623046875, 2.467991455078125, 2.468906982421875, 2.46822802734375, 2.468630615234375, 2.469042236328125, 2.469347412109375, 2.47062939453125, 2.471343017578125, 2.467967041015625, 2.468644775390625, 2.46843701171875, 2.4686162109375, 2.473323486328125, 5.0889111328125, 2.468672607421875, 2.4683642578125, 2.468341796875, 2.468705322265625, 2.468413330078125, 2.46852197265625, 2.468662353515625, 2.468912109375, 2.46987158203125, 2.470073486328125, 2.46920068359375, 2.469920654296875, 2.46879638671875, 2.4687412109375, 2.4697333984375, 2.4695244140625, 2.4696298828125, 2.4695244140625, 2.470005859375, 2.468862060546875, 2.47003125, 2.4691630859375, 2.468483154296875, 2.4710625, 2.469022705078125, 2.46885888671875, 2.4692705078125, 2.469971923828125, 2.47102783203125, 2.47071435546875, 2.4704482421875, 2.471227294921875, 2.470994873046875, 2.469182373046875, 2.46873095703125, 2.47003125, 2.468622314453125, 2.4721171875, 2.468535400390625, 2.47090673828125, 2.470928466796875, 2.470781982421875, 2.470194091796875, 2.470340576171875, 2.468509765625, 2.468907958984375, 2.469888916015625, 2.469969970703125, 2.47013671875, 2.472005859375, 2.47183740234375, 2.47176708984375, 2.471166015625, 2.471228515625, 2.4702783203125, 2.47012451171875, 2.468314208984375, 2.46890087890625, 2.46814208984375, 2.46915283203125, 2.467857421875, 2.46899609375, 5.0967275390625, 2.468865966796875, 2.47231494140625, 2.4689765625, 2.469561279296875, 2.469706787109375, 2.468801513671875, 2.468005859375, 2.46970068359375, 2.469357666015625, 2.470153076171875, 2.47166455078125, 2.4716962890625, 2.47134716796875, 2.4716728515625, 2.469540771484375, 2.46961865234375, 2.46875439453125, 2.468744140625, 2.469203857421875, 2.46841748046875, 2.469667724609375, 2.4690810546875, 2.46848828125, 2.468600830078125, 2.469570556640625, 2.46897265625, 2.468644775390625, 2.46887109375, 2.46993310546875, 2.469518310546875, 2.468893798828125, 2.469458740234375, 2.4689755859375, 2.469214111328125, 2.468822998046875, 2.468765625, 2.469544921875, 2.46893359375, 2.46794140625, 2.46793115234375, 2.468474853515625, 2.468509765625, 2.469032958984375, 2.468798583984375, 2.46907177734375, 2.469205078125, 2.46849853515625, 2.468729736328125, 2.469506103515625, 2.4690390625, 2.4686376953125, 2.4710419921875, 2.469719970703125, 2.4685322265625, 2.468116455078125, 2.468727783203125, 2.4698369140625, 2.46875244140625, 2.46940869140625, 2.469123046875, 2.47010205078125, 2.468192138671875, 5.09570947265625, 2.469843994140625, 2.469245849609375, 2.46862939453125, 2.469123046875, 2.470595703125, 2.46889990234375, 2.468484130859375, 2.468884521484375, 2.469498779296875, 2.46938720703125, 2.468440185546875, 2.4687841796875, 2.46934326171875, 2.468601806640625, 2.469178466796875, 2.46820654296875, 2.469718017578125, 2.469128173828125, 2.468865966796875, 2.468513671875, 2.46963525390625, 2.468724609375, 2.4697353515625, 2.468211669921875, 2.468957275390625, 2.4694794921875, 2.46911083984375, 2.4683447265625, 2.4698837890625, 2.468577392578125, 2.468810791015625, 2.468404296875, 2.46858251953125, 2.469440673828125, 2.469866455078125, 2.468959228515625, 2.4695625, 2.46875732421875, 2.46845849609375, 2.469051513671875, 2.46925830078125, 2.470119384765625, 2.468737060546875, 2.468697998046875, 2.469822509765625, 2.469697509765625, 2.4686376953125, 2.4682802734375, 2.468843505859375, 2.471318603515625, 2.4700693359375, 2.469341064453125, 2.469017578125, 2.469211181640625, 2.469179443359375, 2.468263916015625, 2.46847998046875, 2.4693955078125, 2.468968505859375, 2.469316650390625, 2.46883740234375, 2.46984814453125]",tokens/s,0.39888423348603846,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-4.5B,facebook/xglm-4.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1469, in _autoset_attn_implementation + cls._check_and_enable_flash_attn_2( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1560, in _check_and_enable_flash_attn_2 + raise ValueError( +ValueError: XGLMForCausalLM does not support Flash Attention 2.0 yet. Please request to add support where the model is hosted, on its model hub page: https://huggingface.co//tmp/tmpj_4j1da9/no_weights_model/discussions/new or in the Transformers GitHub repo: https://github.com/huggingface/transformers/issues/new + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,TencentARC/Mistral_Pro_8B_v0.1,TencentARC/Mistral_Pro_8B_v0.1,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,2454.69184,7298.613248,0.0,6652.166144,6323.221504,s,10,7.735312255859376,0.7735312255859375,0.0028429451285762273,0.7729794311523437,0.7762465698242188,0.7783619323730468,0.7800542224121094,"[0.780477294921875, 0.7757764892578125, 0.7711934204101563, 0.770524658203125, 0.7719502563476562, 0.7706873168945313, 0.7725225219726563, 0.7734363403320312, 0.7744036254882812, 0.77434033203125]",tokens/s,330.9497943099635,kWh,9.102796527565035e-06,4.9879214771863185e-06,4.3937086431680494e-05,5.802780443643185e-05,tokens/kWh,4411678.202997362,MB,2454.69184,7298.613248,0.0,6652.166144,6382.564864,s,10,458.5464296875,45.85464296875,0.0073453057278341505,45.852935546875,45.866698828124996,45.8672068359375,45.8676132421875,"[45.8568515625, 45.857828125, 45.85469140625, 45.8511796875, 45.845234375, 45.8491953125, 45.85046484375, 45.84668359375, 45.8665859375, 45.86771484375]",tokens/s,1.3739066738112995,kWh,0.0005414656486152074,0.00029677022395511814,0.002569976393586324,0.003408212266156649,tokens/kWh,18484.764175514054,,s,629,464.77344433593714,0.7389084965595191,0.09178579233429768,0.727773193359375,0.7285446655273438,0.7288094848632812,1.4987512353515624,"[0.72828515625, 0.7276810302734374, 0.7278029174804688, 0.72793701171875, 0.7272877807617187, 0.727383056640625, 0.7272335205078125, 0.7274444580078125, 0.7278551025390625, 0.727720947265625, 0.7276083374023438, 0.7281663818359375, 0.7275867919921875, 0.72804248046875, 0.72800048828125, 0.7276687622070312, 0.7281151733398438, 0.7279267578125, 0.7276728515625, 0.7276615600585937, 0.7278981323242187, 0.7287255249023438, 0.7286405029296875, 0.7275847778320312, 0.7281510620117188, 0.7281069946289063, 0.7279093627929687, 0.7283015747070313, 0.7283804321289062, 0.7276973876953124, 0.7277127685546875, 0.7276400756835938, 0.7282882690429687, 0.7277936401367188, 0.7284705200195313, 0.727784423828125, 0.727952392578125, 0.727930908203125, 0.728369140625, 0.728827880859375, 0.7278551025390625, 0.72924365234375, 0.7275161743164062, 0.7274977416992188, 0.7274014892578125, 0.727947265625, 0.7273707275390625, 0.7273738403320312, 0.7270891723632813, 0.72770458984375, 0.727604248046875, 0.7274721069335938, 0.728226806640625, 0.7281571655273438, 0.7279820556640625, 0.7282749633789063, 0.7275151977539063, 0.7279564208984375, 0.7273912353515625, 0.7277537231445312, 0.7282349853515625, 0.7278714599609375, 1.503824951171875, 0.72740966796875, 0.7277650146484375, 0.7272868041992188, 0.7274219360351563, 0.7278981323242187, 0.7287039794921875, 0.7276185302734375, 0.7274321899414062, 0.7271383056640625, 0.72722021484375, 0.7271024780273437, 0.7274373168945313, 0.7272499389648438, 0.7280332641601562, 0.7276113891601562, 0.7276943359375, 0.7272652587890625, 0.7283251342773438, 0.7285104370117188, 0.7281961059570312, 0.7280137939453125, 0.7283834838867187, 0.7276124267578125, 0.7279022216796875, 0.7284910278320312, 0.7285155639648437, 0.728142822265625, 0.7280394287109375, 0.7274547119140625, 0.7279277954101563, 0.7275509643554687, 0.7277987670898437, 0.7278591918945313, 0.728226806640625, 0.7301795654296875, 0.726887451171875, 0.727372802734375, 0.7274393310546875, 0.7276984252929688, 0.72789404296875, 0.7279503173828125, 0.7277659912109375, 0.7282677612304688, 0.7285678100585937, 0.7279226684570312, 0.7286671142578125, 0.7283712158203125, 0.7282360229492187, 0.727741455078125, 0.7277588500976563, 0.7276482543945313, 0.727857177734375, 0.7283681030273438, 0.7278796997070313, 0.7283937377929688, 0.727920654296875, 0.7278295288085938, 0.728015869140625, 0.7276452026367187, 0.7281694946289062, 0.7286343383789062, 0.7274926147460937, 1.4987724609375, 0.7273533325195313, 0.727457763671875, 0.7278960571289063, 0.7272560424804687, 0.7277005004882813, 0.7277772827148438, 0.7281305541992188, 0.7275653076171875, 0.727499755859375, 0.7276851196289063, 0.7280271606445312, 0.7270768432617187, 0.7270553588867188, 0.7272847290039063, 0.7274874877929688, 0.7273748779296875, 0.727520263671875, 0.727920654296875, 0.7277373657226562, 0.7273338623046876, 0.728322021484375, 0.7278253784179688, 0.7282452392578125, 0.72855859375, 0.7283507080078125, 0.7274547119140625, 0.727931884765625, 0.7276277465820312, 0.7273011474609375, 0.7271905517578126, 0.727736328125, 0.7286302490234375, 0.728036376953125, 0.727710693359375, 0.7286405029296875, 0.728036376953125, 0.7283230590820312, 0.72854833984375, 0.7284182739257813, 0.7284367065429688, 0.729618408203125, 0.7277322387695313, 0.7279042358398438, 0.7280332641601562, 0.7283035888671875, 0.7280281372070313, 0.7276656494140625, 0.7277639770507812, 0.7281858520507812, 0.7287337036132813, 0.7282145385742187, 0.7282606201171875, 0.7274495849609375, 0.727203857421875, 0.7272243041992188, 0.727235595703125, 0.7272662963867188, 0.7274495849609375, 0.727531494140625, 0.7289282836914063, 0.7277404174804688, 0.7281520385742187, 1.498346435546875, 0.7277035522460937, 0.7284224243164062, 0.7275243530273438, 0.727741455078125, 0.727731201171875, 0.7279708251953125, 0.7275581665039063, 0.7280773315429687, 0.7281356811523437, 0.7283568725585937, 0.7276912841796875, 0.7285411987304687, 0.7277701416015625, 0.7284715576171875, 0.72749462890625, 0.7280516967773437, 0.7273963623046875, 0.7273779296875, 0.7273543701171875, 0.7274270629882813, 0.7274116821289063, 0.727689208984375, 0.7282974853515625, 0.7277066040039063, 0.7288258666992188, 0.72807421875, 0.727720947265625, 0.728501220703125, 0.7287817993164063, 0.728057861328125, 0.7277598876953125, 0.72751513671875, 0.7276728515625, 0.7286773681640625, 0.7274321899414062, 0.7279042358398438, 0.727920654296875, 0.727573486328125, 0.7276513061523437, 0.7273072509765625, 0.7275028686523437, 0.7276226806640625, 0.727731201171875, 0.7275745239257813, 0.727709716796875, 0.7274833984375, 0.7272919311523437, 0.7273953247070313, 0.7276328735351563, 0.7272529907226563, 0.7279380493164063, 0.7277659912109375, 0.7278837890625, 0.7274598388671875, 0.7278622436523438, 0.727636962890625, 0.7282718505859375, 0.7279830932617187, 0.7275888671875, 0.72747314453125, 0.7274137573242188, 0.7276810302734374, 1.500190673828125, 0.7274137573242188, 0.727794677734375, 0.72776806640625, 0.7275325317382813, 0.728131591796875, 0.7274035034179688, 0.7273656616210937, 0.727741455078125, 0.7269683227539062, 0.7274475708007813, 0.727498779296875, 0.7274158325195312, 0.7276226806640625, 0.7277803344726562, 0.7272796020507812, 0.7277967529296875, 0.7278120727539062, 0.7281182861328125, 0.7282227172851562, 0.729248779296875, 0.7279073486328125, 0.727394287109375, 0.7274772338867187, 0.7273768920898438, 0.7279462280273438, 0.7280506591796875, 0.7277485961914063, 0.7276800537109375, 0.7273492431640625, 0.7277352905273438, 0.7280148315429688, 0.7277557983398437, 0.7279503173828125, 0.72796875, 0.727677978515625, 0.7273717651367188, 0.7274014892578125, 0.7277393798828125, 0.7273564453125, 0.7273922729492187, 0.7276431274414062, 0.7273267211914063, 0.727183349609375, 0.7276964111328125, 0.7270574340820313, 0.7279892578125, 0.7274506225585937, 0.7283281860351563, 0.7276236572265625, 0.7274864501953126, 0.7277271118164063, 0.7274741821289062, 0.7276431274414062, 0.727984130859375, 0.7277352905273438, 0.727568359375, 0.7290419311523437, 0.7279063110351562, 0.7274669799804687, 0.7279380493164063, 0.7273656616210937, 0.727736328125, 1.4986966552734375, 0.7277783203125, 0.7279697875976563, 0.7273380126953125, 0.7279759521484375, 0.7274772338867187, 0.7276113891601562, 0.72751513671875, 0.727520263671875, 0.7276339111328125, 0.7279185791015625, 0.7278212890625, 0.7281500244140625, 0.7277086791992188, 0.7276032104492187, 0.7277926635742188, 0.7276728515625, 0.7275827026367188, 0.7275140991210938, 0.7280179443359375, 0.7281879272460937, 0.7278212890625, 0.727699462890625, 0.72736767578125, 0.7276728515625, 0.7279083251953125, 0.7273850708007813, 0.7275069580078125, 0.7274024658203125, 0.7273922729492187, 0.7274700927734375, 0.7275847778320312, 0.7282288818359375, 0.7278090209960938, 0.7279667358398437, 0.7283240966796874, 0.7274004516601562, 0.7275899047851563, 0.729038818359375, 0.728173583984375, 0.727530517578125, 0.7272581176757813, 0.727572509765625, 0.7277127685546875, 0.72736767578125, 0.7276503295898438, 0.7283455810546875, 0.72766259765625, 0.7285360717773437, 0.7277281494140625, 0.72745166015625, 0.7276943359375, 0.7278653564453125, 0.7278253784179688, 0.7273615112304688, 0.7273717651367188, 0.7274547119140625, 0.7275899047851563, 0.7274557495117188, 0.7279779663085938, 0.7282565307617187, 0.728158203125, 0.7288606567382813, 1.500291015625, 0.7285186767578125, 0.7282554931640625, 0.727962646484375, 0.72707275390625, 0.7270717163085938, 0.7278837890625, 0.7273738403320312, 0.7277557983398437, 0.7276339111328125, 0.728394775390625, 0.7282606201171875, 0.728369140625, 0.727878662109375, 0.7280169067382812, 0.727794677734375, 0.7282175903320313, 0.7275038452148438, 0.7273502807617187, 0.7271946411132812, 0.7275499267578125, 0.7277998046875, 0.7273502807617187, 0.7283394775390625, 0.7281551513671874, 0.7276728515625, 0.729017333984375, 0.7274383544921875, 0.7284019165039063, 0.7275847778320312, 0.727709716796875, 0.7274024658203125, 0.7274024658203125, 0.7272263793945313, 0.7275448608398437, 0.727203857421875, 0.7272734985351562, 0.7281172485351562, 0.72766259765625, 0.7277854614257813, 0.7278192749023438, 0.7277168579101563, 0.7283128051757812, 0.7274014892578125, 0.7279892578125, 0.7273103637695313, 0.7277905883789062, 0.72785302734375, 0.7278776245117188, 0.727984130859375, 0.7275591430664062, 0.7280732421875, 0.728015869140625, 0.7281817626953125, 0.7277322387695313, 0.72749365234375, 0.7281449584960937, 0.7275447387695313, 0.7287470092773437, 0.727857177734375, 0.727446533203125, 0.7272703857421875, 0.7274813232421875, 1.5015628662109375, 0.7280670776367187, 0.7274475708007813, 0.7279002075195312, 0.7276564331054688, 0.727930908203125, 0.7274649658203125, 0.7278028564453125, 0.7277875366210937, 0.7275980834960938, 0.7279329223632812, 0.7276553955078126, 0.7276932983398438, 0.7277240600585938, 0.7281275024414062, 0.7278305053710937, 0.7283988647460937, 0.727973876953125, 0.7275120849609376, 0.7283138427734375, 0.7280885620117188, 0.728431640625, 0.7279882202148438, 0.7285555419921875, 0.727773193359375, 0.7274137573242188, 0.727572509765625, 0.7277168579101563, 0.7279124755859375, 0.7280169067382812, 0.7274721069335938, 0.7271577758789063, 0.7273502807617187, 0.7277578125, 0.72791552734375, 0.7280302124023438, 0.7272263793945313, 0.727066650390625, 0.7275479125976563, 0.7274383544921875, 0.727099365234375, 0.7272929077148438, 0.7277578125, 0.728215576171875, 0.727783447265625, 0.7276656494140625, 0.72743115234375, 0.7274690551757812, 0.7273277587890625, 0.7281571655273438, 0.7275397338867188, 0.727414794921875, 0.7275867919921875, 0.7271116943359375, 0.727257080078125, 0.727667724609375, 0.7276851196289063, 0.7272703857421875, 0.7279493408203125, 0.7277250366210938, 0.7286610107421875, 0.7273246459960937, 0.7279380493164063, 1.50266162109375, 0.72736767578125, 0.7276441650390625, 0.727141357421875, 0.7274188842773438, 0.7278018798828125, 0.727625732421875, 0.727541748046875, 0.7280783081054687, 0.7277342529296875, 0.72876953125, 0.728369140625, 0.7288411865234375, 0.7279749145507812, 0.7283046264648437, 0.72797900390625, 0.7282718505859375, 0.7281930541992188, 0.728658935546875, 0.7289108276367188, 0.7287183227539062, 0.7283169555664063, 0.7286814575195313, 0.7278428344726563, 0.7285463256835938, 0.72789404296875, 0.72875830078125, 0.72804248046875, 0.7291146240234375, 0.7289763793945313, 0.7286661376953125, 0.728784912109375, 0.728363037109375, 0.7279595336914062, 0.7274700927734375, 0.7275233154296875, 0.7272816772460937, 0.7279974365234375, 0.7284090576171875, 0.72762060546875, 0.728300537109375, 0.727457763671875, 0.7274711303710938, 0.727183349609375, 0.7278960571289063, 0.7272171630859375, 0.7289682006835938, 0.7285933837890625, 0.7276564331054688, 0.7285442504882812, 0.7269273681640624, 0.72795751953125, 0.7275642700195313, 0.7277936401367188, 0.7290480346679687, 0.72764208984375, 0.727804931640625, 0.7280322265625, 0.7278858032226563, 0.727383056640625, 0.7281940307617187, 0.7273421020507812, 0.72789404296875, 1.502066650390625, 0.7280179443359375, 0.7290101928710937, 0.7292303466796874, 0.7280660400390625, 0.728395751953125, 0.72835791015625, 0.72821142578125, 0.7285330200195312, 0.7277250366210938, 0.728616943359375, 0.7284172973632812, 0.7285718994140625, 0.728131591796875, 0.7278837890625, 0.7286558837890625, 0.7279124755859375, 0.7291678466796875, 0.728326171875, 0.7273584594726562, 0.7271588134765625, 0.7275980834960938, 0.7273738403320312, 0.7279912719726562, 0.72772607421875, 0.7297515258789062, 0.7277824096679687, 0.7272632446289062, 0.7278909301757812, 0.7273318481445312, 0.7274424438476562, 0.7272509155273438, 0.7275796508789063, 0.7279483032226562, 0.7278059692382812, 0.7278919677734375, 0.7276881713867187, 0.7284940795898438, 0.7285985107421875, 0.7283128051757812, 0.7282175903320313, 0.7282974853515625, 0.728426513671875, 0.7279165649414062, 0.7284940795898438, 0.7287070922851563, 0.7288955078125, 0.7288780517578125, 0.7277035522460937, 0.72800048828125, 0.72765234375, 0.728056884765625, 0.7274495849609375, 0.7280240478515625, 0.727583740234375, 0.727520263671875, 0.7282698364257812, 0.7273543701171875, 0.7287091064453125, 0.7278960571289063, 0.7277035522460937, 0.7276973876953124, 0.727141357421875]",tokens/s,1.3533475452727455,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-564M,facebook/xglm-564M,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1469, in _autoset_attn_implementation + cls._check_and_enable_flash_attn_2( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1560, in _check_and_enable_flash_attn_2 + raise ValueError( +ValueError: XGLMForCausalLM does not support Flash Attention 2.0 yet. Please request to add support where the model is hosted, on its model hub page: https://huggingface.co//tmp/tmpqd8xlz04/no_weights_model/discussions/new or in the Transformers GitHub repo: https://github.com/huggingface/transformers/issues/new + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-350m,facebook/opt-350m,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,1312.935936,1023.934464,0.0,377.48736,290.348032,s,10,0.6976366119384766,0.06976366119384766,0.0018590846735664541,0.0696058578491211,0.07176354598999024,0.07230553245544434,0.07273912162780762,"[0.07284751892089844, 0.06800761413574219, 0.06769139099121094, 0.06779033660888673, 0.07104889678955079, 0.07164310455322266, 0.07125389099121093, 0.06824649810791016, 0.07096521759033203, 0.06814214324951172]",tokens/s,3669.5321836488733,kWh,8.134236630148721e-07,4.4571909736688417e-07,2.1182893658638067e-06,3.377432126245563e-06,tokens/kWh,75797230.09402885,MB,1312.935936,1023.934464,0.0,377.48736,337.281536,s,10,43.73100244140625,4.373100244140625,0.07142248616314475,4.400765380859375,4.4402288085937505,4.445402490234375,4.449541435546875,"[4.35173583984375, 4.25024462890625, 4.26398681640625, 4.30877587890625, 4.41468994140625, 4.4390791015625, 4.43881396484375, 4.42625927734375, 4.3868408203125, 4.450576171875]",tokens/s,14.40625562709468,kWh,5.082961926562783e-05,2.7857636841584544e-05,0.0001254377571233551,0.00020412501323056748,tokens/kWh,308634.39518232364,,s,629,44.28504171752928,0.07040547172898139,0.008395184351578173,0.07026687622070313,0.07091015625,0.07132590332031251,0.13567237548828126,"[0.0723609619140625, 0.07140863800048829, 0.07176195526123047, 0.07163388824462891, 0.07094169616699218, 0.07094783782958984, 0.07148441314697265, 0.07121202850341797, 0.07107174682617187, 0.07111577606201172, 0.07114342498779297, 0.07076761627197266, 0.07108198547363281, 0.07081574249267578, 0.07077072143554687, 0.07092835235595703, 0.07097344207763671, 0.0714598388671875, 0.07095500946044922, 0.07107481384277343, 0.07127347564697266, 0.07087923431396484, 0.07049420928955077, 0.07076659393310547, 0.07170150756835937, 0.071552001953125, 0.0690360336303711, 0.06716006469726563, 0.06747545623779297, 0.0673433609008789, 0.06707405090332032, 0.06747135925292969, 0.06733312225341796, 0.06719999694824219, 0.06728806304931641, 0.0673095703125, 0.06710784149169922, 0.06706175994873047, 0.0676126708984375, 0.06733926391601562, 0.06688256072998047, 0.06733414459228515, 0.06721126556396484, 0.0674703369140625, 0.06733926391601562, 0.06750822448730469, 0.06778880310058594, 0.06814105224609375, 0.0675225601196289, 0.06693376159667969, 0.06731980895996094, 0.06751334381103516, 0.07044198608398437, 0.0684615707397461, 0.0671098861694336, 0.06701465606689454, 0.06718463897705078, 0.06769561767578125, 0.06716422271728516, 0.06652819061279297, 0.06804582214355469, 0.06770995330810547, 0.13573631286621093, 0.0672204818725586, 0.06833356475830078, 0.06807555389404298, 0.06771196746826172, 0.06805197143554688, 0.06715392303466797, 0.06708428955078125, 0.06746112060546874, 0.06681702423095703, 0.06697062683105469, 0.06744166564941406, 0.06732288360595703, 0.06716006469726563, 0.06725325012207031, 0.06715602874755859, 0.06704019165039063, 0.06741094207763672, 0.06709555053710937, 0.06650572967529297, 0.06745292663574219, 0.06664704132080078, 0.06701261138916016, 0.06722150421142578, 0.06713855743408204, 0.0671098861694336, 0.06744268798828125, 0.06728704071044922, 0.0696258544921875, 0.06788813018798828, 0.06791372680664062, 0.06849024200439453, 0.06736895751953124, 0.06989209747314454, 0.06821990203857421, 0.06771097564697266, 0.06738432312011719, 0.06751948547363282, 0.06726348876953125, 0.06710784149169922, 0.06749183654785157, 0.06747545623779297, 0.06864179229736328, 0.0675563507080078, 0.06683135986328125, 0.06710681915283204, 0.06729011535644532, 0.06674534606933594, 0.06728089904785156, 0.06722354888916016, 0.06724607849121093, 0.06693990325927734, 0.0674703369140625, 0.06723481750488282, 0.06712223815917968, 0.06780818939208984, 0.06794751739501953, 0.06755840301513671, 0.0675758056640625, 0.06719692993164063, 0.0672573471069336, 0.06723481750488282, 0.06736589050292968, 0.13550796508789062, 0.06692147064208985, 0.06734950256347656, 0.07442227172851562, 0.0679557113647461, 0.06727680206298828, 0.06684569549560547, 0.06775091552734375, 0.06815436553955079, 0.06903807830810547, 0.06718975830078125, 0.06693376159667969, 0.06734031677246094, 0.06811030578613281, 0.06972621154785157, 0.06792601776123047, 0.06736486053466798, 0.06915481567382813, 0.06836736297607422, 0.06893875122070313, 0.07033753967285156, 0.06752665710449218, 0.0672194595336914, 0.06725325012207031, 0.06720822143554687, 0.06680572509765625, 0.06687026977539062, 0.0672143325805664, 0.06730445098876953, 0.06715904235839844, 0.06759014129638671, 0.06702899169921875, 0.06747443389892578, 0.06825984191894531, 0.0673280029296875, 0.06705152130126953, 0.06734130859375, 0.06769664001464844, 0.06806425476074218, 0.06706380462646484, 0.06730137634277343, 0.0671825942993164, 0.0676157455444336, 0.06711398315429687, 0.06727577972412109, 0.06933606719970703, 0.06894182586669922, 0.06846259307861328, 0.06729523468017579, 0.06725836944580078, 0.06733312225341796, 0.0672511978149414, 0.0672890853881836, 0.06693376159667969, 0.0676341781616211, 0.06666957092285156, 0.06659993743896485, 0.0665722885131836, 0.0667514877319336, 0.0672368621826172, 0.06741712188720703, 0.06712726593017578, 0.06705049896240234, 0.1353912353515625, 0.0671488037109375, 0.06749900817871093, 0.0674703369140625, 0.06737715148925781, 0.0673064956665039, 0.06735257720947266, 0.06739974212646484, 0.0676075210571289, 0.07006716918945312, 0.06824960327148437, 0.06842777252197266, 0.06737100982666015, 0.06794239807128906, 0.06855276489257812, 0.0674927978515625, 0.06724813079833984, 0.06728294372558594, 0.06814617919921875, 0.06813286590576172, 0.06723583984375, 0.06725222778320313, 0.06753689575195312, 0.06841139221191406, 0.06750617980957031, 0.06819737243652343, 0.06738438415527344, 0.06882195281982421, 0.06757997131347657, 0.06839907073974609, 0.06708633422851562, 0.06791574096679688, 0.06755020904541016, 0.06742527770996094, 0.06748569488525391, 0.066587646484375, 0.06756761932373047, 0.06675154876708984, 0.06733510589599609, 0.06650163269042969, 0.06747647857666016, 0.06739250946044922, 0.06733824157714843, 0.07072870635986328, 0.07021260833740234, 0.07095807647705078, 0.07044915008544922, 0.07039590454101563, 0.07007437133789063, 0.06753177642822265, 0.06737203216552734, 0.07054438018798828, 0.06973747253417968, 0.07040415954589843, 0.07051052856445313, 0.0710830078125, 0.0708136978149414, 0.06919782257080079, 0.06926131439208984, 0.06997299194335938, 0.07050035095214843, 0.0701286392211914, 0.07048499298095703, 0.14131712341308594, 0.06722457885742188, 0.06749183654785157, 0.06760550689697266, 0.07033446502685547, 0.07015731048583984, 0.07011225891113282, 0.06978765106201172, 0.0705269775390625, 0.06983372497558593, 0.06974156951904296, 0.07031603240966797, 0.07043583679199218, 0.0705802230834961, 0.07040614318847656, 0.07013069152832031, 0.0699504623413086, 0.07051058959960937, 0.07024332427978516, 0.06977126312255859, 0.06987673950195313, 0.07046553802490234, 0.07214806365966797, 0.07040809631347657, 0.06932179260253907, 0.07022994995117188, 0.07043071746826172, 0.07018402862548828, 0.07030262756347656, 0.07083622741699219, 0.07088333129882812, 0.06744166564941406, 0.06735155487060547, 0.06759731292724609, 0.06742733001708984, 0.0693780517578125, 0.07071334075927735, 0.07068876647949218, 0.07128985595703125, 0.07055465698242187, 0.07048291015625, 0.07026483154296875, 0.07063346862792969, 0.07024947357177734, 0.0704563217163086, 0.07028530883789062, 0.0700212173461914, 0.07084226989746094, 0.0705269775390625, 0.07041228485107422, 0.07060889434814453, 0.07039590454101563, 0.07050342559814453, 0.0704686050415039, 0.07056486511230468, 0.07071641540527343, 0.07057920074462891, 0.07043276977539062, 0.07038159942626954, 0.07048802947998047, 0.06970470428466796, 0.07084134674072265, 0.07183257293701172, 0.14264422607421876, 0.07059661102294922, 0.0704337921142578, 0.07088127899169921, 0.07030681610107421, 0.07067340850830078, 0.07019110107421875, 0.06985215759277344, 0.07026483154296875, 0.07035391998291016, 0.0703815689086914, 0.07054847717285156, 0.07035903930664063, 0.07019519805908203, 0.0714567642211914, 0.0703662109375, 0.07029964447021485, 0.07025766754150391, 0.07012351989746093, 0.07053107452392578, 0.07036518096923829, 0.07038873291015625, 0.07040204620361327, 0.07045836639404297, 0.07014093017578125, 0.07181926727294922, 0.07061196899414063, 0.07031705474853515, 0.07053209686279296, 0.07042457580566407, 0.07070719909667969, 0.07003033447265625, 0.0699525146484375, 0.07056486511230468, 0.07099187469482422, 0.07034060668945312, 0.07056281280517578, 0.07041228485107422, 0.070940673828125, 0.07032319641113281, 0.0704901123046875, 0.07044915008544922, 0.07032627105712891, 0.07004057312011719, 0.07051776123046875, 0.07030169677734376, 0.07029043579101563, 0.07063859558105469, 0.07095500946044922, 0.07047993469238281, 0.0705125732421875, 0.07045734405517579, 0.07045222473144531, 0.07037542724609375, 0.07044096374511719, 0.0705638427734375, 0.07061196899414063, 0.07037747192382812, 0.07041024017333984, 0.06980403137207031, 0.07040614318847656, 0.07026585388183594, 0.07064780426025391, 0.14216192626953125, 0.07040102386474609, 0.07040819549560547, 0.07026483154296875, 0.07061504364013672, 0.07052496337890625, 0.07054332733154296, 0.07030989074707031, 0.07011020660400391, 0.07049727630615234, 0.07108812713623047, 0.07056179046630859, 0.07050444793701172, 0.07025971221923828, 0.07032217407226563, 0.07038566589355469, 0.07023513793945313, 0.07034368133544922, 0.07049625396728515, 0.07012556457519531, 0.07023721313476562, 0.06932579040527344, 0.07020543670654297, 0.07085977935791016, 0.07015219116210937, 0.07016242980957031, 0.07007129669189453, 0.07072870635986328, 0.07037542724609375, 0.07045836639404297, 0.07033446502685547, 0.07036313629150391, 0.07039180755615235, 0.07009587097167969, 0.07063346862792969, 0.0701839370727539, 0.07108812713623047, 0.07090892791748046, 0.07049727630615234, 0.07038259124755859, 0.07044403076171875, 0.0706519012451172, 0.070761474609375, 0.0694466552734375, 0.0697528305053711, 0.07026380920410157, 0.07064268493652344, 0.0705433578491211, 0.0703477783203125, 0.07133491516113281, 0.0704368667602539, 0.07131238555908204, 0.07170355224609375, 0.0705771484375, 0.07050752258300781, 0.07046451568603515, 0.07079840087890625, 0.07022892761230469, 0.07035084533691406, 0.07033344268798829, 0.0702402572631836, 0.07078604888916015, 0.07056896209716797, 0.13596263122558594, 0.06696959686279297, 0.06721331024169921, 0.06699417877197265, 0.06940467071533203, 0.07078604888916015, 0.07069593811035156, 0.07035187530517578, 0.07053107452392578, 0.07046246337890626, 0.0706344985961914, 0.07024642944335938, 0.07162467193603515, 0.07080652618408204, 0.07045529937744141, 0.07055359649658204, 0.0706519012451172, 0.07040102386474609, 0.06990335845947265, 0.07026080322265625, 0.07056377410888671, 0.07085670471191406, 0.06976921844482421, 0.07021260833740234, 0.06979583740234375, 0.07011634826660157, 0.0700426254272461, 0.07040716552734375, 0.07058329772949219, 0.0699658203125, 0.07026892852783204, 0.0703272933959961, 0.07030989074707031, 0.07069593811035156, 0.07077683258056641, 0.07038259124755859, 0.07024230194091798, 0.07033548736572266, 0.07012454223632812, 0.07299378967285156, 0.07067033386230469, 0.07059865570068359, 0.07034368133544922, 0.07035699462890625, 0.07052902221679687, 0.070687744140625, 0.06977433776855468, 0.07029862213134766, 0.0704901123046875, 0.07047270202636718, 0.07048191833496094, 0.07052799987792968, 0.0706170883178711, 0.07022182464599609, 0.0703272933959961, 0.070255615234375, 0.07052082824707032, 0.07059967803955078, 0.07068160247802735, 0.0705054702758789, 0.07004364776611328, 0.07048089599609375, 0.07136255645751953, 0.14220700073242187, 0.07030985260009766, 0.07026892852783204, 0.0702033920288086, 0.07120588684082031, 0.07035289764404297, 0.07038361358642578, 0.07045222473144531, 0.070181884765625, 0.07042969512939454, 0.07047577667236328, 0.07076150512695313, 0.07077168273925781, 0.07013990020751953, 0.07032422637939453, 0.07050752258300781, 0.07027200317382812, 0.07128268432617188, 0.070614013671875, 0.07047475433349609, 0.07060889434814453, 0.07019827270507813, 0.07051264190673828, 0.07065702056884765, 0.07134515380859376, 0.07106253051757813, 0.07054950714111329, 0.07042253112792969, 0.07038873291015625, 0.0702740478515625, 0.07026687622070313, 0.06861004638671875, 0.0671272964477539, 0.06711507415771484, 0.0671866226196289, 0.06729523468017579, 0.06710169219970703, 0.06714166259765625, 0.06752764892578125, 0.06734130859375, 0.06720921325683593, 0.06747135925292969, 0.06729933166503907, 0.0674150390625, 0.06712422180175781, 0.06744882965087891, 0.06737612915039062, 0.06713958740234376, 0.06726656341552735, 0.0685823974609375, 0.07029043579101563, 0.07070515441894532, 0.0720742416381836, 0.07064166259765625, 0.07065087890625, 0.07042559814453125, 0.07116806030273437, 0.07063648223876953, 0.07048703765869141, 0.07075430297851562, 0.07056896209716797, 0.07044608306884766, 0.07033036804199219, 0.13839053344726562, 0.07068978881835937, 0.07101030731201172, 0.07077279663085938, 0.07043782043457031, 0.070614013671875, 0.07037542724609375, 0.07067443084716797, 0.070181884765625, 0.07043788909912109, 0.07035391998291016, 0.0702003173828125, 0.07099903869628907, 0.07056690979003906, 0.07074406433105469, 0.07023308563232422, 0.07023616027832032, 0.0702208023071289, 0.070150146484375, 0.0707799072265625, 0.07061199951171875, 0.0703927993774414, 0.07016448211669922, 0.07032115173339844, 0.07010406494140625, 0.07089151763916016, 0.07061913299560547, 0.07181926727294922, 0.07085977935791016, 0.07089663696289063, 0.07025459289550781, 0.07071952056884766, 0.07063139343261719, 0.07043276977539062, 0.07035391998291016, 0.07057920074462891, 0.07042867279052735, 0.07073075103759766, 0.07066726684570312, 0.07058124542236328, 0.07049215698242188, 0.07038668823242188, 0.07035903930664063, 0.07047987365722656, 0.07088025665283203, 0.07053414154052734, 0.07050342559814453, 0.07065395355224609, 0.07077375793457032, 0.07083213043212891, 0.070761474609375, 0.07080242919921875, 0.07068364715576173, 0.0717496337890625, 0.07135539245605468, 0.07119155120849609, 0.07112703704833985, 0.07091506958007812, 0.07129190063476562, 0.07068876647949218, 0.07067545318603516, 0.07053619384765625, 0.07090688323974609]",tokens/s,14.20344151445213,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-6.7b,facebook/opt-6.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,2034.978816,5539.10272,0.0,4892.655616,4542.610432,s,10,5.622600402832031,0.5622600402832031,0.0013230028627895694,0.5620163879394531,0.5633294921875001,0.564493603515625,0.565424892578125,"[0.56565771484375, 0.5620866088867188, 0.5614910888671875, 0.5609406127929687, 0.5618070678710938, 0.5607069091796875, 0.5619461669921875, 0.5621692504882813, 0.5627241821289063, 0.56307080078125]",tokens/s,455.30534211724546,kWh,6.621943920114895e-06,3.6285156736691226e-06,3.075088879823785e-05,4.100134839202187e-05,tokens/kWh,6243697.098747441,MB,2035.28192,5539.10272,0.0,4892.655616,4726.279168,s,10,331.3280078125,33.13280078125,0.008409894982587748,33.12982421875,33.142355468750004,33.145203125,33.14748125,"[33.14805078125, 33.124625, 33.1231640625, 33.1260859375, 33.14172265625, 33.1400703125, 33.12555859375, 33.12609375, 33.13908203125, 33.1335546875]",tokens/s,1.9014390125344305,kWh,0.00039121345305516403,0.00021441890828907464,0.0018001254370127661,0.0024057577983570048,tokens/kWh,26187.174803309546,,s,629,335.85907128906234,0.5339571880589229,0.06674281532897068,0.5258792724609375,0.5263243408203125,0.5265016845703125,1.0872363134765626,"[0.5258577880859375, 0.5256417236328125, 0.5264219970703125, 0.5263196411132812, 0.5257113647460937, 0.5258464965820312, 0.525486083984375, 0.5262151489257813, 0.5262940063476562, 0.52614453125, 0.5258035278320312, 0.52607080078125, 0.5257379760742188, 0.5260421142578126, 0.5259489135742188, 0.526060546875, 0.5258331909179688, 0.5260062866210937, 0.5258424072265625, 0.52600732421875, 0.5258045654296875, 0.5261404418945312, 0.5261957397460938, 0.5260748901367187, 0.5259468994140625, 0.5261731567382812, 0.5258363037109375, 0.5262284545898438, 0.5258250122070313, 0.5260421142578126, 0.526023681640625, 0.526244873046875, 0.52615576171875, 0.5261332397460937, 0.52611376953125, 0.5259898681640625, 0.526482421875, 0.526271484375, 0.5264271240234375, 0.5263646850585938, 0.5265366821289063, 0.52632373046875, 0.5263370361328125, 0.5259735107421875, 0.5261414184570312, 0.526107666015625, 0.5262571411132813, 0.5263308715820313, 0.5263267822265625, 0.5261199340820313, 0.52626025390625, 0.5261434936523437, 0.526482421875, 0.5264199829101562, 0.5262622680664063, 0.5264076538085938, 0.5264937133789063, 0.5264998168945313, 0.5264230346679688, 0.5268643798828125, 0.5268623657226562, 0.526497802734375, 1.0878065185546875, 0.5256908569335937, 0.5257564086914063, 0.5259059448242187, 0.5258055419921875, 0.5255567626953125, 0.52566015625, 0.5258875122070312, 0.526002197265625, 0.5259949951171875, 0.5256663208007812, 0.5261107177734375, 0.5259857788085938, 0.5258189086914062, 0.525576171875, 0.525970458984375, 0.5256294555664063, 0.5262294921875, 0.5257728271484375, 0.5257666625976563, 0.525465576171875, 0.5258946533203125, 0.5259468994140625, 0.5260144653320312, 0.5255526123046875, 0.5257932739257812, 0.5255004272460938, 0.5259530029296875, 0.52562841796875, 0.5257584838867188, 0.5257738037109375, 0.525844482421875, 0.5255690307617188, 0.5256724243164063, 0.5258956909179687, 0.5257461547851563, 0.5255751953125, 0.5257482299804688, 0.5255536499023438, 0.5257677001953125, 0.5256314697265625, 0.5257533569335937, 0.5254666137695313, 0.52566015625, 0.5253980102539062, 0.5257482299804688, 0.5259049072265625, 0.5260809936523437, 0.5258383178710937, 0.5262018432617187, 0.5256294555664063, 0.525750244140625, 0.525849609375, 0.5261281127929688, 0.5256796264648438, 0.5259478759765625, 0.5256181640625, 0.5258219604492187, 0.5256539916992188, 0.525886474609375, 0.5259243774414063, 0.5260851440429688, 0.5258331909179688, 1.087382568359375, 0.526060546875, 0.5257789306640624, 0.5258936157226562, 0.5258137817382813, 0.5256263427734374, 0.525675537109375, 0.5257164916992187, 0.525739013671875, 0.525718505859375, 0.5260646362304687, 0.5256406860351562, 0.5256908569335937, 0.5256744995117187, 0.525433837890625, 0.525717529296875, 0.5254287109375, 0.5256908569335937, 0.5254031372070312, 0.5255137329101562, 0.5256263427734374, 0.525770751953125, 0.5254891357421875, 0.5259683837890625, 0.5255454711914063, 0.5256990966796875, 0.5255372924804688, 0.5261035766601563, 0.5256632080078125, 0.5258690795898437, 0.5257625732421874, 0.525728759765625, 0.5260155029296875, 0.5258189086914062, 0.5256539916992188, 0.5256417236328125, 0.5256273803710938, 0.5258055419921875, 0.5257083129882812, 0.525971435546875, 0.5258916015625, 0.52600830078125, 0.5256673583984375, 0.5257686767578125, 0.5254942626953125, 0.52560791015625, 0.5257482299804688, 0.5258803100585937, 0.5259755249023438, 0.5260676879882813, 0.5256151123046875, 0.5258536987304687, 0.5258526611328125, 0.5258741455078125, 0.5255966796875, 0.5259386596679687, 0.5257666625976563, 0.5259059448242187, 0.5259223022460937, 0.5260523681640625, 0.52583935546875, 0.5260933227539063, 0.5256048583984375, 1.0868602294921874, 0.5259120483398437, 0.5255465087890625, 0.5258884887695312, 0.5255741577148437, 0.525613037109375, 0.5256007690429687, 0.5258076171875, 0.5256734619140625, 0.5257328491210937, 0.5254573974609374, 0.525791259765625, 0.525454345703125, 0.5258527221679687, 0.5255577392578125, 0.525912109375, 0.5255772094726563, 0.525727783203125, 0.5255833129882812, 0.5263431396484375, 0.5257984008789063, 0.525792236328125, 0.525896728515625, 0.52569189453125, 0.5255977172851563, 0.52600732421875, 0.525549560546875, 0.5257738037109375, 0.5256058959960938, 0.5258147583007813, 0.52556494140625, 0.5260534057617188, 0.5259837646484375, 0.5260646362304687, 0.5258803100585937, 0.5260369873046875, 0.5255034790039063, 0.5258485717773438, 0.5255608520507813, 0.52600830078125, 0.52562841796875, 0.52608203125, 0.5257349243164062, 0.5257574462890625, 0.5256406860351562, 0.5257717895507813, 0.5257625732421874, 0.5258875122070312, 0.525686767578125, 0.5260492553710937, 0.5258639526367187, 0.5261209716796875, 0.5259990844726562, 0.5260728149414062, 0.5258884887695312, 0.5258946533203125, 0.52564990234375, 0.5258731689453126, 0.5256837158203125, 0.5259202270507812, 0.5259827270507812, 0.5264097290039063, 0.526213134765625, 1.0886195068359374, 0.5266964721679688, 0.5265223388671875, 0.52691455078125, 0.5266104125976563, 0.5267548217773438, 0.5260175170898438, 0.5263544311523437, 0.5261270751953125, 0.5263964233398437, 0.5262694702148437, 0.5260482788085937, 0.5259909057617187, 0.5260492553710937, 0.5259929809570313, 0.5264066772460938, 0.525787109375, 0.5257963256835938, 0.5255126953125, 0.5257431030273437, 0.5255669555664062, 0.5256345825195312, 0.5254779052734375, 0.5258198852539062, 0.5258219604492187, 0.5260687255859375, 0.5260800170898438, 0.525802490234375, 0.5258506469726563, 0.525781005859375, 0.5256857299804687, 0.5258485717773438, 0.52577587890625, 0.5258373413085937, 0.5261895751953125, 0.52682958984375, 0.526614501953125, 0.52650390625, 0.5255485229492187, 0.5258424072265625, 0.525549560546875, 0.5258352661132812, 0.5255639038085937, 0.525822998046875, 0.5257195434570312, 0.525739013671875, 0.5257267456054687, 0.5263533935546875, 0.5261547241210938, 0.5263810424804688, 0.5262264404296875, 0.526551025390625, 0.5261895751953125, 0.5266165771484375, 0.5261998291015625, 0.5261404418945312, 0.5261486206054687, 0.5264189453125, 0.52577587890625, 0.5259304809570312, 0.5256898803710938, 0.5258782958984375, 0.5257636108398438, 1.08657666015625, 0.5256980590820313, 0.52586083984375, 0.5259304809570312, 0.5256939697265625, 0.5258782958984375, 0.5258782958984375, 0.52608203125, 0.5260462036132812, 0.5260513305664063, 0.525549560546875, 0.5258875122070312, 0.5257984008789063, 0.5262018432617187, 0.5257328491210937, 0.5260298461914062, 0.5255608520507813, 0.5259683837890625, 0.52613427734375, 0.52609228515625, 0.5256028442382813, 0.52607080078125, 0.5256304931640625, 0.5257891845703125, 0.525643798828125, 0.5260062866210937, 0.5256611938476563, 0.5261844482421875, 0.5262510375976562, 0.5261025390625, 0.5262673950195312, 0.5262622680664063, 0.5260779418945313, 0.5261721801757813, 0.5261353149414062, 0.5261588745117187, 0.5259642944335937, 0.52661865234375, 0.52625, 0.5263790283203125, 0.5262643432617188, 0.5265131225585937, 0.5259069213867188, 0.526286865234375, 0.5260626220703125, 0.5262888793945313, 0.52619775390625, 0.526224365234375, 0.5260155029296875, 0.526613525390625, 0.526581787109375, 0.5259366455078125, 0.5256642456054688, 0.5257686767578125, 0.5256632080078125, 0.5258198852539062, 0.5258588256835938, 0.5261209716796875, 0.525928466796875, 0.5265029296875, 0.5259878540039062, 0.5262069702148438, 0.5264834594726563, 1.088301025390625, 0.5262653198242188, 0.5256724243164063, 0.5255659790039062, 0.525475830078125, 0.5257636108398438, 0.5255300903320312, 0.5257246704101562, 0.5256058959960938, 0.5256345825195312, 0.5254676513671875, 0.5258137817382813, 0.5256581420898437, 0.5258731689453126, 0.526055419921875, 0.526497802734375, 0.5257297973632813, 0.5260103759765625, 0.5256406860351562, 0.5259428100585938, 0.5259376831054687, 0.5261486206054687, 0.5255321655273437, 0.5259735107421875, 0.5258168334960938, 0.52605029296875, 0.5256571044921875, 0.5257420654296875, 0.5257471923828125, 0.525717529296875, 0.525749267578125, 0.5258168334960938, 0.5255885009765625, 0.5259366455078125, 0.5256325073242187, 0.5258721313476562, 0.5255454711914063, 0.5259120483398437, 0.5255874633789063, 0.5258823852539063, 0.5260584716796874, 0.5260093383789063, 0.525638671875, 0.5257984008789063, 0.5254819946289062, 0.5257584838867188, 0.5255413818359375, 0.5256263427734374, 0.5254993896484375, 0.5260482788085937, 0.5255536499023438, 0.5256744995117187, 0.5257471923828125, 0.5260912475585937, 0.5257963256835938, 0.526045166015625, 0.5258997802734375, 0.5258854370117187, 0.5258956909179687, 0.526097412109375, 0.5258956909179687, 0.52596630859375, 0.525486083984375, 1.0877613525390626, 0.525769775390625, 0.5254829711914063, 0.5259612426757813, 0.525929443359375, 0.5262827758789063, 0.5257748413085938, 0.5258516235351562, 0.5254768676757813, 0.5256775512695312, 0.5254871215820313, 0.5258475341796875, 0.525770751953125, 0.5260123901367187, 0.5260103759765625, 0.52575537109375, 0.525591552734375, 0.5260431518554688, 0.5255874633789063, 0.5257769165039062, 0.52560693359375, 0.5258577880859375, 0.5256089477539062, 0.5257195434570312, 0.52571337890625, 0.5260534057617188, 0.5259540405273437, 0.526023681640625, 0.5256468505859375, 0.5257000732421875, 0.5256878051757813, 0.5258270874023437, 0.5255669555664062, 0.5259325561523438, 0.5256478881835938, 0.5259059448242187, 0.5256539916992188, 0.5259171752929688, 0.525533203125, 0.5258956909179687, 0.5256396484375, 0.5259151611328124, 0.5259089965820313, 0.5258045654296875, 0.525470703125, 0.526087158203125, 0.525570068359375, 0.526012451171875, 0.5258239135742188, 0.526266357421875, 0.5257000732421875, 0.5258916015625, 0.5262540893554688, 0.525970458984375, 0.5258168334960938, 0.5259356079101563, 0.5257216186523438, 0.5257953491210937, 0.5261250610351562, 0.5259458618164062, 0.5255300903320312, 0.5259765625, 0.5258004760742188, 1.0886134033203125, 0.52569189453125, 0.5255536499023438, 0.5258383178710937, 0.5257799682617188, 0.5257584838867188, 0.5256837158203125, 0.5262305297851563, 0.5256099853515624, 0.525781005859375, 0.5258270874023437, 0.5262387084960938, 0.5258424072265625, 0.5259765625, 0.5260114135742188, 0.5261588745117187, 0.5259622192382812, 0.5260318603515625, 0.5257083129882812, 0.526166015625, 0.5256325073242187, 0.52587109375, 0.5258260498046875, 0.5261178588867188, 0.5256458129882813, 0.525897705078125, 0.5261752319335937, 0.5257431030273437, 0.5257799682617188, 0.526298095703125, 0.5260574951171875, 0.52619775390625, 0.52611279296875, 0.526271484375, 0.5262151489257813, 0.5269708862304687, 0.5263012084960937, 0.52657666015625, 0.5264937133789063, 0.5267752685546875, 0.5261527099609375, 0.5262315673828125, 0.5261690673828125, 0.5258997802734375, 0.5257728271484375, 0.5265807495117187, 0.525681640625, 0.5258721313476562, 0.5259049072265625, 0.525970458984375, 0.525633544921875, 0.5261414184570312, 0.5260697631835938, 0.5258782958984375, 0.5258792724609375, 0.52611376953125, 0.5257523193359375, 0.5261752319335937, 0.5259192504882813, 0.52607080078125, 0.5259898681640625, 0.5263278198242187, 0.5262356567382812, 1.089292236328125, 0.5257205810546876, 0.5256959838867188, 0.5258516235351562, 0.525675537109375, 0.525707275390625, 0.525644775390625, 0.5258741455078125, 0.5256427612304687, 0.52611376953125, 0.5256857299804687, 0.5256929321289062, 0.525865966796875, 0.5259274291992188, 0.5254942626953125, 0.52583935546875, 0.5259089965820313, 0.5260984497070312, 0.5255659790039062, 0.52596630859375, 0.5255341796875, 0.5257932739257812, 0.5256161499023437, 0.5257615356445312, 0.5255608520507813, 0.5256908569335937, 0.5256058959960938, 0.5258137817382813, 0.525822998046875, 0.5260390625, 0.5258168334960938, 0.526033935546875, 0.5262326049804688, 0.5259100341796875, 0.5258065795898438, 0.52625, 0.5258567504882813, 0.52642919921875, 0.5261015014648438, 0.525955078125, 0.5256345825195312, 0.5259990844726562, 0.5259017944335938, 0.525970458984375, 0.5258311767578125, 0.5259765625, 0.5258895263671876, 0.5263216552734375, 0.5260062866210937, 0.5263206176757812, 0.5261220092773438, 0.5258639526367187, 0.5263790283203125, 0.52634521484375, 0.52650390625, 0.5263104248046875, 0.5256294555664063, 0.5261782836914063, 0.5257164916992187, 0.526087158203125, 0.526256103515625, 0.52645068359375, 0.5261045532226563]",tokens/s,1.8728093232254581,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-72B,Qwen/Qwen-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 550, in from_pretrained + model_class = get_class_from_dynamic_module( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 498, in get_class_from_dynamic_module + final_module = get_cached_module_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 323, in get_cached_module_file + modules_needed = check_imports(resolved_module_file) + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 181, in check_imports + raise ImportError( +ImportError: This modeling file requires the following packages that were not found in your environment: transformers_stream_generator. Run `pip install transformers_stream_generator` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-6B,,cuda,0,42,,,,,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,,MB,1608.957952,5448.925184,0.0,4802.47808,4489.12128,s,10,5.066485656738281,0.5066485656738281,0.0011948895817140703,0.5068220977783202,0.5079840667724609,0.5080279983520508,0.5080631436157227,"[0.5075723266601563, 0.5061219482421875, 0.504515380859375, 0.5053372497558594, 0.5061441345214843, 0.5055963745117188, 0.5075000610351562, 0.5076519470214844, 0.5080719299316406, 0.5079743041992187]",tokens/s,505.2812093912223,kWh,5.964234835571714e-06,3.2679267622643234e-06,2.7223952334698433e-05,3.6456113932534466e-05,tokens/kWh,7022141.758547072,MB,1610.637312,5448.925184,0.0,4802.47808,4557.793792,s,10,299.1010234375,29.910102343749998,0.01023537902845777,29.913578125,29.921135546875,29.9219037109375,29.9225182421875,"[29.899591796875, 29.901322265625, 29.913013671875, 29.890712890625, 29.920396484375, 29.922671875, 29.92096484375, 29.9152421875, 29.90296484375, 29.914142578125]",tokens/s,2.106311749654192,kWh,0.00035296389131082433,0.00019345485010443554,0.0015722178549955057,0.0021186365964107655,tokens/kWh,29736.104864198915,,s,629,303.18425177001956,0.4820099392210167,0.06023141623810366,0.4746004333496094,0.4757639221191406,0.4761423828125,0.98077423828125,"[0.474365966796875, 0.47431475830078124, 0.4748533630371094, 0.47396148681640626, 0.47435162353515625, 0.47414068603515624, 0.47463320922851565, 0.47518923950195313, 0.47532952880859375, 0.4745502624511719, 0.4740556945800781, 0.47391131591796876, 0.47444989013671873, 0.47430245971679685, 0.4739420166015625, 0.4737402954101563, 0.473987060546875, 0.47486770629882813, 0.4739993591308594, 0.475114501953125, 0.47402597045898437, 0.4752353210449219, 0.47403109741210936, 0.4739563598632813, 0.4750960693359375, 0.4755916748046875, 0.4742686767578125, 0.47484210205078126, 0.4754851989746094, 0.4751022033691406, 0.4755619812011719, 0.4745973815917969, 0.47523736572265624, 0.474829833984375, 0.474850341796875, 0.47421026611328126, 0.4759234619140625, 0.47534796142578123, 0.4750888977050781, 0.4741591186523437, 0.4745441284179687, 0.473744384765625, 0.4741099548339844, 0.473807861328125, 0.47429119873046877, 0.4740597839355469, 0.4742881164550781, 0.47400347900390627, 0.4744253234863281, 0.47619073486328123, 0.4757626953125, 0.4742860717773438, 0.47489535522460935, 0.47489950561523436, 0.47460345458984377, 0.4742686767578125, 0.4739921875, 0.4739420166015625, 0.47433624267578123, 0.4748810119628906, 0.4744110107421875, 0.47424615478515625, 0.980806640625, 0.4736993408203125, 0.47429733276367186, 0.4741478271484375, 0.4741099548339844, 0.47420416259765624, 0.4742000732421875, 0.47553228759765626, 0.4755077209472656, 0.4747817077636719, 0.47485235595703124, 0.4747202453613281, 0.4752384033203125, 0.4742318115234375, 0.4740137023925781, 0.4751011962890625, 0.47453387451171875, 0.4759981994628906, 0.4747898864746094, 0.4742266845703125, 0.47406185913085935, 0.4743239440917969, 0.4744366149902344, 0.47420416259765624, 0.47499981689453125, 0.47433114624023437, 0.47446322631835935, 0.47578317260742187, 0.47489739990234375, 0.4754503784179688, 0.4746977233886719, 0.47437005615234373, 0.4743680114746094, 0.47590194702148436, 0.47595211791992187, 0.47569818115234375, 0.474777587890625, 0.474113037109375, 0.4756346740722656, 0.4741652526855469, 0.47452056884765625, 0.4742778930664063, 0.474176513671875, 0.47388363647460935, 0.473754638671875, 0.47433624267578123, 0.474029052734375, 0.47412841796875, 0.4740433654785156, 0.47558758544921875, 0.4756910400390625, 0.4747693786621094, 0.47467111206054685, 0.4742625427246094, 0.4742799377441406, 0.473807861328125, 0.474387451171875, 0.47398092651367185, 0.4740495300292969, 0.47483187866210935, 0.4742236022949219, 0.4746342468261719, 0.4739717102050781, 0.98069091796875, 0.4739993591308594, 0.4743341979980469, 0.4740321350097656, 0.4741048278808594, 0.47378021240234375, 0.4756019287109375, 0.4752998352050781, 0.47476327514648436, 0.47476531982421877, 0.47458303833007814, 0.474603515625, 0.4745164794921875, 0.47447552490234374, 0.4745482177734375, 0.47400344848632814, 0.4754913330078125, 0.47445709228515626, 0.4743076171875, 0.4742041320800781, 0.47417855834960937, 0.4741171264648438, 0.47425332641601564, 0.47636785888671873, 0.4741949462890625, 0.4747110290527344, 0.4743659973144531, 0.47549435424804687, 0.47543499755859375, 0.4749854736328125, 0.4751646728515625, 0.4754985046386719, 0.4756643981933594, 0.475114501953125, 0.4750315551757813, 0.47576882934570314, 0.47484927368164065, 0.47597467041015623, 0.4747806701660156, 0.47479296875, 0.47538177490234373, 0.4749906005859375, 0.47488409423828126, 0.4746322021484375, 0.4749168701171875, 0.47477862548828126, 0.4759582824707031, 0.47688909912109373, 0.47617330932617186, 0.47571865844726563, 0.4744898681640625, 0.4741591186523437, 0.47446939086914064, 0.4745513000488281, 0.4750899047851562, 0.4740403137207031, 0.474461181640625, 0.4741929016113281, 0.47546676635742186, 0.47404339599609374, 0.474176513671875, 0.47427685546875, 0.47412841796875, 0.979821533203125, 0.4742574157714844, 0.4741734313964844, 0.47405874633789064, 0.47411508178710937, 0.47479910278320314, 0.476015625, 0.47458816528320313, 0.4742225952148437, 0.47418679809570313, 0.4743782043457031, 0.47417138671875, 0.4748810119628906, 0.47444174194335936, 0.4738058166503906, 0.4744478759765625, 0.47413861083984377, 0.47446221923828125, 0.47416116333007813, 0.47388262939453124, 0.47408843994140626, 0.47393484497070315, 0.4740771789550781, 0.47400857543945313, 0.47406695556640627, 0.4738518981933594, 0.4751022033691406, 0.4763402099609375, 0.47449191284179687, 0.4742512512207031, 0.4740157470703125, 0.4743352355957031, 0.47394509887695313, 0.473849853515625, 0.47430963134765625, 0.4738713684082031, 0.47423590087890627, 0.4743987121582031, 0.47423590087890627, 0.4738887634277344, 0.4744407043457031, 0.4742297668457031, 0.4740741271972656, 0.47423590087890627, 0.47649996948242185, 0.4742901611328125, 0.4741949462890625, 0.47464346313476563, 0.476073974609375, 0.47489022827148436, 0.4758292541503906, 0.47497113037109373, 0.47431890869140625, 0.4741651916503906, 0.4740843505859375, 0.47435162353515625, 0.47437310791015624, 0.4743291015625, 0.47532852172851564, 0.474313720703125, 0.474640380859375, 0.47469473266601564, 0.4747908630371094, 0.9814691772460937, 0.4747683715820312, 0.4747376708984375, 0.4743690185546875, 0.4749127807617187, 0.4760391540527344, 0.4749434814453125, 0.4753387451171875, 0.4744488830566406, 0.4743935852050781, 0.4743935852050781, 0.47374334716796873, 0.47430349731445315, 0.47411404418945313, 0.47499264526367185, 0.47437823486328123, 0.4740597839355469, 0.4755292053222656, 0.474429443359375, 0.4746004333496094, 0.47424716186523436, 0.4757176208496094, 0.474365966796875, 0.4744038391113281, 0.47409765625, 0.47461581420898435, 0.47587124633789063, 0.47467724609375, 0.47449700927734373, 0.4752076721191406, 0.47526400756835935, 0.4758815002441406, 0.4752086791992188, 0.4745318298339844, 0.4740362548828125, 0.47401980590820314, 0.474777587890625, 0.47447760009765627, 0.47450723266601563, 0.4745646057128906, 0.4748257141113281, 0.4746987609863281, 0.4746670227050781, 0.4746844177246094, 0.4750899047851562, 0.474967041015625, 0.4748011474609375, 0.47588555908203123, 0.47625112915039064, 0.4750325622558594, 0.4751790161132812, 0.47611801147460936, 0.4751523742675781, 0.4750223388671875, 0.4746065979003906, 0.4746875, 0.47461068725585936, 0.47841998291015625, 0.4753489990234375, 0.47590911865234375, 0.4750417785644531, 0.474998779296875, 0.47562240600585937, 0.9817487182617187, 0.47520870971679685, 0.4753449096679688, 0.47469158935546873, 0.4756121520996094, 0.47580160522460935, 0.47535000610351563, 0.47563983154296874, 0.47536431884765623, 0.4756459655761719, 0.4758149108886719, 0.47476327514648436, 0.47486669921875, 0.4744960021972656, 0.4750315551757813, 0.4745902099609375, 0.4752404479980469, 0.4754586181640625, 0.47481234741210937, 0.47426150512695314, 0.4745635986328125, 0.4747796630859375, 0.4742758483886719, 0.4749291381835937, 0.4753070068359375, 0.4757319641113281, 0.47559988403320314, 0.4744366149902344, 0.4746649475097656, 0.47449087524414063, 0.47499365234375, 0.47450009155273437, 0.47397579956054686, 0.47451034545898435, 0.47416116333007813, 0.4750878601074219, 0.47417752075195313, 0.47466085815429687, 0.4771686401367187, 0.4746567687988281, 0.47440997314453126, 0.4744816589355469, 0.4743618469238281, 0.47448269653320313, 0.47479705810546874, 0.47491070556640624, 0.47615179443359373, 0.4757442626953125, 0.4753827819824219, 0.4749906005859375, 0.4746219482421875, 0.4746925964355469, 0.4743485412597656, 0.4746967163085937, 0.47494552612304686, 0.4743270263671875, 0.47532647705078124, 0.47462603759765626, 0.4751585388183594, 0.4746270751953125, 0.47493734741210936, 0.47487387084960936, 0.4749823913574219, 0.9828372192382813, 0.4750807189941406, 0.4744366149902344, 0.4755558471679687, 0.47547698974609376, 0.47533465576171874, 0.4759132080078125, 0.4759879760742188, 0.47586407470703124, 0.4761640930175781, 0.4759255065917969, 0.4748451843261719, 0.474919921875, 0.476189697265625, 0.4745257263183594, 0.4750560913085938, 0.475325439453125, 0.474986572265625, 0.4742686157226563, 0.4745994262695313, 0.4747817077636719, 0.4749609069824219, 0.4750120849609375, 0.47644467163085935, 0.47719833374023435, 0.47524453735351563, 0.4750530700683594, 0.4744356384277344, 0.47469769287109376, 0.4745891418457031, 0.47447756958007814, 0.4750274658203125, 0.47433831787109376, 0.47436083984375, 0.4751431579589844, 0.4747120666503906, 0.474977294921875, 0.47543295288085935, 0.4742215576171875, 0.47395327758789063, 0.474292236328125, 0.473849853515625, 0.4742840270996094, 0.4745062255859375, 0.47460455322265627, 0.475863037109375, 0.4749680786132813, 0.4751790161132812, 0.4745482177734375, 0.47489227294921876, 0.4745994873046875, 0.47445599365234375, 0.47442022705078124, 0.47415603637695314, 0.47442739868164063, 0.474777587890625, 0.47418060302734377, 0.4745369567871094, 0.47485952758789063, 0.4748001708984375, 0.474187744140625, 0.47446221923828125, 0.47530703735351565, 0.9833133544921875, 0.4750899047851562, 0.4764405822753906, 0.47526705932617186, 0.4746055603027344, 0.47418368530273436, 0.4742369384765625, 0.4743475341796875, 0.474524658203125, 0.474323974609375, 0.4742778930664063, 0.47433010864257813, 0.4754462585449219, 0.47468850708007815, 0.474829833984375, 0.4747304992675781, 0.47638424682617186, 0.4761231994628906, 0.47679379272460937, 0.4750878601074219, 0.4745164794921875, 0.47426150512695314, 0.4747745361328125, 0.47573504638671876, 0.4750520324707031, 0.474893310546875, 0.4743792724609375, 0.4743096923828125, 0.474462158203125, 0.47433831787109376, 0.4749609069824219, 0.4743250427246094, 0.4742419738769531, 0.47488204956054686, 0.47455438232421876, 0.4743075866699219, 0.4756408386230469, 0.47438339233398436, 0.4743741760253906, 0.47458706665039063, 0.47492404174804687, 0.47545547485351564, 0.47435980224609375, 0.47457485961914064, 0.4752353210449219, 0.47622964477539065, 0.47467416381835936, 0.47408126831054687, 0.47500799560546875, 0.47463320922851565, 0.4743915405273437, 0.4742758483886719, 0.47516363525390626, 0.47539404296875, 0.4750796813964844, 0.4738518981933594, 0.4742266845703125, 0.477939697265625, 0.474281982421875, 0.47408126831054687, 0.4742758483886719, 0.4744366149902344, 0.4743372802734375, 0.9826785278320312, 0.4751769714355469, 0.47564697265625, 0.47452569580078124, 0.4744765319824219, 0.47413861083984377, 0.47415090942382815, 0.4741754760742187, 0.4749885559082031, 0.47412017822265623, 0.4737423400878906, 0.4744488830566406, 0.4739051513671875, 0.4738478088378906, 0.4745400390625, 0.47486770629882813, 0.47466085815429687, 0.47589376831054686, 0.47535205078125, 0.4755548095703125, 0.4743424072265625, 0.47426663208007813, 0.47488409423828126, 0.47611911010742186, 0.47500384521484373, 0.4744591369628906, 0.47539813232421874, 0.47478475952148436, 0.47490567016601565, 0.47446112060546874, 0.475104248046875, 0.4745164794921875, 0.47489535522460935, 0.47453900146484373, 0.4745594787597656, 0.47434445190429686, 0.47417752075195313, 0.47448269653320313, 0.4744724731445312, 0.4748072509765625, 0.47473458862304685, 0.4740843505859375, 0.4747591552734375, 0.4748912658691406, 0.4760514526367188, 0.47554150390625, 0.4754565124511719, 0.47558041381835936, 0.4751912841796875, 0.47503564453125, 0.47414990234375, 0.4742850646972656, 0.47414886474609375, 0.47481036376953123, 0.47404852294921873, 0.4742584228515625, 0.4739358825683594, 0.4744171447753906, 0.4741263427734375, 0.47414169311523435, 0.4739491882324219, 0.47383755493164065, 0.4742522888183594, 0.983693359375, 0.47566241455078123, 0.47446722412109377, 0.4746977233886719, 0.4751247253417969, 0.47474688720703123, 0.47455032348632814, 0.4745143737792969, 0.474745849609375, 0.4746485900878906, 0.4754575500488281, 0.47418881225585935, 0.474029052734375, 0.47392666625976565, 0.4740771789550781, 0.4740321350097656, 0.4739317626953125, 0.4744591369628906, 0.47454925537109377, 0.47516058349609375, 0.47410687255859374, 0.4743935852050781, 0.4761661376953125, 0.4749885559082031, 0.47419699096679685, 0.47460250854492186, 0.47660543823242185, 0.4745430908203125, 0.4743280639648437, 0.47440179443359376, 0.4743813171386719, 0.47454721069335937, 0.4742799377441406, 0.4744407043457031, 0.4743854064941406, 0.47423898315429686, 0.47517593383789064, 0.4753879089355469, 0.47501516723632814, 0.47492202758789065, 0.4744867858886719, 0.47409762573242187, 0.47553741455078125, 0.4761282653808594, 0.47627365112304687, 0.47592141723632814, 0.474745849609375, 0.47454931640625, 0.47427374267578126, 0.47429937744140627, 0.47435775756835935, 0.47429937744140627, 0.475188232421875, 0.4755568542480469, 0.4755281982421875, 0.4754483337402344, 0.4758005676269531, 0.47536639404296877, 0.4753592224121094, 0.4763494567871094, 0.47446011352539064, 0.4744376220703125, 0.4739686279296875]",tokens/s,2.074646015839662,,,main,False,False,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen2-beta-72B,Qwen/Qwen2-beta-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 96.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-2.7B,,cuda,0,42,,,,,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,,MB,2234.441728,2932.342784,0.0,2285.89568,2082.575872,s,10,2.433214202880859,0.24332142028808593,0.0008617435568583196,0.24325077056884764,0.24459763488769531,0.24462198333740234,0.24464146209716797,"[0.24380015563964844, 0.24204917907714843, 0.24258213806152343, 0.24336614990234376, 0.24313539123535155, 0.24226588439941407, 0.24288755798339845, 0.24388919067382814, 0.24464633178710937, 0.24459222412109374]",tokens/s,1052.1063032465574,kWh,2.856464352872637e-06,1.5652168180297718e-06,1.2995936322662033e-05,1.7417617493564442e-05,tokens/kWh,14697762.199370166,MB,2234.441728,2959.60576,0.0,2313.158656,2180.684288,s,10,139.734884765625,13.9734884765625,0.01146198033243705,13.9710908203125,13.98807626953125,13.992370166015625,13.995805283203126,"[13.9871220703125, 13.9770771484375, 13.9966640625, 13.980609375, 13.95928515625, 13.96146875, 13.9670087890625, 13.9751728515625, 13.965376953125, 13.965099609375]",tokens/s,4.508537728833344,kWh,0.00016483188493384256,9.03412226583766e-05,0.0007461166431891384,0.0010012897507813577,tokens/kWh,62918.85036358144,,s,629,141.7022607421874,0.22528181358058424,0.02898003334479797,0.22168063354492187,0.2224162872314453,0.22265180053710937,0.4647246643066406,"[0.22235647583007812, 0.22139903259277344, 0.22141644287109374, 0.22147891235351563, 0.2213693389892578, 0.22170726013183595, 0.22163967895507813, 0.2214256591796875, 0.22199398803710937, 0.22159564208984375, 0.22243431091308594, 0.22190284729003906, 0.22202674865722657, 0.2210723876953125, 0.22141133117675782, 0.22209945678710938, 0.22192127990722657, 0.2224189453125, 0.22194073486328125, 0.22234317016601562, 0.22204518127441406, 0.22228582763671875, 0.2221793212890625, 0.22186904907226562, 0.222603271484375, 0.22207487487792968, 0.22233599853515626, 0.22205238342285155, 0.22235133361816406, 0.2235494384765625, 0.22263296508789063, 0.22236671447753906, 0.22127410888671875, 0.2215219268798828, 0.2213519287109375, 0.22189056396484375, 0.22161100769042968, 0.22196940612792967, 0.22203187561035156, 0.22196531677246092, 0.22140007019042968, 0.2213396453857422, 0.2219325408935547, 0.2218577880859375, 0.22200933837890624, 0.22186495971679687, 0.22244557189941405, 0.22216499328613282, 0.22224179077148437, 0.221955078125, 0.2217943115234375, 0.2224179229736328, 0.22189776611328124, 0.2225540771484375, 0.22166732788085938, 0.22277632141113282, 0.22537420654296875, 0.22221209716796875, 0.22146047973632813, 0.22176870727539064, 0.22216294860839844, 0.22160383605957032, 0.46420379638671877, 0.2221997833251953, 0.22188236999511718, 0.22198272705078126, 0.2217697296142578, 0.22201139831542968, 0.2215034942626953, 0.22119526672363282, 0.22219879150390626, 0.22249369812011718, 0.22189773559570314, 0.22246092224121095, 0.22249574279785156, 0.22148197937011718, 0.22124134826660155, 0.221802490234375, 0.22128536987304687, 0.2213744659423828, 0.22234214782714845, 0.22216806030273437, 0.22141746520996095, 0.22162637329101562, 0.22157516479492187, 0.22105191040039063, 0.2215782470703125, 0.2218741760253906, 0.2216898498535156, 0.22149427795410156, 0.2215536651611328, 0.2222725067138672, 0.22197760009765624, 0.22207795715332032, 0.2216847381591797, 0.22156083679199218, 0.2215741424560547, 0.2220595245361328, 0.22160179138183594, 0.22255001831054688, 0.22156492614746093, 0.22203904724121093, 0.22160179138183594, 0.22140109252929688, 0.22128640747070313, 0.22133660888671874, 0.2213908233642578, 0.22154035949707032, 0.22508236694335937, 0.22183013916015626, 0.22202879333496095, 0.2222407684326172, 0.2216785888671875, 0.22169293212890626, 0.22160691833496093, 0.22218649291992187, 0.2214686737060547, 0.22141439819335937, 0.222487548828125, 0.2225991668701172, 0.22195301818847657, 0.22147276306152344, 0.22199909973144533, 0.22176666259765626, 0.22219879150390626, 0.4654428100585937, 0.2218987579345703, 0.22210765075683594, 0.2219008026123047, 0.22165504455566407, 0.22182707214355468, 0.22211891174316406, 0.22192640686035156, 0.221770751953125, 0.22130482482910158, 0.22165913391113282, 0.22189263916015625, 0.2217932434082031, 0.22167552185058595, 0.22172467041015625, 0.22150143432617186, 0.22293606567382812, 0.22162535095214844, 0.2228019256591797, 0.22176255798339845, 0.22248243713378907, 0.22172979736328124, 0.22229402160644532, 0.22310092163085937, 0.22198988342285156, 0.22210150146484375, 0.22238616943359374, 0.22228480529785155, 0.22175334167480468, 0.22205337524414062, 0.22236058044433593, 0.2222407684326172, 0.2220369873046875, 0.2219929656982422, 0.2221854705810547, 0.2224127960205078, 0.22233395385742188, 0.22231858825683593, 0.22264627075195312, 0.22218751525878908, 0.22241587829589843, 0.22255718994140625, 0.2218741760253906, 0.22225920104980468, 0.22183935546875, 0.22192332458496095, 0.22207795715332032, 0.223752197265625, 0.22219673156738282, 0.22203289794921874, 0.22212197875976564, 0.2223206329345703, 0.22238616943359374, 0.22253575134277342, 0.22210047912597655, 0.22248442077636718, 0.2219356231689453, 0.2220912628173828, 0.22202572631835937, 0.22224896240234376, 0.22215168762207033, 0.22308761596679688, 0.22262168884277345, 0.46642996215820315, 0.22230323791503906, 0.22220390319824218, 0.22247935485839843, 0.22223155212402343, 0.22242303466796876, 0.22273023986816406, 0.2226411590576172, 0.22237493896484375, 0.22229808044433594, 0.2220185546875, 0.22259507751464844, 0.22249984741210938, 0.22244659423828125, 0.22167654418945312, 0.22273228454589844, 0.2222581787109375, 0.22126182556152343, 0.2230200653076172, 0.22316950988769532, 0.22224179077148437, 0.22140518188476563, 0.22184857177734374, 0.22160281372070312, 0.22178099060058593, 0.22235443115234374, 0.2213928985595703, 0.22145535278320314, 0.22126591491699218, 0.22138368225097657, 0.22140518188476563, 0.22138983154296876, 0.22134886169433593, 0.22174208068847656, 0.22144720458984374, 0.22132118225097655, 0.22146354675292967, 0.2214001007080078, 0.22144508361816406, 0.2214297637939453, 0.22163763427734376, 0.221517822265625, 0.22244044494628906, 0.22324838256835938, 0.22501274108886718, 0.22167861938476563, 0.22164067077636718, 0.2215004119873047, 0.22143283081054688, 0.22134579467773438, 0.22140518188476563, 0.22115122985839844, 0.22140415954589843, 0.2217195587158203, 0.22167449951171875, 0.2215034942626953, 0.22143795776367187, 0.2215854034423828, 0.22149530029296874, 0.22169497680664063, 0.22140931701660158, 0.22172157287597657, 0.22152806091308594, 0.4667914123535156, 0.22145228576660156, 0.2217512969970703, 0.2216468505859375, 0.22147789001464843, 0.221444091796875, 0.22145126342773438, 0.22144717407226563, 0.22157005310058595, 0.22128947448730468, 0.22147071838378907, 0.22156903076171874, 0.22139801025390626, 0.2215116729736328, 0.2211778564453125, 0.22171034240722656, 0.22124748229980468, 0.22265548706054689, 0.22149836730957032, 0.22151065063476563, 0.22136422729492186, 0.22120550537109376, 0.22162431335449218, 0.22184754943847657, 0.2219530487060547, 0.22183625793457032, 0.22151065063476563, 0.22187826538085936, 0.2214799346923828, 0.22132736206054687, 0.22136524963378906, 0.22137344360351563, 0.22129356384277343, 0.22121778869628905, 0.22129254150390626, 0.2211164093017578, 0.22158746337890625, 0.22140415954589843, 0.22128128051757812, 0.22147071838378907, 0.22140313720703125, 0.22121273803710936, 0.22131808471679687, 0.2216048583984375, 0.22186810302734375, 0.22218540954589844, 0.221623291015625, 0.22134477233886718, 0.22152088928222657, 0.22141746520996095, 0.22138368225097657, 0.22134169006347656, 0.22129458618164063, 0.22138983154296876, 0.22134169006347656, 0.2212956085205078, 0.221159423828125, 0.2240184326171875, 0.22193971252441405, 0.2216089630126953, 0.22141952514648439, 0.22154853820800782, 0.2216837158203125, 0.46434201049804685, 0.22140415954589843, 0.22150758361816406, 0.22161817932128905, 0.2220298309326172, 0.22168780517578124, 0.2222530517578125, 0.22144613647460937, 0.2217840576171875, 0.22155264282226564, 0.22125669860839844, 0.22169197082519532, 0.22176358032226562, 0.22181779479980468, 0.22157926940917969, 0.2215782470703125, 0.2214246368408203, 0.22152294921875, 0.22142771911621092, 0.22154547119140625, 0.22132429504394532, 0.22135606384277343, 0.22152394104003906, 0.22135398864746095, 0.22151577758789062, 0.2220185546875, 0.22214349365234376, 0.22169088745117188, 0.2212464599609375, 0.22148403930664062, 0.221370361328125, 0.22125363159179687, 0.22161305236816406, 0.22148197937011718, 0.22166015625, 0.2215977020263672, 0.22134783935546876, 0.2212351989746094, 0.2241535949707031, 0.22163558959960938, 0.2216785888671875, 0.22160691833496093, 0.221549560546875, 0.22167141723632813, 0.22142054748535156, 0.2214297637939453, 0.22139903259277344, 0.2217943115234375, 0.22149119567871095, 0.221876220703125, 0.22163250732421874, 0.221765625, 0.2215004119873047, 0.2217400360107422, 0.22149427795410156, 0.2213939208984375, 0.2211758117675781, 0.22143487548828125, 0.22141644287109374, 0.22124339294433593, 0.22147071838378907, 0.22166630554199218, 0.22171136474609374, 0.46487347412109375, 0.22141133117675782, 0.22127206420898438, 0.221338623046875, 0.2219622344970703, 0.22154342651367187, 0.22202265930175782, 0.22145330810546876, 0.22258995056152345, 0.22142874145507813, 0.22155264282226564, 0.22175334167480468, 0.2215741424560547, 0.22162739562988282, 0.22164378356933595, 0.2218803253173828, 0.22163250732421874, 0.22160076904296874, 0.22158848571777343, 0.2215116729736328, 0.2215679931640625, 0.22120550537109376, 0.22134477233886718, 0.22316543579101564, 0.22167039489746093, 0.22168063354492187, 0.22168678283691406, 0.2215116729736328, 0.2215188751220703, 0.22155363464355468, 0.22205235290527345, 0.22159359741210938, 0.22261351013183595, 0.2219059143066406, 0.22219161987304686, 0.22189773559570314, 0.22165504455566407, 0.22153114318847655, 0.2218076171875, 0.22219570922851561, 0.22192445373535155, 0.22225091552734375, 0.22204415893554688, 0.22189260864257812, 0.22191104125976563, 0.2215352325439453, 0.22143385314941405, 0.22180557250976562, 0.22147584533691406, 0.22139698791503906, 0.22178713989257812, 0.22213325500488282, 0.22149221801757812, 0.22222848510742188, 0.22150860595703126, 0.2214686737060547, 0.22136627197265624, 0.22127001953125, 0.22115020751953124, 0.22117478942871094, 0.22150143432617186, 0.22140211486816405, 0.22120550537109376, 0.46591384887695314, 0.22151065063476563, 0.22132838439941407, 0.22122496032714845, 0.22145433044433593, 0.22153318786621093, 0.2213519287109375, 0.22116044616699218, 0.22144717407226563, 0.22135296630859375, 0.2212833251953125, 0.22198477172851563, 0.22165811157226561, 0.22226739501953124, 0.22169804382324218, 0.22206361389160156, 0.22276095581054686, 0.22171136474609374, 0.22175949096679687, 0.22171034240722656, 0.22183833312988283, 0.2217021484375, 0.22223257446289063, 0.221876220703125, 0.22166323852539063, 0.22185369873046876, 0.22207693481445312, 0.22190386962890624, 0.2217379913330078, 0.22185165405273438, 0.22234214782714845, 0.22208717346191406, 0.22299136352539062, 0.22185061645507811, 0.22382899475097656, 0.22139187622070314, 0.22159461975097655, 0.22185369873046876, 0.2214072265625, 0.22193971252441405, 0.22215577697753905, 0.22185267639160156, 0.22164480590820312, 0.22156594848632813, 0.2215188751220703, 0.22147990417480468, 0.22186189270019532, 0.22172262573242188, 0.22174310302734376, 0.22215577697753905, 0.22219366455078124, 0.2219448699951172, 0.2216007385253906, 0.22184959411621094, 0.22170930480957032, 0.22189164733886718, 0.2221639404296875, 0.22170518493652344, 0.22201344299316406, 0.2217830352783203, 0.22169395446777343, 0.22200831604003907, 0.22163148498535157, 0.46692044067382815, 0.2217400360107422, 0.22165196228027345, 0.22197555541992187, 0.22193458557128906, 0.22203904724121093, 0.22154655456542968, 0.22138873291015626, 0.22145330810546876, 0.22146354675292967, 0.22145024108886718, 0.22175640869140625, 0.22180351257324218, 0.22217727661132813, 0.22153727722167968, 0.221896728515625, 0.2214164123535156, 0.2225797119140625, 0.22176051330566407, 0.22177381896972656, 0.22154649353027345, 0.221233154296875, 0.2214615020751953, 0.22134681701660155, 0.22129356384277343, 0.22127615356445313, 0.22128640747070313, 0.22148095703125, 0.22134169006347656, 0.22124339294433593, 0.22114816284179686, 0.2213191680908203, 0.2215188446044922, 0.22131814575195313, 0.22146969604492187, 0.22117990112304686, 0.22134375, 0.22164889526367187, 0.22181581115722657, 0.222171142578125, 0.22175640869140625, 0.22172979736328124, 0.22134066772460936, 0.22151986694335937, 0.22140313720703125, 0.22141030883789062, 0.22220594787597656, 0.22215887451171876, 0.22227349853515624, 0.22183116149902343, 0.22192640686035156, 0.22173184204101562, 0.2216898498535156, 0.22144717407226563, 0.22147584533691406, 0.22146354675292967, 0.22156903076171874, 0.22176051330566407, 0.22247833251953125, 0.221802490234375, 0.22252543640136718, 0.22219468688964844, 0.22166015625, 0.46707403564453126, 0.22175640869140625, 0.22192536926269532, 0.22187315368652344, 0.22208204650878907, 0.22162535095214844, 0.22142361450195314, 0.22128025817871094, 0.22170008850097656, 0.2215782470703125, 0.22144613647460937, 0.22178201293945313, 0.22167654418945312, 0.22178816223144532, 0.2213816375732422, 0.2216816711425781, 0.22158131408691406, 0.22198066711425782, 0.22218853759765625, 0.221770751953125, 0.22207693481445312, 0.2219069366455078, 0.22215782165527342, 0.2219069366455078, 0.2217830352783203, 0.2216417236328125, 0.2232033233642578, 0.22143283081054688, 0.22127719116210937, 0.22148300170898438, 0.22127104187011717, 0.2214615020751953, 0.22147584533691406, 0.22175538635253905, 0.22177690124511718, 0.2215290832519531, 0.22144717407226563, 0.22135296630859375, 0.22129356384277343, 0.22165196228027345, 0.22160076904296874, 0.22165402221679686, 0.22142771911621092, 0.22134783935546876, 0.22133555603027344, 0.22152088928222657, 0.22161509704589843, 0.2215034942626953, 0.22159257507324218, 0.22131712341308593, 0.22130586242675782, 0.22127410888671875, 0.22141850280761718, 0.22161305236816406, 0.22188954162597657, 0.22163456726074218, 0.2218014678955078, 0.2217902069091797, 0.22162124633789063, 0.22154444885253907, 0.22161305236816406, 0.2218956756591797, 0.22184857177734374]",tokens/s,4.438884720014453,,,main,False,False,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-66b,facebook/opt-66b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 162.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,internlm/internlm2-20b,internlm/internlm2-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,4008.419328,15760.621568,0.0,15114.174464,14046.123008,s,10,15.848047607421876,1.5848047607421876,0.001058458314234293,1.584490295410156,1.5859812255859376,1.5864652832031252,1.5868525292968751,"[1.5851588134765624, 1.583447265625, 1.5845269775390625, 1.583828125, 1.58445361328125, 1.583693603515625, 1.5843594970703125, 1.5857567138671875, 1.5869493408203126, 1.5858736572265626]",tokens/s,161.53409324698862,kWh,1.869280167751842e-05,1.0243654597979911e-05,8.793465368099884e-05,0.00011687110995649717,tokens/kWh,2190447.23794692,MB,4008.419328,15760.621568,0.0,15114.174464,14170.904576,s,10,927.2718203125,92.72718203125001,0.005797509915113206,92.7283046875,92.73256640625,92.73465820312501,92.736331640625,"[92.7307421875, 92.72909375, 92.73675, 92.7321015625, 92.7264921875, 92.7241015625, 92.7158671875, 92.7194609375, 92.7296953125, 92.727515625]",tokens/s,0.6794124292353493,kWh,0.0010946948910421795,0.0005999896856933628,0.005209015139431006,0.006903699716166547,tokens/kWh,9125.541751543959,,s,629,940.0627094726563,1.494535309177514,0.18825786848907494,1.47182177734375,1.4724240966796875,1.4726346923828124,3.055808037109375,"[1.4722620849609376, 1.4717921142578125, 1.47181982421875, 1.4719560546875, 1.4718125, 1.4717183837890624, 1.472183349609375, 1.47190576171875, 1.4715218505859375, 1.47171630859375, 1.47156787109375, 1.4714951171875, 1.471922119140625, 1.4721187744140625, 1.4715177001953126, 1.47153515625, 1.4718924560546875, 1.4718126220703125, 1.4714593505859375, 1.4722181396484375, 1.4722314453125, 1.4714234619140625, 1.471951904296875, 1.4719990234375, 1.471762451171875, 1.4720880126953124, 1.472605224609375, 1.471382568359375, 1.4716119384765625, 1.471868896484375, 1.4716876220703126, 1.4716446533203125, 1.472257080078125, 1.4723829345703126, 1.472047119140625, 1.4720819091796875, 1.47234716796875, 1.4716507568359376, 1.47162109375, 1.471951904296875, 1.472247802734375, 1.472111572265625, 1.471752197265625, 1.4720460205078125, 1.4718433837890625, 1.471847412109375, 1.473238037109375, 1.4716826171875, 1.4716129150390624, 1.471711181640625, 1.4720286865234375, 1.4715771484375, 1.47129443359375, 1.4723441162109374, 1.47237890625, 1.4720123291015625, 1.4720389404296874, 1.4725570068359375, 1.4715821533203124, 1.47163232421875, 1.472152587890625, 1.4719139404296875, 3.055659912109375, 1.4719190673828124, 1.4714122314453124, 1.4715750732421875, 1.4726441650390625, 1.4716497802734374, 1.4719395751953126, 1.4714327392578126, 1.4716334228515624, 1.4717982177734374, 1.4725570068359375, 1.472247802734375, 1.4715555419921875, 1.4713795166015624, 1.47167431640625, 1.4718515625, 1.471489990234375, 1.4721033935546874, 1.4721280517578126, 1.471662109375, 1.47139990234375, 1.4714696044921876, 1.471595458984375, 1.4714849853515626, 1.472489501953125, 1.4717716064453126, 1.471542236328125, 1.4713282470703124, 1.4715074462890625, 1.4714327392578126, 1.4716241455078125, 1.4727454833984375, 1.4718914794921876, 1.472322509765625, 1.4719334716796875, 1.47182177734375, 1.4713436279296874, 1.4722703857421875, 1.47228466796875, 1.471656982421875, 1.471455322265625, 1.4724454345703124, 1.4727833251953124, 1.472058349609375, 1.472762939453125, 1.4725816650390624, 1.47173583984375, 1.4722109375, 1.4722037353515625, 1.4719927978515626, 1.4716068115234375, 1.4723389892578125, 1.4720194091796874, 1.4716221923828126, 1.4717747802734376, 1.47186279296875, 1.4714490966796876, 1.471636474609375, 1.4723154296875, 1.471784912109375, 1.47171533203125, 1.4717327880859374, 1.472248779296875, 3.057259521484375, 1.47224267578125, 1.4718597412109375, 1.472489501953125, 1.4721064453125, 1.4717808837890625, 1.471899658203125, 1.4724228515625, 1.472541748046875, 1.47203173828125, 1.472416748046875, 1.47173583984375, 1.471767578125, 1.4718350830078124, 1.472779296875, 1.4719927978515626, 1.47137841796875, 1.4718525390625, 1.4718648681640625, 1.4717603759765625, 1.4714869384765625, 1.47251611328125, 1.4720552978515624, 1.47161083984375, 1.4713856201171875, 1.4714736328125, 1.4710916748046876, 1.471134765625, 1.4725887451171875, 1.4723441162109374, 1.4716630859375, 1.471782958984375, 1.4716395263671875, 1.4718515625, 1.4722734375, 1.472278564453125, 1.4714920654296875, 1.47167431640625, 1.47209423828125, 1.4719703369140624, 1.47245263671875, 1.4724505615234376, 1.47195703125, 1.47188525390625, 1.472077880859375, 1.4723031005859375, 1.4720552978515624, 1.4719139404296875, 1.4722252197265624, 1.4726502685546874, 1.4722242431640624, 1.471540283203125, 1.472048095703125, 1.4718955078125, 1.4713743896484375, 1.47283154296875, 1.4720828857421875, 1.472143310546875, 1.472288818359375, 1.4720460205078125, 1.4718065185546876, 1.471910888671875, 1.472711669921875, 3.055857666015625, 1.471562744140625, 1.4717071533203125, 1.471541259765625, 1.4722652587890626, 1.47290625, 1.4720972900390625, 1.4717244873046875, 1.4717379150390626, 1.47239111328125, 1.4715084228515625, 1.4723399658203125, 1.47213720703125, 1.47205224609375, 1.472205810546875, 1.4718515625, 1.471974365234375, 1.4717972412109375, 1.472838623046875, 1.472647216796875, 1.47232666015625, 1.47219970703125, 1.4716876220703126, 1.4716630859375, 1.4720552978515624, 1.4729554443359374, 1.471932373046875, 1.471942626953125, 1.471614990234375, 1.4719990234375, 1.4717265625, 1.4722620849609376, 1.47253857421875, 1.47146337890625, 1.4715965576171874, 1.4714061279296875, 1.471909912109375, 1.4716488037109374, 1.4721239013671874, 1.472101318359375, 1.4714869384765625, 1.471604736328125, 1.4719661865234375, 1.4716927490234375, 1.4718228759765626, 1.471889404296875, 1.4719180908203124, 1.4718177490234374, 1.4717174072265624, 1.471542236328125, 1.471983642578125, 1.4715238037109375, 1.4717327880859374, 1.472067626953125, 1.471873046875, 1.4716876220703126, 1.4719764404296876, 1.4723194580078125, 1.4719017333984374, 1.4723870849609375, 1.4717716064453126, 1.4715146484375, 1.4714757080078125, 3.056364501953125, 1.4723338623046875, 1.472585693359375, 1.4714593505859375, 1.471646728515625, 1.4715494384765626, 1.4717880859375, 1.472016357421875, 1.4726204833984375, 1.4721822509765625, 1.4717747802734376, 1.4717244873046875, 1.4717716064453126, 1.47205322265625, 1.4717276611328125, 1.472067626953125, 1.471805419921875, 1.4715555419921875, 1.471551513671875, 1.4714962158203124, 1.47133544921875, 1.4717747802734376, 1.471994873046875, 1.47158837890625, 1.471520751953125, 1.4713466796875, 1.4714747314453125, 1.4717276611328125, 1.4718135986328125, 1.4724290771484374, 1.4716488037109374, 1.4714132080078124, 1.4716231689453125, 1.4723389892578125, 1.4713077392578124, 1.47160888671875, 1.471899658203125, 1.4723778076171874, 1.471910888671875, 1.4719447021484375, 1.47195703125, 1.47152685546875, 1.472251953125, 1.471962158203125, 1.4716077880859375, 1.471457275390625, 1.4723092041015624, 1.4719969482421875, 1.4716549072265626, 1.4718648681640625, 1.47160986328125, 1.4718760986328125, 1.471826904296875, 1.4713251953125, 1.47182177734375, 1.4716077880859375, 1.472786376953125, 1.472522216796875, 1.4720809326171875, 1.4717982177734374, 1.4721248779296876, 1.471731689453125, 1.4721033935546874, 3.055680419921875, 1.4718280029296875, 1.4716129150390624, 1.4713046875, 1.4718914794921876, 1.4720511474609375, 1.4723931884765624, 1.471605712890625, 1.4720911865234374, 1.47203173828125, 1.4718822021484375, 1.4717869873046876, 1.4723829345703126, 1.4720809326171875, 1.471552490234375, 1.4721136474609375, 1.471494140625, 1.4717767333984375, 1.471177734375, 1.472689208984375, 1.4716497802734374, 1.4718802490234375, 1.4714306640625, 1.472184326171875, 1.4717860107421874, 1.47194677734375, 1.47276904296875, 1.4713814697265626, 1.4714398193359375, 1.471753173828125, 1.472689208984375, 1.472320556640625, 1.472069580078125, 1.47273828125, 1.4716497802734374, 1.4718289794921875, 1.4717060546875, 1.471951904296875, 1.4716866455078126, 1.47211669921875, 1.471942626953125, 1.4714388427734375, 1.471298583984375, 1.4713538818359375, 1.4712965087890626, 1.4712689208984375, 1.4717244873046875, 1.472357421875, 1.471731689453125, 1.4715218505859375, 1.471215576171875, 1.4711285400390626, 1.4713046875, 1.47230615234375, 1.4722723388671874, 1.4720819091796875, 1.471382568359375, 1.4713026123046875, 1.472006103515625, 1.47163232421875, 1.472415771484375, 1.471131591796875, 1.4713272705078124, 3.056773193359375, 1.4716077880859375, 1.4716282958984375, 1.471858642578125, 1.4712208251953125, 1.4718197021484376, 1.472443359375, 1.4715648193359374, 1.471594482421875, 1.472443359375, 1.4719036865234374, 1.47152587890625, 1.4711285400390626, 1.4714869384765625, 1.4713077392578124, 1.4713907470703125, 1.47195703125, 1.47201123046875, 1.4714542236328125, 1.4714757080078125, 1.471456298828125, 1.471595458984375, 1.47175732421875, 1.4725191650390625, 1.4716539306640626, 1.4716416015625, 1.47211474609375, 1.4713333740234376, 1.47162109375, 1.4716436767578125, 1.4722816162109376, 1.47161083984375, 1.471515625, 1.4715064697265625, 1.4715709228515625, 1.47129345703125, 1.4712586669921874, 1.4722181396484375, 1.471456298828125, 1.4713170166015626, 1.4715576171875, 1.47175732421875, 1.47146337890625, 1.472247802734375, 1.471603759765625, 1.4716273193359375, 1.4715709228515625, 1.472047119140625, 1.4721197509765624, 1.472300048828125, 1.4724617919921874, 1.472036865234375, 1.4715872802734375, 1.47150537109375, 1.47152685546875, 1.4712227783203125, 1.4709483642578125, 1.4714481201171874, 1.4717808837890625, 1.4714285888671874, 1.4712176513671875, 1.4715013427734376, 1.471331298828125, 3.058272216796875, 1.4714920654296875, 1.4712022705078125, 1.47119921875, 1.4717470703125, 1.471373291015625, 1.4719886474609376, 1.4716968994140625, 1.4716273193359375, 1.4718914794921876, 1.471751220703125, 1.4717174072265624, 1.471711181640625, 1.472184326171875, 1.472363525390625, 1.471656982421875, 1.4721873779296875, 1.4718863525390624, 1.4719764404296876, 1.4720511474609375, 1.4719754638671876, 1.4718924560546875, 1.4713037109375, 1.4712841796875, 1.47162109375, 1.4717725830078126, 1.472, 1.4716497802734374, 1.4716558837890625, 1.4713426513671874, 1.471215576171875, 1.4712484130859376, 1.47142041015625, 1.471537109375, 1.4719190673828124, 1.4712698974609375, 1.471478759765625, 1.47183203125, 1.4715013427734376, 1.4721925048828124, 1.47177978515625, 1.472447509765625, 1.4714500732421876, 1.4717501220703124, 1.47160888671875, 1.4718299560546875, 1.471974365234375, 1.4720081787109376, 1.472126953125, 1.4715023193359376, 1.4715084228515625, 1.47169482421875, 1.471330322265625, 1.471171630859375, 1.4719794921875, 1.4723297119140626, 1.471705078125, 1.4717818603515624, 1.4718740234375, 1.4721126708984376, 1.471266845703125, 1.472437255859375, 1.4715064697265625, 3.05955126953125, 1.4719764404296876, 1.47195703125, 1.47245361328125, 1.472227294921875, 1.4713221435546875, 1.4720296630859375, 1.4715545654296875, 1.4716138916015624, 1.4715606689453125, 1.47230517578125, 1.4725037841796875, 1.471837158203125, 1.47186376953125, 1.47192724609375, 1.4720911865234374, 1.47192626953125, 1.4723450927734374, 1.4718555908203126, 1.47135791015625, 1.4716026611328126, 1.471425537109375, 1.4714224853515625, 1.4716395263671875, 1.4727669677734374, 1.4724013671875, 1.4724403076171875, 1.47211669921875, 1.4722672119140625, 1.472236572265625, 1.47209423828125, 1.4720142822265625, 1.471340576171875, 1.4719610595703125, 1.471952880859375, 1.4716343994140626, 1.4717020263671876, 1.472510009765625, 1.4722958984375, 1.47205322265625, 1.471494140625, 1.471520751953125, 1.4711224365234374, 1.4714521484375, 1.471753173828125, 1.4722447509765626, 1.4713282470703124, 1.47150341796875, 1.472290771484375, 1.471889404296875, 1.4715074462890625, 1.4724771728515624, 1.4729246826171876, 1.4722078857421874, 1.4716180419921876, 1.471309814453125, 1.47177978515625, 1.47196923828125, 1.4722120361328126, 1.471774658203125, 1.4713466796875, 1.471520751953125, 1.4712811279296876, 3.058113525390625, 1.4730526123046874, 1.4717235107421875, 1.4721668701171875, 1.471916015625, 1.472654296875, 1.4716385498046876, 1.4719088134765625, 1.4720716552734374, 1.4713763427734374, 1.471236083984375, 1.471573974609375, 1.4717041015625, 1.471826904296875, 1.4721710205078125, 1.472248779296875, 1.471709228515625, 1.47167333984375, 1.4722354736328125, 1.47183203125, 1.4724495849609376, 1.47257861328125, 1.4722344970703125, 1.471520751953125, 1.4716630859375, 1.47169384765625, 1.472069580078125, 1.4725938720703124, 1.472268310546875, 1.471952880859375, 1.4720880126953124, 1.4720225830078124, 1.471952880859375, 1.4723450927734374, 1.472373779296875, 1.472194580078125, 1.4716927490234375, 1.471921142578125, 1.471922119140625, 1.472109619140625, 1.471762451171875, 1.4723779296875, 1.4719671630859374, 1.4716590576171875, 1.47093603515625, 1.4710968017578125, 1.4713907470703125, 1.47113671875, 1.471277099609375, 1.4721495361328125, 1.4713795166015624, 1.4710025634765624, 1.471087646484375, 1.47147265625, 1.4714593505859375, 1.4727730712890625, 1.472385986328125, 1.4715115966796875, 1.47182177734375, 1.471372314453125, 1.471321044921875, 1.4717603759765625, 1.47186181640625]",tokens/s,0.6691042987470994,,,,,,,, +4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-6.7b,EleutherAI/pythia-6.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.1,,,,MB,1308.962816,6256.328704,0.0,5609.8816,5292.617216,s,10,5.571213256835938,0.5571213256835937,0.0007666840176307737,0.5568714599609375,0.557456005859375,0.5584093994140625,0.5591721142578125,"[0.55936279296875, 0.556702392578125, 0.5567649536132813, 0.5570609741210938, 0.5566168212890625, 0.5568453979492187, 0.557244140625, 0.5567742919921875, 0.5569439697265625, 0.5568975219726563]",tokens/s,459.5049376110048,kWh,6.58186607890659e-06,3.606554268359711e-06,3.10157809854445e-05,4.1204201332710805e-05,tokens/kWh,6212958.672172324,MB,1308.962816,6256.328704,0.0,5609.8816,5503.948288,s,10,327.20778125,32.720778124999995,0.005308491364446453,32.7202353515625,32.7266998046875,32.72907548828125,32.73097603515625,"[32.72102734375, 32.713080078125, 32.72320703125, 32.723677734375, 32.719443359375, 32.726171875, 32.7142578125, 32.71880859375, 32.71665625, 32.731451171875]",tokens/s,1.9253820847208232,kWh,0.00038621306252148413,0.00021167820178359155,0.0018056511605691545,0.00240354242487423,tokens/kWh,26211.31183207494,,s,629,331.6996517333982,0.527344438367883,0.06619643329461054,0.519362548828125,0.5196838745117187,0.5197864990234375,1.0765131103515626,"[0.5193318481445313, 0.5191546630859375, 0.5195264282226563, 0.5193011474609375, 0.519161865234375, 0.5189949340820312, 0.519024658203125, 0.5189488525390625, 0.51919873046875, 0.518950927734375, 0.5190645751953125, 0.5191629028320313, 0.5196636352539062, 0.5194833984375, 0.5193001098632812, 0.5194168090820312, 0.5192283935546875, 0.5194127807617187, 0.5191557006835937, 0.5193123779296875, 0.5200199584960937, 0.5194977416992187, 0.5196605224609375, 0.519141357421875, 0.5195980834960937, 0.5191690063476563, 0.5194024658203125, 0.51964208984375, 0.5193021240234375, 0.5191884765625, 0.5192212524414063, 0.5194844360351563, 0.5194926147460938, 0.5195140991210937, 0.519436279296875, 0.5195079956054688, 0.5191536865234375, 0.5194281005859375, 0.5195120849609375, 0.5192755126953125, 0.5196728515625, 0.5195980834960937, 0.5193359375, 0.5193533325195312, 0.5194619140625, 0.5194895629882812, 0.5192806396484375, 0.5191629028320313, 0.51945166015625, 0.519245849609375, 0.5195632934570312, 0.519736328125, 0.5192724609375, 0.5192140502929687, 0.5193277587890625, 0.5196113891601563, 0.5192888793945313, 0.519089111328125, 0.5195858154296875, 0.519647216796875, 0.5195612182617187, 0.5193564453125, 1.0763663330078126, 0.518898681640625, 0.5189488525390625, 0.5190789184570312, 0.5190236206054688, 0.5189979858398438, 0.5195140991210937, 0.5191475219726562, 0.5191536865234375, 0.5191444702148438, 0.5191588134765625, 0.519103515625, 0.5190625, 0.5190420532226563, 0.5190901489257812, 0.5192744750976562, 0.5193707275390625, 0.5191854248046875, 0.5194649658203125, 0.5193666381835937, 0.5193236694335938, 0.5194281005859375, 0.5191649169921875, 0.5191342163085938, 0.519046142578125, 0.5191383056640625, 0.5194874877929687, 0.5194271240234375, 0.5192376098632813, 0.5192274169921876, 0.5190830078125, 0.519161865234375, 0.51907275390625, 0.51944140625, 0.5194660034179688, 0.5193697509765625, 0.5191290893554688, 0.5192161254882812, 0.51917822265625, 0.51928271484375, 0.5191895141601562, 0.5192591552734375, 0.5195140991210937, 0.51934619140625, 0.5191177978515625, 0.51966259765625, 0.5194188842773437, 0.5195499267578125, 0.5193359375, 0.5193554077148438, 0.5193215942382813, 0.5193707275390625, 0.5193646240234375, 0.5194526977539062, 0.5195642700195312, 0.5192232666015625, 0.5191710815429688, 0.519193603515625, 0.5191710815429688, 0.519245849609375, 0.5192489013671875, 0.519278564453125, 0.5190645751953125, 1.076937744140625, 0.5192662963867187, 0.5193164672851562, 0.5193584594726562, 0.5193011474609375, 0.5191680297851563, 0.5191188354492188, 0.5191270141601563, 0.5195693969726562, 0.5193901977539063, 0.5192611694335938, 0.519352294921875, 0.519352294921875, 0.5192232666015625, 0.5193594970703125, 0.51930419921875, 0.519245849609375, 0.5191546630859375, 0.5192427368164062, 0.5194291381835937, 0.5196943359375, 0.519520263671875, 0.5194066162109375, 0.5193318481445313, 0.5193011474609375, 0.5194178466796875, 0.5193533325195312, 0.5193809814453125, 0.5195612182617187, 0.5193554077148438, 0.5194926147460938, 0.5192898559570313, 0.5194373168945312, 0.5194137573242188, 0.519341064453125, 0.5194496459960938, 0.5192283935546875, 0.51919873046875, 0.5195048828125, 0.5196339111328125, 0.5195560913085937, 0.519572509765625, 0.51938916015625, 0.5192929077148437, 0.5191874389648438, 0.5194547119140625, 0.5193114013671875, 0.5195682983398437, 0.5193922729492187, 0.5194547119140625, 0.5196246948242188, 0.5193451538085937, 0.5195867919921875, 0.5193400268554688, 0.5194721069335938, 0.5192376098632813, 0.51959912109375, 0.5194547119140625, 0.5198776245117187, 0.5195980834960937, 0.519794677734375, 0.5196093139648438, 0.5194988403320312, 1.07694482421875, 0.51944140625, 0.5193348999023437, 0.519572509765625, 0.5193656616210938, 0.5192765502929687, 0.5190225830078125, 0.5192960205078125, 0.5193001098632812, 0.519414794921875, 0.5196113891601563, 0.5192478637695312, 0.5194208984375, 0.5195591430664063, 0.5194547119140625, 0.5194055786132813, 0.5193932495117187, 0.5193236694335938, 0.519299072265625, 0.519203857421875, 0.519172119140625, 0.519413818359375, 0.5194751586914063, 0.5194956665039062, 0.5193031616210938, 0.5192714233398438, 0.5192642822265625, 0.5195704345703125, 0.5194055786132813, 0.5192109985351563, 0.5192069702148437, 0.5193225708007813, 0.5193994140625, 0.5194393310546875, 0.5193717651367188, 0.51938916015625, 0.5194823608398438, 0.5195161743164063, 0.5194229736328125, 0.5193123779296875, 0.5194803466796875, 0.5196267700195313, 0.5195346069335938, 0.5195284423828125, 0.5194066162109375, 0.5194495849609375, 0.5196309204101562, 0.5191699829101563, 0.5192222900390625, 0.5191680297851563, 0.5192140502929687, 0.5192509155273437, 0.5197455444335938, 0.51945166015625, 0.5199011840820312, 0.5196165161132813, 0.5196021728515625, 0.51944140625, 0.519468017578125, 0.5196871948242188, 0.5198428344726562, 0.5196503295898437, 0.5193164672851562, 1.076336669921875, 0.5190471801757812, 0.51890380859375, 0.5190809326171875, 0.5190072021484375, 0.5190051879882812, 0.519067626953125, 0.5190942993164063, 0.5192550659179688, 0.5191608276367188, 0.51911474609375, 0.5190000610351563, 0.5190625, 0.5189099731445312, 0.5189284057617187, 0.5190532836914062, 0.5191976928710937, 0.5195181884765625, 0.5195192260742187, 0.5192243041992187, 0.5194373779296875, 0.5194208374023438, 0.5193380126953125, 0.51938818359375, 0.5193707275390625, 0.5193901977539063, 0.5194732055664063, 0.5196451416015625, 0.5194977416992187, 0.5196513061523438, 0.5195796508789062, 0.5192212524414063, 0.5195172119140625, 0.5192601318359376, 0.5197967529296875, 0.5195601806640625, 0.5197598876953125, 0.5193154296875, 0.5192755126953125, 0.5190604858398438, 0.519140380859375, 0.519161865234375, 0.5192714233398438, 0.5192130737304688, 0.519172119140625, 0.5193922729492187, 0.5197404174804687, 0.5195489501953126, 0.5196461791992187, 0.5194475708007813, 0.5194373168945312, 0.5194741821289063, 0.5194240112304688, 0.5196431274414063, 0.51991552734375, 0.5195888671875, 0.5194813232421875, 0.5196830444335937, 0.5193574829101563, 0.5195447387695312, 0.519552001953125, 0.5194495849609375, 0.5193912353515625, 1.0770308837890625, 0.519299072265625, 0.519161865234375, 0.519352294921875, 0.519202880859375, 0.519232421875, 0.5192161254882812, 0.5191495971679687, 0.5193451538085937, 0.5194649658203125, 0.5192969970703125, 0.51949365234375, 0.5189734497070313, 0.5189652709960938, 0.5190738525390625, 0.5190205078125, 0.5188731079101563, 0.5189837036132813, 0.5191946411132813, 0.5191895141601562, 0.5194915771484375, 0.5194823608398438, 0.5194649658203125, 0.5193861083984375, 0.519541748046875, 0.5195222778320312, 0.5193441162109375, 0.5196687622070313, 0.5194107055664062, 0.5194291381835937, 0.5196728515625, 0.519736328125, 0.5194956665039062, 0.5196932983398438, 0.51970458984375, 0.5194557495117188, 0.5194752197265625, 0.5195489501953126, 0.5196452026367188, 0.51976904296875, 0.5196585693359375, 0.5195037841796875, 0.5197609252929688, 0.5197005004882812, 0.5194229736328125, 0.51953564453125, 0.5195899047851562, 0.5200025634765625, 0.5196328735351563, 0.5195530395507812, 0.5195929565429688, 0.5197322387695312, 0.519456787109375, 0.5192703857421875, 0.5194905395507813, 0.5194588012695313, 0.5194240112304688, 0.5194588012695313, 0.519572509765625, 0.5200353393554688, 0.5195612182617187, 0.5195438232421875, 0.5196062622070312, 1.0774559326171875, 0.5191044921875, 0.5191157836914062, 0.5191874389648438, 0.5194168090820312, 0.519161865234375, 0.5189949340820312, 0.5192335205078125, 0.5191475219726562, 0.5193502807617187, 0.5192642822265625, 0.5191946411132813, 0.5190123291015625, 0.5190532836914062, 0.5194086303710937, 0.5190615234375, 0.5189734497070313, 0.5190645751953125, 0.5192376098632813, 0.5190942993164063, 0.5191700439453125, 0.5191280517578125, 0.5192765502929687, 0.5192151489257812, 0.5193707275390625, 0.5190543212890625, 0.5191290893554688, 0.5193359375, 0.5194874877929687, 0.5193267211914062, 0.5189846801757813, 0.51919873046875, 0.5191116943359375, 0.5194178466796875, 0.51928369140625, 0.5192315063476562, 0.51917724609375, 0.5193430786132812, 0.5191895141601562, 0.5191710815429688, 0.5193011474609375, 0.5194956665039062, 0.5192806396484375, 0.519357421875, 0.5193850708007812, 0.5195530395507812, 0.5191905517578125, 0.519109619140625, 0.5193072509765625, 0.5192550659179688, 0.5195325317382813, 0.5192079467773437, 0.5195632934570312, 0.5195438232421875, 0.5192283935546875, 0.5193164672851562, 0.5193871459960937, 0.519130126953125, 0.5190543212890625, 0.519066650390625, 0.519762939453125, 0.5196871948242188, 0.51964208984375, 1.0765701904296876, 0.5193727416992188, 0.5190225830078125, 0.51900927734375, 0.51913525390625, 0.5190553588867187, 0.5191137084960937, 0.519404541015625, 0.5193369750976562, 0.5193421020507812, 0.5189796142578125, 0.5190471801757812, 0.5191260375976563, 0.5190000610351563, 0.5188761596679687, 0.5193082885742187, 0.5192929077148437, 0.5189151000976563, 0.5193871459960937, 0.5191874389648438, 0.5193277587890625, 0.5193389892578125, 0.5194967041015625, 0.5193380126953125, 0.518867919921875, 0.5191762084960938, 0.5194178466796875, 0.5198345947265625, 0.5194864501953125, 0.519278564453125, 0.5193430786132812, 0.5191434326171875, 0.51901953125, 0.519066650390625, 0.5193963623046876, 0.5190963134765625, 0.51917822265625, 0.5191843872070312, 0.5193840942382812, 0.5189468383789062, 0.5197189331054688, 0.5195069580078125, 0.5195929565429688, 0.5196318969726562, 0.5192345581054687, 0.5199708251953125, 0.5198817138671875, 0.5194895629882812, 0.5193318481445313, 0.51960009765625, 0.519520263671875, 0.5197168579101562, 0.5195089721679688, 0.5197127685546875, 0.5197619018554688, 0.5193789672851562, 0.5196011352539063, 0.5199605712890625, 0.5192171630859375, 0.519773193359375, 0.5193154296875, 0.5193421020507812, 0.5191127319335938, 1.076674560546875, 0.5193175048828125, 0.5198510131835937, 0.5194219360351563, 0.5192171630859375, 0.5192335205078125, 0.5190491943359375, 0.5191076049804687, 0.519288818359375, 0.519351318359375, 0.51945166015625, 0.5192109985351563, 0.5194281005859375, 0.5190523071289063, 0.518887451171875, 0.5189160766601563, 0.5193380126953125, 0.5191219482421875, 0.5194137573242188, 0.5189949340820312, 0.5194024658203125, 0.5194833984375, 0.5191393432617187, 0.5193871459960937, 0.5191823120117187, 0.519035888671875, 0.5191137084960937, 0.5192263793945312, 0.5193850708007812, 0.5195980834960937, 0.519099365234375, 0.5191802978515625, 0.519103515625, 0.5190819702148437, 0.5191076049804687, 0.5194823608398438, 0.519362548828125, 0.5193380126953125, 0.5192847290039062, 0.5194066162109375, 0.5196646118164062, 0.519319580078125, 0.5197086791992187, 0.5193389892578125, 0.5191321411132812, 0.5191423950195313, 0.5191177978515625, 0.51974658203125, 0.5195028686523437, 0.5193082885742187, 0.5194967041015625, 0.5192734985351563, 0.5193185424804687, 0.5191393432617187, 0.5193789672851562, 0.5193001098632812, 0.519245849609375, 0.5193215942382813, 0.5193871459960937, 0.5200670776367188, 0.5194752197265625, 0.5193564453125, 0.519161865234375, 1.0765865478515626, 0.5188423461914062, 0.5193768920898437, 0.5195335693359375, 0.5195069580078125, 0.5196103515625, 0.5192714233398438, 0.5193687133789062, 0.5194761962890625, 0.5191976928710937, 0.5193594970703125, 0.5194977416992187, 0.5192212524414063, 0.5192960205078125, 0.5195693969726562, 0.51974755859375, 0.519372802734375, 0.5196605224609375, 0.5194926147460938, 0.5194332275390625, 0.5192581176757812, 0.5195612182617187, 0.519888916015625, 0.52008251953125, 0.519522216796875, 0.519762939453125, 0.5197742309570312, 0.5199441528320312, 0.5196011352539063, 0.5197128295898438, 0.519615478515625, 0.5195438232421875, 0.5197282104492188, 0.52025439453125, 0.52010595703125, 0.5197557983398438, 0.5197957153320313, 0.5193380126953125, 0.5193861083984375, 0.5192591552734375, 0.5193103637695312, 0.5194598388671875, 0.519741455078125, 0.5195346069335938, 0.519635986328125, 0.5194772338867187, 0.5194158325195313, 0.5195806884765625, 0.5193267211914062, 0.5193666381835937, 0.5194485473632813, 0.5195233154296875, 0.5197117309570313, 0.519857177734375, 0.519552001953125, 0.519488525390625, 0.5193113403320313, 0.519372802734375, 0.5194302368164062, 0.5193245849609375, 0.5197352905273438, 0.51986328125, 0.5195817260742187]",tokens/s,1.8962938209701676,,,,,,,, +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-1.4b,EleutherAI/pythia-1.4b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Salesforce/codegen-6B-nl,Salesforce/codegen-6B-nl,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: CodeGenForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-70m,EleutherAI/pythia-70m,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-110B,Qwen/Qwen1.5-110B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 32.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciCoder-1b,Deci/DeciCoder-1b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 609, in resolve_trust_remote_code + answer = input( +EOFError: EOF when reading a line + +During handling of the above exception, another exception occurred: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 228, in load_model_with_no_weights + self.create_no_weights_model() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 210, in create_no_weights_model + meta_model = self.automodel_class.from_config(self.pretrained_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 418, in from_config + trust_remote_code = resolve_trust_remote_code( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 622, in resolve_trust_remote_code + raise ValueError( +ValueError: The repository for Deci/DeciCoder-1b contains custom code which must be executed to correctly load the model. You can inspect the repository content at https://hf.co/Deci/DeciCoder-1b. +Please pass the argument `trust_remote_code=True` to allow custom code to be run. + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/recurrentgemma-2b,google/recurrentgemma-2b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/recurrentgemma-2b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aaa6e-2addff2d3636d5c66f0d7b0e;540f09a1-f8d5-44f7-9a31-4c35f6e45cb8) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/recurrentgemma-2b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/recurrentgemma-2b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,internlm/internlm-20b,internlm/internlm-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 609, in resolve_trust_remote_code + answer = input( +EOFError: EOF when reading a line + +During handling of the above exception, another exception occurred: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 228, in load_model_with_no_weights + self.create_no_weights_model() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 210, in create_no_weights_model + meta_model = self.automodel_class.from_config(self.pretrained_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 418, in from_config + trust_remote_code = resolve_trust_remote_code( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 622, in resolve_trust_remote_code + raise ValueError( +ValueError: The repository for internlm/internlm-20b contains custom code which must be executed to correctly load the model. You can inspect the repository content at https://hf.co/internlm/internlm-20b. +Please pass the argument `trust_remote_code=True` to allow custom code to be run. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-2.7b,facebook/opt-2.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: OPTForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-125m,facebook/opt-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: OPTForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/gemma-7b,google/gemma-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/gemma-7b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aaa1c-007c37da082a575d102a98bd;d62f780e-9e6a-4dd1-9710-246195337376) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/gemma-7b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/gemma-7b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,huggyllama/llama-13b,huggyllama/llama-13b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-7B,Qwen/Qwen-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 609, in resolve_trust_remote_code + answer = input( +EOFError: EOF when reading a line + +During handling of the above exception, another exception occurred: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 228, in load_model_with_no_weights + self.create_no_weights_model() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 210, in create_no_weights_model + meta_model = self.automodel_class.from_config(self.pretrained_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 418, in from_config + trust_remote_code = resolve_trust_remote_code( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 622, in resolve_trust_remote_code + raise ValueError( +ValueError: The repository for Qwen/Qwen-7B contains custom code which must be executed to correctly load the model. You can inspect the repository content at https://hf.co/Qwen/Qwen-7B. +Please pass the argument `trust_remote_code=True` to allow custom code to be run. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-72B,Qwen/Qwen1.5-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 96.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciLM-7B,Deci/DeciLM-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 609, in resolve_trust_remote_code + answer = input( +EOFError: EOF when reading a line + +During handling of the above exception, another exception occurred: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 228, in load_model_with_no_weights + self.create_no_weights_model() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 210, in create_no_weights_model + meta_model = self.automodel_class.from_config(self.pretrained_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 418, in from_config + trust_remote_code = resolve_trust_remote_code( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 622, in resolve_trust_remote_code + raise ValueError( +ValueError: The repository for Deci/DeciLM-7B contains custom code which must be executed to correctly load the model. You can inspect the repository content at https://hf.co/Deci/DeciLM-7B. +Please pass the argument `trust_remote_code=True` to allow custom code to be run. + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-j-6b,EleutherAI/gpt-j-6b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTJForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-0.5B,Qwen/Qwen1.5-0.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-160m,EleutherAI/pythia-160m,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Salesforce/codegen-16B-nl,Salesforce/codegen-16B-nl,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: CodeGenForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-MoE-A2.7B,Qwen/Qwen1.5-MoE-A2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-1.3b,EleutherAI/pythia-1.3b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-2.7b,EleutherAI/pythia-2.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-14B,Qwen/Qwen1.5-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen2-beta-14B,Qwen/Qwen2-beta-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-30b,facebook/opt-30b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: OPTForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-1.3B,EleutherAI/gpt-neo-1.3B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,databricks/dbrx-base,databricks/dbrx-base,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/databricks/dbrx-base/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aa537-319e27b2631f0106330d203e;1f015b9a-cc9a-4442-a1ca-3f308f8f238f) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/databricks/dbrx-base/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like databricks/dbrx-base is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,huggyllama/llama-7b,huggyllama/llama-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neox-20b,EleutherAI/gpt-neox-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-4B,Qwen/Qwen1.5-4B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-1.8B,Qwen/Qwen1.5-1.8B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-14B,Qwen/Qwen-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 609, in resolve_trust_remote_code + answer = input( +EOFError: EOF when reading a line + +During handling of the above exception, another exception occurred: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 228, in load_model_with_no_weights + self.create_no_weights_model() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 210, in create_no_weights_model + meta_model = self.automodel_class.from_config(self.pretrained_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 418, in from_config + trust_remote_code = resolve_trust_remote_code( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 622, in resolve_trust_remote_code + raise ValueError( +ValueError: The repository for Qwen/Qwen-14B contains custom code which must be executed to correctly load the model. You can inspect the repository content at https://hf.co/Qwen/Qwen-14B. +Please pass the argument `trust_remote_code=True` to allow custom code to be run. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-7.5B,facebook/xglm-7.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: XGLMForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-12b,EleutherAI/pythia-12b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-7B,Qwen/Qwen1.5-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-34B,01-ai/Yi-34B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/gemma-2b,google/gemma-2b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/gemma-2b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aa9c2-758b38da2fab44a872a7eefb;dfcf6b04-c4b9-4836-aeac-32beee847fc8) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/gemma-2b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/gemma-2b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/polyglot-ko-12.8b,EleutherAI/polyglot-ko-12.8b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-125m,EleutherAI/gpt-neo-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-13b,facebook/opt-13b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: OPTForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/recurrentgemma-7b,google/recurrentgemma-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 404 Client Error: Not Found for url: https://huggingface.co/google/recurrentgemma-7b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1823, in _raise_on_head_call_error + raise head_call_error + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 352, in hf_raise_for_status + raise RepositoryNotFoundError(message, response) from e +huggingface_hub.utils._errors.RepositoryNotFoundError: 404 Client Error. (Request ID: Root=1-664aaac0-2256a0e5174bdade23205d9c;d17d72c7-67d3-42ab-9d95-4f698844f852) + +Repository Not Found for url: https://huggingface.co/google/recurrentgemma-7b/resolve/main/config.json. +Please make sure you specified the correct `repo_id` and `repo_type`. +If you are trying to access a private or gated repo, make sure you are authenticated. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 422, in cached_file + raise EnvironmentError( +OSError: google/recurrentgemma-7b is not a local folder and is not a valid model identifier listed on 'https://huggingface.co/models' +If this is a private repository, make sure to pass a token having permission to this repo either by logging in with `huggingface-cli login` or by passing `token=` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-410m,EleutherAI/pythia-410m,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-32B,Qwen/Qwen1.5-32B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-4.5B,facebook/xglm-4.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: XGLMForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,TencentARC/Mistral_Pro_8B_v0.1,TencentARC/Mistral_Pro_8B_v0.1,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-564M,facebook/xglm-564M,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: XGLMForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-350m,facebook/opt-350m,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: OPTForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-6.7b,facebook/opt-6.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: OPTForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-72B,Qwen/Qwen-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 609, in resolve_trust_remote_code + answer = input( +EOFError: EOF when reading a line + +During handling of the above exception, another exception occurred: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 228, in load_model_with_no_weights + self.create_no_weights_model() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 210, in create_no_weights_model + meta_model = self.automodel_class.from_config(self.pretrained_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 418, in from_config + trust_remote_code = resolve_trust_remote_code( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 622, in resolve_trust_remote_code + raise ValueError( +ValueError: The repository for Qwen/Qwen-72B contains custom code which must be executed to correctly load the model. You can inspect the repository content at https://hf.co/Qwen/Qwen-72B. +Please pass the argument `trust_remote_code=True` to allow custom code to be run. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-6B,01-ai/Yi-6B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen2-beta-72B,Qwen/Qwen2-beta-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 96.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-2.7B,EleutherAI/gpt-neo-2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-66b,facebook/opt-66b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: OPTForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,internlm/internlm2-20b,internlm/internlm2-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 609, in resolve_trust_remote_code + answer = input( +EOFError: EOF when reading a line + +During handling of the above exception, another exception occurred: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 228, in load_model_with_no_weights + self.create_no_weights_model() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 210, in create_no_weights_model + meta_model = self.automodel_class.from_config(self.pretrained_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 418, in from_config + trust_remote_code = resolve_trust_remote_code( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 622, in resolve_trust_remote_code + raise ValueError( +ValueError: The repository for internlm/internlm2-20b contains custom code which must be executed to correctly load the model. You can inspect the repository content at https://hf.co/internlm/internlm2-20b. +Please pass the argument `trust_remote_code=True` to allow custom code to be run. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-6.7b,EleutherAI/pythia-6.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-1.4b,EleutherAI/pythia-1.4b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Salesforce/codegen-6B-nl,Salesforce/codegen-6B-nl,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: CodeGenForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-70m,EleutherAI/pythia-70m,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-110B,Qwen/Qwen1.5-110B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 32.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciCoder-1b,Deci/DeciCoder-1b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 68, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1164, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 968, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 713, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) +TypeError: DeciCoderAttention.forward() got an unexpected keyword argument 'cache_position' + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/recurrentgemma-2b,google/recurrentgemma-2b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/recurrentgemma-2b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aaa5f-63892e1e661fe8ce4e9deab1;691eba98-1cd4-4777-a25a-e64bc7bfde6c) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/recurrentgemma-2b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/recurrentgemma-2b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,internlm/internlm-20b,internlm/internlm-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 558, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3626, in from_pretrained + model = cls(config, *model_args, **model_kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm-20b/80729bcf52fbc4553d965926b27304ac5e156d98/modeling_internlm.py"", line 906, in __init__ + self.model = InternLMModel(config) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm-20b/80729bcf52fbc4553d965926b27304ac5e156d98/modeling_internlm.py"", line 729, in __init__ + self.layers = nn.ModuleList([InternLMDecoderLayer(config) for _ in range(config.num_hidden_layers)]) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm-20b/80729bcf52fbc4553d965926b27304ac5e156d98/modeling_internlm.py"", line 729, in + self.layers = nn.ModuleList([InternLMDecoderLayer(config) for _ in range(config.num_hidden_layers)]) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm-20b/80729bcf52fbc4553d965926b27304ac5e156d98/modeling_internlm.py"", line 545, in __init__ + self.self_attn = INTERNLM_ATTENTION_CLASSES[config.attn_implementation](config=config) +KeyError: 'sdpa' + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-2.7b,facebook/opt-2.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: OPTForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-125m,facebook/opt-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: OPTForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/gemma-7b,google/gemma-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/gemma-7b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aaa0d-7090eb320a3a9677364f74d7;8e660734-5141-4aff-845a-e1a4de473d20) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/gemma-7b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/gemma-7b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,huggyllama/llama-13b,huggyllama/llama-13b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1164, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 968, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 713, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 615, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-7B,Qwen/Qwen-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 550, in from_pretrained + model_class = get_class_from_dynamic_module( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 498, in get_class_from_dynamic_module + final_module = get_cached_module_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 323, in get_cached_module_file + modules_needed = check_imports(resolved_module_file) + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 181, in check_imports + raise ImportError( +ImportError: This modeling file requires the following packages that were not found in your environment: transformers_stream_generator. Run `pip install transformers_stream_generator` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-72B,Qwen/Qwen1.5-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 96.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciLM-7B,Deci/DeciLM-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 558, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: DeciLMForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-j-6b,EleutherAI/gpt-j-6b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTJForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-0.5B,Qwen/Qwen1.5-0.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1149, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1034, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 748, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 644, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-160m,EleutherAI/pythia-160m,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Salesforce/codegen-16B-nl,Salesforce/codegen-16B-nl,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: CodeGenForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-MoE-A2.7B,Qwen/Qwen1.5-MoE-A2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3640, in from_pretrained + hf_quantizer.preprocess_model( + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 182, in preprocess_model + return self._process_model_before_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 85, in _process_model_before_weight_loading + model, has_been_replaced = replace_with_awq_linear( + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 178, in replace_with_awq_linear + _, has_been_replaced = replace_with_awq_linear( + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 178, in replace_with_awq_linear + _, has_been_replaced = replace_with_awq_linear( + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 178, in replace_with_awq_linear + _, has_been_replaced = replace_with_awq_linear( + [Previous line repeated 1 more time] + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 165, in replace_with_awq_linear + model._modules[name] = target_cls( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 46, in __init__ + assert out_features % (32 // self.w_bit) == 0 +AssertionError + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-1.3b,EleutherAI/pythia-1.3b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-2.7b,EleutherAI/pythia-2.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-14B,Qwen/Qwen1.5-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1149, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1034, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 748, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 644, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen2-beta-14B,Qwen/Qwen2-beta-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1149, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1034, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 748, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 644, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-30b,facebook/opt-30b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: OPTForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-1.3B,EleutherAI/gpt-neo-1.3B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,databricks/dbrx-base,databricks/dbrx-base,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/databricks/dbrx-base/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aa529-6094d41f55ed664a136e5a2f;0a009b52-eafb-476c-bdc3-9a83a02e5974) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/databricks/dbrx-base/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like databricks/dbrx-base is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,huggyllama/llama-7b,huggyllama/llama-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1164, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 968, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 713, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 615, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neox-20b,EleutherAI/gpt-neox-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-4B,Qwen/Qwen1.5-4B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1149, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1034, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 748, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 644, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-1.8B,Qwen/Qwen1.5-1.8B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1149, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1034, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 748, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 644, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-14B,Qwen/Qwen-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 550, in from_pretrained + model_class = get_class_from_dynamic_module( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 498, in get_class_from_dynamic_module + final_module = get_cached_module_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 323, in get_cached_module_file + modules_needed = check_imports(resolved_module_file) + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 181, in check_imports + raise ImportError( +ImportError: This modeling file requires the following packages that were not found in your environment: transformers_stream_generator. Run `pip install transformers_stream_generator` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-7.5B,facebook/xglm-7.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: XGLMForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-12b,EleutherAI/pythia-12b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-7B,Qwen/Qwen1.5-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1149, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1034, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 748, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 644, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-34B,01-ai/Yi-34B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 68, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1164, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 968, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 713, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 615, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/gemma-2b,google/gemma-2b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/gemma-2b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aa9b1-67e0fd51178f04d9724b0a74;3cf4d4ee-2a88-4567-90a1-31421aa01148) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/gemma-2b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/gemma-2b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/polyglot-ko-12.8b,EleutherAI/polyglot-ko-12.8b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-125m,EleutherAI/gpt-neo-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-13b,facebook/opt-13b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: OPTForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/recurrentgemma-7b,google/recurrentgemma-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 404 Client Error: Not Found for url: https://huggingface.co/google/recurrentgemma-7b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1823, in _raise_on_head_call_error + raise head_call_error + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 352, in hf_raise_for_status + raise RepositoryNotFoundError(message, response) from e +huggingface_hub.utils._errors.RepositoryNotFoundError: 404 Client Error. (Request ID: Root=1-664aaab2-0518e2d112534a360e150f71;aec3db9e-6a2b-4e9e-a0cd-1529f6d19130) + +Repository Not Found for url: https://huggingface.co/google/recurrentgemma-7b/resolve/main/config.json. +Please make sure you specified the correct `repo_id` and `repo_type`. +If you are trying to access a private or gated repo, make sure you are authenticated. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 422, in cached_file + raise EnvironmentError( +OSError: google/recurrentgemma-7b is not a local folder and is not a valid model identifier listed on 'https://huggingface.co/models' +If this is a private repository, make sure to pass a token having permission to this repo either by logging in with `huggingface-cli login` or by passing `token=` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-410m,EleutherAI/pythia-410m,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-32B,Qwen/Qwen1.5-32B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1149, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 1034, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 748, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/qwen2/modeling_qwen2.py"", line 644, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-4.5B,facebook/xglm-4.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: XGLMForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,TencentARC/Mistral_Pro_8B_v0.1,TencentARC/Mistral_Pro_8B_v0.1,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 64, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run + _ = backend.generate(self.inputs, self.config.generate_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate + return self.pretrained_model.generate(**inputs, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample + outputs = self( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/mistral/modeling_mistral.py"", line 1139, in forward + outputs = self.model( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/mistral/modeling_mistral.py"", line 1024, in forward + layer_outputs = decoder_layer( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/mistral/modeling_mistral.py"", line 738, in forward + hidden_states, self_attn_weights, present_key_value = self.self_attn( + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/mistral/modeling_mistral.py"", line 639, in forward + query_states = self.q_proj(hidden_states) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl + return self._call_impl(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl + return forward_call(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context + return func(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward + assert AWQ_INSTALLED, ( +AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-564M,facebook/xglm-564M,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: XGLMForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-350m,facebook/opt-350m,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: OPTForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-6.7b,facebook/opt-6.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: OPTForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-72B,Qwen/Qwen-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 550, in from_pretrained + model_class = get_class_from_dynamic_module( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 498, in get_class_from_dynamic_module + final_module = get_cached_module_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 323, in get_cached_module_file + modules_needed = check_imports(resolved_module_file) + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 181, in check_imports + raise ImportError( +ImportError: This modeling file requires the following packages that were not found in your environment: transformers_stream_generator. Run `pip install transformers_stream_generator` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-6B,01-ai/Yi-6B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 68, in run + report = scenario.run(backend) + File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run _ = backend.generate(self.inputs, self.config.generate_kwargs) File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context return func(*args, **kwargs) @@ -2222,33 +21175,33 @@ ChildProcessError: Traceback (most recent call last): return self.pretrained_model.generate(**inputs, **kwargs) File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1576, in generate - result = self._greedy_search( - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2494, in _greedy_search + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1736, in generate + result = self._sample( + File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2375, in _sample outputs = self( File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl return self._call_impl(*args, **kwargs) File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1211, in forward + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1164, in forward outputs = self.model( File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl return self._call_impl(*args, **kwargs) File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1018, in forward + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 968, in forward layer_outputs = decoder_layer( File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl return self._call_impl(*args, **kwargs) File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 741, in forward + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 713, in forward hidden_states, self_attn_weights, present_key_value = self.self_attn( File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl return self._call_impl(*args, **kwargs) File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 351, in forward + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 615, in forward query_states = self.q_proj(hidden_states) File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl return self._call_impl(*args, **kwargs) @@ -2260,329 +21213,1071 @@ ChildProcessError: Traceback (most recent call last): assert AWQ_INSTALLED, ( AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-125m,EleutherAI/gpt-neo-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen2-beta-72B,Qwen/Qwen2-beta-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 96.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-2.7B,EleutherAI/gpt-neo-2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-66b,facebook/opt-66b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: OPTForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,internlm/internlm2-20b,internlm/internlm2-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 558, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3626, in from_pretrained + model = cls(config, *model_args, **model_kwargs) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm2-20b/f363ea8a116b3ea829c7a068ca24bc9d3e668083/modeling_internlm2.py"", line 974, in __init__ + self.model = InternLM2Model(config) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm2-20b/f363ea8a116b3ea829c7a068ca24bc9d3e668083/modeling_internlm2.py"", line 796, in __init__ + self.layers = nn.ModuleList([InternLM2DecoderLayer(config) for _ in range(config.num_hidden_layers)]) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm2-20b/f363ea8a116b3ea829c7a068ca24bc9d3e668083/modeling_internlm2.py"", line 796, in + self.layers = nn.ModuleList([InternLM2DecoderLayer(config) for _ in range(config.num_hidden_layers)]) + File ""/root/.cache/huggingface/modules/transformers_modules/internlm/internlm2-20b/f363ea8a116b3ea829c7a068ca24bc9d3e668083/modeling_internlm2.py"", line 598, in __init__ + self.attention = INTERNLM2_ATTENTION_CLASSES[config.attn_implementation](config=config) +KeyError: 'sdpa' + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-6.7b,EleutherAI/pythia-6.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3620, in from_pretrained + config = cls._autoset_attn_implementation( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1478, in _autoset_attn_implementation + config = cls._check_and_enable_sdpa( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1644, in _check_and_enable_sdpa + raise ValueError( +ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-1.4b,EleutherAI/pythia-1.4b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Salesforce/codegen-6B-nl,Salesforce/codegen-6B-nl,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-70m,EleutherAI/pythia-70m,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-110B,Qwen/Qwen1.5-110B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 32.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciCoder-1b,Deci/DeciCoder-1b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 609, in resolve_trust_remote_code + answer = input( +EOFError: EOF when reading a line + +During handling of the above exception, another exception occurred: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 228, in load_model_with_no_weights + self.create_no_weights_model() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 210, in create_no_weights_model + meta_model = self.automodel_class.from_config(self.pretrained_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 418, in from_config + trust_remote_code = resolve_trust_remote_code( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 622, in resolve_trust_remote_code + raise ValueError( +ValueError: The repository for Deci/DeciCoder-1b contains custom code which must be executed to correctly load the model. You can inspect the repository content at https://hf.co/Deci/DeciCoder-1b. +Please pass the argument `trust_remote_code=True` to allow custom code to be run. + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/recurrentgemma-2b,google/recurrentgemma-2b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/recurrentgemma-2b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aaa52-0f35cf87603651cb4a6737df;6ff300d9-260a-42bd-97a7-879160c6673f) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/recurrentgemma-2b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/recurrentgemma-2b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,internlm/internlm-20b,internlm/internlm-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 609, in resolve_trust_remote_code + answer = input( +EOFError: EOF when reading a line + +During handling of the above exception, another exception occurred: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 228, in load_model_with_no_weights + self.create_no_weights_model() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 210, in create_no_weights_model + meta_model = self.automodel_class.from_config(self.pretrained_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 418, in from_config + trust_remote_code = resolve_trust_remote_code( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 622, in resolve_trust_remote_code + raise ValueError( +ValueError: The repository for internlm/internlm-20b contains custom code which must be executed to correctly load the model. You can inspect the repository content at https://hf.co/internlm/internlm-20b. +Please pass the argument `trust_remote_code=True` to allow custom code to be run. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-2.7b,facebook/opt-2.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-125m,facebook/opt-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/gemma-7b,google/gemma-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/gemma-7b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aa9ff-22ae56ac6d35a58840a8c5fe;e9439071-a0e6-4776-8c01-3f5de3095ae3) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/gemma-7b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 64, in run - report = scenario.run(backend) - File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run - _ = backend.generate(self.inputs, self.config.generate_kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate - return self.pretrained_model.generate(**inputs, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1576, in generate - result = self._greedy_search( - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2494, in _greedy_search - outputs = self( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 974, in forward - transformer_outputs = self.transformer( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 842, in forward - outputs = block( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 569, in forward - attn_outputs = self.attn( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 521, in forward - return self.attention( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 263, in forward - query = self.q_proj(hidden_states) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward - assert AWQ_INSTALLED, ( -AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/gemma-7b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,huggyllama/llama-13b,huggyllama/llama-13b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-7B,Qwen/Qwen-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 609, in resolve_trust_remote_code + answer = input( +EOFError: EOF when reading a line + +During handling of the above exception, another exception occurred: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 228, in load_model_with_no_weights + self.create_no_weights_model() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 210, in create_no_weights_model + meta_model = self.automodel_class.from_config(self.pretrained_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 418, in from_config + trust_remote_code = resolve_trust_remote_code( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 622, in resolve_trust_remote_code + raise ValueError( +ValueError: The repository for Qwen/Qwen-7B contains custom code which must be executed to correctly load the model. You can inspect the repository content at https://hf.co/Qwen/Qwen-7B. +Please pass the argument `trust_remote_code=True` to allow custom code to be run. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-72B,Qwen/Qwen1.5-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 96.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciLM-7B,Deci/DeciLM-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 609, in resolve_trust_remote_code + answer = input( +EOFError: EOF when reading a line + +During handling of the above exception, another exception occurred: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 228, in load_model_with_no_weights + self.create_no_weights_model() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 210, in create_no_weights_model + meta_model = self.automodel_class.from_config(self.pretrained_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 418, in from_config + trust_remote_code = resolve_trust_remote_code( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 622, in resolve_trust_remote_code + raise ValueError( +ValueError: The repository for Deci/DeciLM-7B contains custom code which must be executed to correctly load the model. You can inspect the repository content at https://hf.co/Deci/DeciLM-7B. +Please pass the argument `trust_remote_code=True` to allow custom code to be run. + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-j-6b,EleutherAI/gpt-j-6b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-0.5B,Qwen/Qwen1.5-0.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-160m,EleutherAI/pythia-160m,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Salesforce/codegen-16B-nl,Salesforce/codegen-16B-nl,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-6B,01-ai/Yi-6B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-MoE-A2.7B,Qwen/Qwen1.5-MoE-A2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 64, in run - report = scenario.run(backend) - File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run - _ = backend.generate(self.inputs, self.config.generate_kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate - return self.pretrained_model.generate(**inputs, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1576, in generate - result = self._greedy_search( - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2494, in _greedy_search - outputs = self( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1211, in forward - outputs = self.model( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1018, in forward - layer_outputs = decoder_layer( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 741, in forward - hidden_states, self_attn_weights, present_key_value = self.self_attn( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 351, in forward - query_states = self.q_proj(hidden_states) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward - assert AWQ_INSTALLED, ( -AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemv-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-2.7B,EleutherAI/gpt-neo-2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-1.3b,EleutherAI/pythia-1.3b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 64, in run - report = scenario.run(backend) - File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run - _ = backend.generate(self.inputs, self.config.generate_kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate - return self.pretrained_model.generate(**inputs, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1576, in generate - result = self._greedy_search( - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2494, in _greedy_search - outputs = self( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 974, in forward - transformer_outputs = self.transformer( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 842, in forward - outputs = block( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 569, in forward - attn_outputs = self.attn( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 521, in forward - return self.attention( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/gpt_neo/modeling_gpt_neo.py"", line 263, in forward - query = self.q_proj(hidden_states) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward - assert AWQ_INSTALLED, ( -AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciCoder-1b,Deci/DeciCoder-1b,cuda,0,42,,,True,True,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-2.7b,EleutherAI/pythia-2.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 64, in run - report = scenario.run(backend) - File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run - _ = backend.generate(self.inputs, self.config.generate_kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate - return self.pretrained_model.generate(**inputs, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1576, in generate - result = self._greedy_search( - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2494, in _greedy_search - outputs = self( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1211, in forward - outputs = self.model( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1018, in forward - layer_outputs = decoder_layer( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 741, in forward - hidden_states, self_attn_weights, present_key_value = self.self_attn( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) -TypeError: DeciCoderAttention.forward() got an unexpected keyword argument 'cache_position' + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciLM-7B,,cuda,0,42,,,,,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,main,False,False,MB,1662.808064,5516.034048,0.0,4869.586944,4743.593472,s,10,6.1237835083007806,0.6123783508300781,0.001253925145814982,0.6120383911132812,0.6140244384765625,0.6141366088867187,0.6142263452148438,"[0.6138787841796876, 0.614248779296875, 0.6107066650390625, 0.61129736328125, 0.6111465454101562, 0.6114146728515625, 0.6116507568359375, 0.612426025390625, 0.61399951171875, 0.613014404296875]",tokens/s,418.0422114089963,kWh,7.21943411562178e-06,3.95425479798766e-06,3.477342651153299e-05,4.594711542514243e-05,tokens/kWh,5571622.889299289,MB,1662.808064,5516.034048,0.0,4869.586944,4769.651712,s,10,360.87087890624997,36.087087890625,0.01748383820965051,36.082986328125,36.11191953125,36.119573046875,36.125695859375,"[36.11021875, 36.1272265625, 36.08849609375, 36.07920703125, 36.0845625, 36.08481640625, 36.08141015625, 36.0650625, 36.07158203125, 36.078296875]",tokens/s,1.7457767773045123,kWh,0.00042595356252458363,0.00023346088450391426,0.0019820382702698735,0.0026414527172983716,tokens/kWh,23850.51210170259,,s,629,365.80599053955063,0.5815675525271077,0.07276675093795772,0.5726207885742187,0.5739550537109376,0.5744299926757812,1.1846730029296875,"[0.5736171264648438, 0.573517822265625, 0.5738854370117188, 0.57415576171875, 0.5742704467773437, 0.5731041259765625, 0.57312255859375, 0.5737461547851562, 0.5739632568359375, 0.5725850219726563, 0.5728665161132812, 0.5735147705078125, 0.5726494750976563, 0.573000732421875, 0.57260546875, 0.5726546020507812, 0.572073974609375, 0.5728809204101563, 0.572885986328125, 0.5725767822265625, 0.5731676025390625, 0.5745838012695312, 0.5729105834960937, 0.5730211791992188, 0.5729566650390625, 0.5723678588867187, 0.5724866333007812, 0.5726197509765625, 0.5735526123046875, 0.5740676879882812, 0.5729720458984375, 0.5732095947265625, 0.5726760864257813, 0.5724497680664062, 0.5724518432617187, 0.572790771484375, 0.5723668823242187, 0.5727159423828125, 0.5731563720703124, 0.5741424560546875, 0.5723576049804687, 0.5724334106445312, 0.5723883666992188, 0.5724548950195313, 0.5725368041992187, 0.5722142944335937, 0.5739407348632812, 0.5740206298828125, 0.572927001953125, 0.573212646484375, 0.5730181274414062, 0.5728184204101563, 0.5730816040039063, 0.5726177368164063, 0.572484619140625, 0.5727293701171875, 0.5740543823242188, 0.5732669677734376, 0.57483056640625, 0.574424072265625, 0.5744517211914062, 0.5744885864257813, 1.18734130859375, 0.5729505004882812, 0.5725030517578125, 0.5725654907226563, 0.5738137817382812, 0.572379150390625, 0.5741465454101562, 0.574750732421875, 0.573475830078125, 0.5734666137695312, 0.5736826782226563, 0.5750292358398438, 0.5742151489257813, 0.5739274291992188, 0.5746913452148438, 0.5744281616210938, 0.5742459106445312, 0.5746903076171875, 0.5750845336914062, 0.5739735107421875, 0.5736888427734375, 0.574476318359375, 0.5740001220703125, 0.573032470703125, 0.5722992553710937, 0.57221630859375, 0.571994140625, 0.573065185546875, 0.573844482421875, 0.5739673461914062, 0.5742745361328125, 0.5743206787109375, 0.5743236694335937, 0.5738577880859375, 0.5742110595703125, 0.5753764038085938, 0.5745797119140625, 0.57402880859375, 0.5725982666015625, 0.5732260131835938, 0.5737594604492188, 0.5739530029296875, 0.5731553344726562, 0.572927978515625, 0.57257470703125, 0.5744312133789062, 0.5755975952148438, 0.5727283325195313, 0.5721149291992188, 0.572199951171875, 0.5723617553710938, 0.572105712890625, 0.57270068359375, 0.5733294067382813, 0.5723873291015625, 0.57191015625, 0.573065185546875, 0.5727866821289063, 0.5724027099609375, 0.5726883544921875, 0.5723258666992187, 0.5720657958984375, 0.5731727294921874, 1.1836395263671875, 0.5721917724609376, 0.5728533325195313, 0.5732484741210937, 0.5725409545898438, 0.5732628784179687, 0.5741773071289062, 0.5730109252929687, 0.5731737670898438, 0.573043701171875, 0.5734686889648437, 0.5732505493164063, 0.5726914672851563, 0.5729924926757812, 0.5729525756835937, 0.5732260131835938, 0.574719970703125, 0.5727836303710937, 0.5730283813476562, 0.5724508056640625, 0.5733519287109375, 0.5728767700195313, 0.5725736694335938, 0.5743431396484375, 0.5745357055664062, 0.573507568359375, 0.572695556640625, 0.5729177856445312, 0.573022216796875, 0.5733980102539062, 0.5731030883789062, 0.5730037841796874, 0.572927978515625, 0.5738895263671875, 0.573137939453125, 0.57310107421875, 0.5718549194335938, 0.5721885986328125, 0.57208935546875, 0.5725255737304688, 0.5722398681640625, 0.5733836669921875, 0.5725736694335938, 0.57238525390625, 0.5722880249023438, 0.5720064086914063, 0.5721456909179687, 0.57385986328125, 0.5721773681640625, 0.5720176391601562, 0.5720484008789063, 0.5730570068359375, 0.5720934448242188, 0.5720852661132813, 0.5719961547851562, 0.5719992065429688, 0.5727928466796876, 0.5721221313476562, 0.5721630859375, 0.572600341796875, 0.5723125610351563, 0.5724631958007812, 0.5727794799804687, 1.1843287353515626, 0.572779541015625, 0.57261669921875, 0.5734041748046875, 0.5739274291992188, 0.57350146484375, 0.57328125, 0.5731553344726562, 0.5729392700195313, 0.5729403076171875, 0.5732833251953126, 0.57289013671875, 0.5743707885742187, 0.5724887084960938, 0.5720043334960937, 0.5718681640625, 0.5722941284179688, 0.5721036987304687, 0.5721763916015625, 0.57200537109375, 0.5717247924804687, 0.572310546875, 0.5729740600585937, 0.5724682006835937, 0.5719664916992188, 0.5723975830078125, 0.5722286376953125, 0.5723289794921875, 0.572221435546875, 0.5728389282226563, 0.5737000732421875, 0.5728737182617187, 0.5730037841796874, 0.572178466796875, 0.5720411376953125, 0.5717974853515625, 0.5721036987304687, 0.5720719604492187, 0.5719695434570312, 0.5730037841796874, 0.5722941284179688, 0.5718763427734375, 0.5725706176757812, 0.5727713012695312, 0.5752719116210937, 0.5729822998046875, 0.5730355224609375, 0.5741107177734375, 0.5727109375, 0.5726207885742187, 0.5727815551757812, 0.5729136352539063, 0.572896240234375, 0.572600341796875, 0.5730191650390625, 0.5724456787109375, 0.573222900390625, 0.57379736328125, 0.5722828979492187, 0.572052490234375, 0.5719766845703125, 0.5719429321289062, 0.5717380981445312, 1.1848990478515624, 0.5739724731445313, 0.5733632202148438, 0.5733693237304688, 0.5734880981445313, 0.5738117065429688, 0.5729403076171875, 0.5730672607421875, 0.5729863891601562, 0.5730140380859375, 0.57394482421875, 0.5726064453125, 0.5722265625, 0.5720391845703126, 0.5731246337890625, 0.5720433349609375, 0.57220703125, 0.5742510375976563, 0.5736365966796875, 0.572663818359375, 0.5723699340820313, 0.5724794921875, 0.5720340576171875, 0.5722705688476563, 0.5724047241210938, 0.57236376953125, 0.5725368041992187, 0.5725255737304688, 0.5728256225585937, 0.5726392211914062, 0.5721354370117188, 0.5721804809570312, 0.5721774291992188, 0.5723484497070312, 0.5721793823242187, 0.572990478515625, 0.5730048217773438, 0.5721558837890625, 0.572242919921875, 0.5721978759765625, 0.5722470703125, 0.5720596313476562, 0.5724047241210938, 0.5724436645507812, 0.5722009887695313, 0.5737861328125, 0.57619970703125, 0.5726668701171875, 0.5722962036132813, 0.5730099487304687, 0.5725439453125, 0.5726085205078125, 0.5728368530273438, 0.5743206176757812, 0.5726044311523437, 0.5726791381835937, 0.5723678588867187, 0.5721937866210938, 0.572822509765625, 0.5727109375, 0.5722070922851562, 0.5722654418945312, 0.5724968872070313, 1.1851029052734374, 0.5723607177734376, 0.57244775390625, 0.5727620849609375, 0.5731502075195313, 0.5729341430664062, 0.5733734130859375, 0.5732301025390625, 0.572663818359375, 0.5735577392578125, 0.5724036865234375, 0.5722737426757812, 0.5723555297851562, 0.5722726440429687, 0.5723494262695312, 0.5725409545898438, 0.5730109252929687, 0.57223681640625, 0.5723402099609375, 0.5722613525390625, 0.572516357421875, 0.5727579956054687, 0.5724467163085938, 0.572221435546875, 0.5731195068359375, 0.5724169921875, 0.5723781127929688, 0.5720811767578124, 0.5718343505859375, 0.5718425903320312, 0.57302734375, 0.5730027465820312, 0.5725501708984375, 0.5730252685546875, 0.57335498046875, 0.5724067993164063, 0.5720145874023438, 0.5724354858398437, 0.5753599853515625, 0.572169189453125, 0.5720125732421875, 0.5726105346679687, 0.5732147216796875, 0.57269970703125, 0.5723483276367187, 0.572010498046875, 0.5722890014648437, 0.5727958984375, 0.5723873901367188, 0.57226953125, 0.5728460693359375, 0.57364892578125, 0.5736980590820312, 0.5732618408203125, 0.5729075317382812, 0.5730396118164063, 0.5730027465820312, 0.5737277221679687, 0.5737984008789062, 0.574740478515625, 0.5733519287109375, 0.5725675659179688, 0.572537841796875, 1.184806884765625, 0.5725839233398438, 0.5726617431640625, 0.573106201171875, 0.5745018920898437, 0.5735618286132812, 0.573259765625, 0.5744578857421875, 0.5749258422851562, 0.5728265991210938, 0.5725081787109375, 0.5727150268554687, 0.5735505981445312, 0.573075439453125, 0.5722562255859375, 0.57212109375, 0.5720760498046875, 0.5723033447265625, 0.5736908569335938, 0.5737195434570312, 0.5725010375976562, 0.57371337890625, 0.5728123168945313, 0.5738147583007812, 0.571978759765625, 0.5718763427734375, 0.5724252319335937, 0.5723299560546875, 0.5721712646484375, 0.5726679077148438, 0.57254296875, 0.5723995971679687, 0.5719193725585937, 0.5722777709960938, 0.571779052734375, 0.5716900024414062, 0.5719306030273438, 0.5717493896484375, 0.571821044921875, 0.5734307861328125, 0.5732904663085937, 0.5728604125976563, 0.572748779296875, 0.57318603515625, 0.5727365112304688, 0.5732413330078125, 0.57249072265625, 0.5729822998046875, 0.573169677734375, 0.5721641235351562, 0.5722808227539062, 0.5722900390625, 0.57253173828125, 0.5722101440429688, 0.5723197631835938, 0.5721200561523437, 0.5727498168945313, 0.5729525756835937, 0.5725030517578125, 0.5723596801757812, 0.5721190185546875, 0.5728788452148438, 0.57232177734375, 1.18618115234375, 0.5729382934570313, 0.5724763793945312, 0.5728051147460937, 0.572410888671875, 0.5722828979492187, 0.5726187744140625, 0.57223681640625, 0.5720278930664062, 0.5722142944335937, 0.5735720825195313, 0.5724139404296875, 0.5722695922851563, 0.5720924072265625, 0.572410888671875, 0.5720698852539062, 0.5738741455078125, 0.5721026611328125, 0.5729136352539063, 0.5728265991210938, 0.5724375, 0.5722224731445312, 0.57202587890625, 0.573212646484375, 0.572484619140625, 0.5720780639648437, 0.57230859375, 0.57369384765625, 0.5728798828125, 0.572020751953125, 0.5721231079101563, 0.5717554931640625, 0.572156982421875, 0.5722654418945312, 0.5718435668945312, 0.5726340942382813, 0.573137939453125, 0.5722542114257813, 0.5726156616210938, 0.572095458984375, 0.572031982421875, 0.5723658447265625, 0.5724303588867188, 0.5722726440429687, 0.5725307006835938, 0.5731655883789063, 0.5721272583007813, 0.5720043334960937, 0.572042236328125, 0.572242919921875, 0.5721856079101563, 0.5724548950195313, 0.5724661865234375, 0.5730764770507812, 0.5732301025390625, 0.5725450439453125, 0.5724866333007812, 0.5721734008789062, 0.5719572143554688, 0.5721549072265625, 0.572000244140625, 0.5720064086914063, 0.5725757446289063, 1.1873709716796874, 0.5724487915039063, 0.5732085571289063, 0.572284912109375, 0.5727989501953125, 0.5728409423828125, 0.5729740600585937, 0.573497314453125, 0.572516357421875, 0.572705810546875, 0.5721886596679687, 0.5721958618164062, 0.5724036865234375, 0.5723381958007813, 0.57270068359375, 0.5734564208984375, 0.5733170776367188, 0.5739089965820312, 0.5724610595703125, 0.5728818969726562, 0.5730293579101563, 0.57270068359375, 0.572495849609375, 0.5729935302734375, 0.5724405517578125, 0.572822509765625, 0.5726105346679687, 0.572315673828125, 0.5717718505859375, 0.5719408569335938, 0.5722603759765625, 0.5722347412109375, 0.5724713134765625, 0.5740123901367188, 0.5722296142578125, 0.5719521484375, 0.572314697265625, 0.572317626953125, 0.5720811767578124, 0.5719705810546875, 0.5717545166015625, 0.5723504638671875, 0.5726515502929688, 0.5724016723632812, 0.5723504638671875, 0.572231689453125, 0.572205078125, 0.5721190185546875, 0.572031982421875, 0.5724467163085938, 0.573043701171875, 0.5730468139648438, 0.5726412353515625, 0.5725030517578125, 0.5725399169921875, 0.5724713134765625, 0.5720606689453125, 0.5723914184570312, 0.5728767700195313, 0.573179931640625, 0.5726597290039063, 0.572295166015625, 0.5719317016601563, 1.1857969970703126, 0.5727313842773437, 0.572368896484375, 0.572347412109375, 0.5732669677734376, 0.5725573120117188, 0.5721907348632812, 0.5723095092773437, 0.5727498168945313, 0.5725081787109375, 0.5725491333007813, 0.5726720581054687, 0.5733457641601563, 0.5733345336914063, 0.5721354370117188, 0.5723043823242188, 0.5719183349609375, 0.5717770385742188, 0.5724968872070313, 0.5726105346679687, 0.573053955078125, 0.5741567993164063, 0.5731102905273437, 0.5720698852539062, 0.5720145874023438, 0.57215185546875, 0.574278564453125, 0.5722521362304688, 0.5720453491210937, 0.5730723266601563, 0.5723924560546875, 0.5727354736328125, 0.572337158203125, 0.5723197631835938, 0.5723668212890625, 0.572221435546875, 0.5729075317382812, 0.5726556396484375, 0.5724456787109375, 0.5738014526367188, 0.5731779174804688, 0.572559326171875, 0.5726812133789062, 0.5728235473632812, 0.572885986328125, 0.5726618041992187, 0.5722745971679688, 0.5731082153320313, 0.5732781982421875, 0.5720135498046875, 0.5721507568359375, 0.57227880859375, 0.5728726806640625, 0.572346435546875, 0.5726781005859375, 0.5728286743164063, 0.5733345336914063, 0.5729228515625, 0.572737548828125, 0.5724548950195313, 0.5741793212890625, 0.5722265625, 0.572494873046875]",tokens/s,1.7194907034525255,,,,, -4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-j-6b,,cuda,0,42,,,,,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,main,False,False,MB,2054.77888,5566.365696,0.0,4919.918592,4635.537408,s,10,5.159750671386719,0.5159750671386718,0.0024202673095528563,0.5153813171386719,0.5170659973144531,0.5199656463623047,0.5222853656005859,"[0.5228652954101562, 0.5143955078125, 0.5157650146484375, 0.5144519653320313, 0.5155265502929688, 0.5141107177734375, 0.516309814453125, 0.5146680908203125, 0.516421630859375, 0.515236083984375]",tokens/s,496.14800463061573,kWh,6.087385805116759e-06,3.3354087369571064e-06,2.7887036198498415e-05,3.7309830740572285e-05,tokens/kWh,6861462.379179726,MB,2054.77888,5566.365696,0.0,4919.918592,4794.439168,s,10,301.012869140625,30.101286914062502,0.007474756939868433,30.101162109375,30.1092380859375,30.11128505859375,30.11292263671875,"[30.104474609375, 30.0874609375, 30.10366796875, 30.09865625, 30.095248046875, 30.1085703125, 30.108783203125, 30.11333203125, 30.097923828125, 30.094751953125]",tokens/s,2.0929337732257594,kWh,0.00035535723721815483,0.00019476662748246779,0.0016039330748120972,0.00215405693951272,tokens/kWh,29247.13773548231,,s,629,305.1713383178704,0.4851690593288888,0.06123176770881389,0.4777687072753906,0.47843901367187497,0.47872921142578123,0.9921051538085938,"[0.47712460327148437, 0.4769034118652344, 0.47702426147460936, 0.47709390258789064, 0.47730584716796876, 0.4771205139160156, 0.4773304443359375, 0.47765298461914063, 0.4770846862792969, 0.4785274963378906, 0.47798681640625, 0.47815167236328127, 0.4778588256835937, 0.4773570556640625, 0.477907958984375, 0.47804620361328126, 0.47764480590820313, 0.4777400207519531, 0.4774901733398437, 0.4775546875, 0.4790906982421875, 0.47841998291015625, 0.4775454711914062, 0.4774205322265625, 0.4777062377929687, 0.4773007507324219, 0.47782296752929687, 0.477517822265625, 0.47927194213867186, 0.47922592163085936, 0.4796824951171875, 0.4796200866699219, 0.478603271484375, 0.47771649169921876, 0.4775034790039063, 0.47786904907226563, 0.4779376525878906, 0.4778526611328125, 0.47831039428710936, 0.47768167114257815, 0.4777820129394531, 0.47838104248046875, 0.4779438171386719, 0.4779376525878906, 0.47762432861328125, 0.47779022216796874, 0.47766015625, 0.4779346008300781, 0.47730584716796876, 0.47776666259765627, 0.4777215881347656, 0.4779438171386719, 0.4776980590820312, 0.4786534423828125, 0.47755673217773437, 0.4775034790039063, 0.47755059814453127, 0.47741543579101564, 0.4771686401367187, 0.47783526611328125, 0.4779069519042969, 0.4776560668945313, 0.9921392822265624, 0.4773294067382812, 0.4772085876464844, 0.477370361328125, 0.4770396728515625, 0.4769504699707031, 0.47741748046875, 0.47746356201171875, 0.4771512451171875, 0.47819979858398437, 0.4786329650878906, 0.47734783935546876, 0.477412353515625, 0.4770017395019531, 0.47757003784179686, 0.47735910034179685, 0.47719219970703125, 0.4772116394042969, 0.4775659484863281, 0.47811892700195313, 0.4780257263183594, 0.4779304809570312, 0.4777492370605469, 0.4775045166015625, 0.4776427612304687, 0.4780738525390625, 0.4773498840332031, 0.47781170654296873, 0.4772178039550781, 0.47686654663085937, 0.4779991149902344, 0.47811175537109374, 0.47742669677734373, 0.4774471740722656, 0.47721368408203124, 0.4768757629394531, 0.47745944213867186, 0.4774481811523438, 0.4774799499511719, 0.4782315979003906, 0.4776785583496094, 0.4772024230957031, 0.477707275390625, 0.47777587890625, 0.47741030883789065, 0.47758950805664063, 0.47756494140625, 0.47724337768554687, 0.4777215881347656, 0.47767962646484374, 0.47741543579101564, 0.47842715454101564, 0.4785602111816406, 0.4774819946289062, 0.4774615173339844, 0.4775465087890625, 0.4775475158691406, 0.4778106994628906, 0.4775577697753906, 0.4773294067382812, 0.47762738037109376, 0.47781375122070313, 0.4781895751953125, 0.9939957885742188, 0.47789157104492186, 0.4779029235839844, 0.4775321044921875, 0.4773294067382812, 0.47682968139648435, 0.4777994384765625, 0.478455810546875, 0.4786780090332031, 0.47789056396484375, 0.47756390380859376, 0.4770979919433594, 0.47737344360351563, 0.4772915344238281, 0.4772536315917969, 0.4778649597167969, 0.47748095703125, 0.4771614685058594, 0.47814862060546875, 0.47750143432617187, 0.4773918762207031, 0.47743896484375, 0.4776662902832031, 0.47875582885742185, 0.4773284606933594, 0.4779539794921875, 0.4778567810058594, 0.47754238891601564, 0.47780154418945314, 0.47773382568359374, 0.4781598815917969, 0.4780472412109375, 0.4777594909667969, 0.477765625, 0.47762841796875, 0.47739596557617187, 0.47771444702148436, 0.47763250732421875, 0.4777687072753906, 0.4783943786621094, 0.47779736328125, 0.4777215881347656, 0.47773284912109376, 0.47762841796875, 0.4779315185546875, 0.4779735107421875, 0.4777461853027344, 0.4787445983886719, 0.47787722778320313, 0.4785080261230469, 0.4784373779296875, 0.47873638916015626, 0.47828274536132814, 0.4777001037597656, 0.47793560791015627, 0.4780277709960937, 0.47800115966796874, 0.47787213134765627, 0.47758950805664063, 0.47742156982421874, 0.47888998413085937, 0.478159912109375, 0.47799700927734373, 0.9920173950195312, 0.47721368408203124, 0.4778260498046875, 0.477939697265625, 0.47749530029296877, 0.4772659912109375, 0.4777287292480469, 0.4779949645996094, 0.47825204467773436, 0.4779581298828125, 0.4773775329589844, 0.4774225769042969, 0.47868109130859376, 0.4778076171875, 0.47798785400390625, 0.478139404296875, 0.47792538452148436, 0.4779366455078125, 0.4774615173339844, 0.47746356201171875, 0.477728759765625, 0.47751473999023436, 0.47717376708984377, 0.47695462036132813, 0.4774942626953125, 0.47761920166015626, 0.4774625244140625, 0.4780001220703125, 0.4778526611328125, 0.47788134765625, 0.4775301513671875, 0.47762127685546873, 0.4773498229980469, 0.4779581298828125, 0.4775413818359375, 0.47721881103515623, 0.47748504638671874, 0.4778824462890625, 0.47803488159179686, 0.47796633911132813, 0.47795098876953124, 0.47843328857421874, 0.4778895263671875, 0.47771749877929687, 0.47763766479492187, 0.47745123291015623, 0.47731610107421873, 0.47720550537109374, 0.47810763549804686, 0.47794686889648436, 0.478023681640625, 0.47817214965820315, 0.4775679931640625, 0.4772167663574219, 0.4781783142089844, 0.4780738525390625, 0.4782264404296875, 0.4780769348144531, 0.4779223022460938, 0.4776662902832031, 0.47806668090820315, 0.4780738525390625, 0.47803903198242187, 0.9936957397460937, 0.47736831665039064, 0.4772413330078125, 0.47760589599609377, 0.47733966064453126, 0.4780892028808594, 0.47794790649414065, 0.4774604797363281, 0.47752294921875, 0.4775792541503906, 0.47779022216796874, 0.4776407165527344, 0.4775096435546875, 0.4774523010253906, 0.4773304443359375, 0.4777021484375, 0.477939697265625, 0.4781055908203125, 0.4771205749511719, 0.477898681640625, 0.47731201171875, 0.47782196044921876, 0.4773775329589844, 0.4771747741699219, 0.4777697143554688, 0.47787213134765627, 0.4778188781738281, 0.4779683837890625, 0.47738983154296877, 0.4772915344238281, 0.4775802917480469, 0.47765914916992186, 0.47731201171875, 0.47755877685546877, 0.4773355407714844, 0.4775382995605469, 0.477918212890625, 0.47747378540039065, 0.4770508728027344, 0.4778793029785156, 0.477528076171875, 0.47719937133789064, 0.4775516052246094, 0.4772341613769531, 0.4773990478515625, 0.47757415771484374, 0.47758746337890623, 0.47739495849609376, 0.47834423828125, 0.4778536682128906, 0.4773058776855469, 0.47772262573242186, 0.4777635498046875, 0.4777185363769531, 0.4785080261230469, 0.4780687255859375, 0.4775168151855469, 0.4807618408203125, 0.47829605102539063, 0.47767245483398435, 0.4783206481933594, 0.4779980773925781, 0.4780625915527344, 0.9916989135742188, 0.47712973022460936, 0.4770652160644531, 0.4772781982421875, 0.47782296752929687, 0.47800833129882814, 0.47793356323242187, 0.47763150024414064, 0.47709390258789064, 0.47777484130859377, 0.4777667236328125, 0.47767550659179686, 0.47757720947265625, 0.4771798400878906, 0.47727410888671873, 0.47779531860351565, 0.4776089477539063, 0.4775792541503906, 0.47766937255859376, 0.47783627319335936, 0.47713690185546875, 0.47779226684570314, 0.4778567810058594, 0.4774615173339844, 0.4775833740234375, 0.47805950927734375, 0.47846502685546877, 0.478561279296875, 0.4782633056640625, 0.4781055908203125, 0.47865753173828124, 0.47780044555664064, 0.47805950927734375, 0.4787906494140625, 0.47858688354492185, 0.47910400390625, 0.47804107666015627, 0.4773498840332031, 0.47763455200195315, 0.4782264404296875, 0.47779736328125, 0.47735809326171874, 0.4779683837890625, 0.4779571228027344, 0.47788134765625, 0.47783627319335936, 0.4788439025878906, 0.47873126220703127, 0.47824075317382814, 0.47784039306640624, 0.47777279663085936, 0.47869952392578125, 0.4775126953125, 0.47748504638671874, 0.4778311767578125, 0.47821005249023435, 0.478708740234375, 0.4786903076171875, 0.4786278381347656, 0.4780533752441406, 0.47773080444335936, 0.47787213134765627, 0.47809127807617186, 0.9952440185546875, 0.4780052490234375, 0.47796734619140624, 0.47861553955078123, 0.47840972900390627, 0.478055419921875, 0.4771358642578125, 0.47734375, 0.4771768188476562, 0.47729766845703125, 0.477949951171875, 0.4772239379882813, 0.4777277526855469, 0.47779531860351565, 0.47714407348632815, 0.47731814575195314, 0.47762841796875, 0.47755877685546877, 0.477528076171875, 0.4772034606933594, 0.47703143310546875, 0.47779531860351565, 0.4779202575683594, 0.4790067138671875, 0.478593017578125, 0.4792412109375, 0.47822540283203124, 0.4779069519042969, 0.4779427795410156, 0.47770932006835937, 0.47803802490234376, 0.4777840576171875, 0.47797247314453123, 0.477939697265625, 0.47781991577148436, 0.4778240051269531, 0.47803289794921877, 0.47818240356445313, 0.47798175048828123, 0.4779959716796875, 0.4778680419921875, 0.4780902404785156, 0.47819674682617186, 0.47809945678710936, 0.4779765625, 0.479388671875, 0.4785356750488281, 0.4774143981933594, 0.4776642456054688, 0.47775238037109374, 0.4779253234863281, 0.4778547058105469, 0.4781363220214844, 0.47750860595703126, 0.4780830688476562, 0.47821728515625, 0.4779703674316406, 0.47782196044921876, 0.4780851135253906, 0.4781322326660156, 0.47786392211914064, 0.4776509399414062, 0.4776099853515625, 0.9929461669921875, 0.47778509521484375, 0.48122674560546874, 0.47778509521484375, 0.4777943115234375, 0.47731814575195314, 0.4780748901367188, 0.47792538452148436, 0.47784756469726564, 0.4781158447265625, 0.4780155029296875, 0.47748403930664063, 0.4777891845703125, 0.4779100036621094, 0.4776908874511719, 0.47778713989257815, 0.47860736083984373, 0.47720037841796875, 0.4777891845703125, 0.4777216491699219, 0.4775577087402344, 0.47795404052734375, 0.47750143432617187, 0.47806668090820315, 0.4786780090332031, 0.477744140625, 0.47708673095703125, 0.47796429443359373, 0.4774686584472656, 0.4776775817871094, 0.4779796447753906, 0.47754855346679687, 0.4775096435546875, 0.4782417907714844, 0.47757107543945315, 0.4774205322265625, 0.4781158447265625, 0.4779376525878906, 0.47806362915039063, 0.47810763549804686, 0.4782090148925781, 0.478413818359375, 0.47833804321289064, 0.4782417907714844, 0.4790927429199219, 0.4781537170410156, 0.47856845092773437, 0.4775301208496094, 0.4781506652832031, 0.4774615173339844, 0.4780472412109375, 0.4780318603515625, 0.47794891357421876, 0.47838311767578123, 0.47825408935546876, 0.4780349426269531, 0.47770111083984373, 0.4779857788085938, 0.47789266967773436, 0.47788128662109375, 0.4787712097167969, 0.47795306396484377, 0.47804415893554686, 0.9943838500976563, 0.4787189636230469, 0.47804107666015627, 0.47781375122070313, 0.477370361328125, 0.4773304443359375, 0.47816705322265624, 0.47811993408203124, 0.4777738342285156, 0.47763250732421875, 0.4774573974609375, 0.47708673095703125, 0.47772467041015626, 0.47725567626953125, 0.47783627319335936, 0.47736831665039064, 0.4771112976074219, 0.478445556640625, 0.47864935302734374, 0.4780984191894531, 0.47761102294921876, 0.4776365966796875, 0.4789073791503906, 0.4771778564453125, 0.47777484130859377, 0.47784756469726564, 0.4784169006347656, 0.4775045166015625, 0.47752191162109375, 0.47713177490234376, 0.47731610107421873, 0.47756390380859376, 0.4773212280273437, 0.47744512939453126, 0.4777543640136719, 0.477222900390625, 0.477707275390625, 0.47724850463867186, 0.47723724365234377, 0.4778823547363281, 0.477412353515625, 0.4772392883300781, 0.4777994384765625, 0.47849676513671874, 0.4779949951171875, 0.47791104125976563, 0.47725466918945314, 0.47723629760742187, 0.47771026611328127, 0.478139404296875, 0.4780195922851562, 0.478129150390625, 0.47857867431640627, 0.47747174072265625, 0.47793869018554686, 0.47770932006835937, 0.4779571228027344, 0.47765509033203124, 0.4775464172363281, 0.47778713989257815, 0.477918212890625, 0.4777820129394531, 0.47804107666015627, 0.9942221069335937, 0.47767550659179686, 0.4771686401367187, 0.477431884765625, 0.4771603698730469, 0.47666278076171875, 0.4774676513671875, 0.47758746337890623, 0.47736114501953125, 0.4776488952636719, 0.47701708984375, 0.47684402465820314, 0.47745944213867186, 0.477633544921875, 0.4778076171875, 0.4772536315917969, 0.47761715698242185, 0.47721881103515623, 0.4775362548828125, 0.47752294921875, 0.4779653015136719, 0.47805645751953124, 0.47783526611328125, 0.47724032592773435, 0.4773447570800781, 0.4780175476074219, 0.4793231506347656, 0.4781588439941406, 0.47761715698242185, 0.47719937133789064, 0.4781537170410156, 0.47752294921875, 0.4775311279296875, 0.47761920166015626, 0.4774993896484375, 0.4768399353027344, 0.4775946960449219, 0.47729248046875, 0.47762841796875, 0.4775372924804687, 0.4772720642089844, 0.4776089477539063, 0.4787261352539062, 0.47772262573242186, 0.47747378540039065, 0.47768780517578124, 0.477528076171875, 0.477149169921875, 0.4776355895996094, 0.47729766845703125, 0.4775628662109375, 0.47870361328125, 0.4778188781738281, 0.4776509399414062, 0.4779898986816406, 0.4782561340332031, 0.47833087158203125, 0.47798681640625, 0.4781035461425781, 0.477955078125, 0.47816192626953125, 0.4783472595214844, 0.47877734375]",tokens/s,2.061137207272146,,,,, -4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-1.3B,,cuda,0,42,,,,,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,main,False,False,MB,1554.096128,1957.167104,0.0,1310.72,1163.82464,s,10,1.286997444152832,0.1286997444152832,0.0010709154419692667,0.12870168304443358,0.13024390106201172,0.13026507034301757,0.13028200576782226,"[0.13028623962402344, 0.13023919677734375, 0.1282391662597656, 0.12805363464355468, 0.1269244155883789, 0.12729046630859375, 0.12865869140625, 0.1287446746826172, 0.1295413055419922, 0.1290196533203125]",tokens/s,1989.1259393177147,kWh,1.4998026454163673e-06,8.218217431655106e-07,6.2844740852399835e-06,8.60609847382186e-06,tokens/kWh,29746347.985524923,MB,1554.096128,1959.264256,0.0,1312.817152,1232.774656,s,10,75.16664355468751,7.516664355468751,0.019324999879587898,7.517852294921875,7.5454459960937506,7.548604345703125,7.551131025390625,"[7.5517626953125, 7.5180322265625, 7.52255224609375, 7.52390966796875, 7.544744140625, 7.51767236328125, 7.49461669921875, 7.49106689453125, 7.50185205078125, 7.5004345703125]",tokens/s,8.3813772999142,kWh,8.89009176429398e-05,4.872418928564946e-05,0.00036540068437377195,0.0005030257913023612,tokens/kWh,125242.08716393956,,s,629,76.20247959136968,0.12114861620249542,0.01533479707935131,0.1192816619873047,0.12045271148681641,0.12072119903564453,0.24720068908691412,"[0.12236902618408203, 0.12192870330810547, 0.12110336303710938, 0.12021862030029297, 0.12056371307373047, 0.11989708709716797, 0.11929497528076172, 0.11829862213134766, 0.11859661102294922, 0.11861504364013672, 0.11821469116210938, 0.11968099212646484, 0.11959193420410157, 0.12045926666259765, 0.12021043395996094, 0.1199452133178711, 0.11970047760009765, 0.12003225708007813, 0.120163330078125, 0.11992678070068359, 0.11977011108398437, 0.11847372436523437, 0.11856588745117187, 0.11883724975585938, 0.11842662048339844, 0.11851676940917968, 0.11948745727539062, 0.11911475372314453, 0.119552001953125, 0.1215068130493164, 0.120257568359375, 0.1204623031616211, 0.12012236785888672, 0.12004659271240234, 0.12008038330078125, 0.12001689910888672, 0.11995954895019531, 0.12004761505126953, 0.12005683135986328, 0.12010189056396485, 0.12003635406494141, 0.11968307495117188, 0.12004863739013671, 0.12036300659179687, 0.11988992309570312, 0.1200558090209961, 0.1202534408569336, 0.12022169494628906, 0.12032921600341796, 0.12026265716552734, 0.12077056121826171, 0.12018585968017578, 0.11901235198974609, 0.1193861083984375, 0.11910348510742187, 0.11925606536865234, 0.11975270080566407, 0.12020941162109375, 0.11999846649169922, 0.12033843231201172, 0.11914854431152344, 0.12003942108154297, 0.2506158142089844, 0.12027391815185547, 0.11959091186523438, 0.11991551971435546, 0.1192816619873047, 0.11801292419433594, 0.11855052947998047, 0.11803955078125, 0.11798016357421875, 0.11803241729736329, 0.11873174285888671, 0.12019200134277344, 0.11996057891845703, 0.11952947235107422, 0.11857612609863281, 0.11973939514160156, 0.11993702697753907, 0.11889049530029297, 0.11839385223388672, 0.11943730926513672, 0.11985203552246093, 0.11852082824707032, 0.11822898864746094, 0.11862739562988281, 0.11842758178710938, 0.11852185821533204, 0.1185054702758789, 0.11835596466064453, 0.11835699462890625, 0.11841331481933594, 0.11834674835205078, 0.11860070037841797, 0.12062003326416015, 0.11992473602294922, 0.12043059539794922, 0.12157746887207031, 0.12022579193115235, 0.11964927673339844, 0.12050534057617188, 0.12049407958984375, 0.12015615844726563, 0.11910553741455078, 0.12046438598632812, 0.1204510726928711, 0.12023603057861328, 0.11975885009765624, 0.12176998138427735, 0.12023705291748046, 0.12054937744140624, 0.11995750427246094, 0.12132864379882813, 0.11925917053222657, 0.1186283187866211, 0.11835699462890625, 0.11851776123046875, 0.11839794921875, 0.11879730987548828, 0.11951107025146485, 0.11866006469726563, 0.11900109100341796, 0.11836723327636718, 0.11853414154052734, 0.11843583679199218, 0.24882893371582032, 0.11982233428955077, 0.11939225769042969, 0.12049919891357422, 0.12002508544921875, 0.1198919677734375, 0.12023910522460937, 0.11994217681884765, 0.11870614624023437, 0.11836006164550782, 0.11821260833740234, 0.11851570892333985, 0.11971686553955078, 0.12000460815429688, 0.11984076690673828, 0.11931136322021485, 0.11905126190185547, 0.11986329650878906, 0.11975373077392579, 0.11980595397949219, 0.1196267547607422, 0.11956735992431641, 0.12024422454833984, 0.12106342315673828, 0.11977728271484375, 0.1197127685546875, 0.11865599822998046, 0.11831404876708984, 0.11899692535400391, 0.11852082824707032, 0.11829145812988281, 0.11852595520019531, 0.11846348571777343, 0.11836518096923829, 0.11844812774658203, 0.11833757019042969, 0.11830780792236328, 0.11964620971679688, 0.11897650909423828, 0.11840921783447265, 0.11968511962890625, 0.11841024017333984, 0.11969535827636718, 0.11970559692382812, 0.12029644775390624, 0.12001078033447266, 0.11961955261230468, 0.12051967620849609, 0.11861196899414063, 0.11842150115966797, 0.11837133026123046, 0.12028313446044922, 0.11990322875976563, 0.1188751983642578, 0.12033939361572266, 0.11918643188476563, 0.1187583999633789, 0.12003942108154297, 0.11934515380859376, 0.12007730865478515, 0.12025138854980469, 0.12019916534423829, 0.12015615844726563, 0.24848793029785157, 0.11997798156738282, 0.11991763305664062, 0.12037625885009766, 0.12008345794677734, 0.11944652557373046, 0.12011827087402344, 0.12008448028564453, 0.11987149047851563, 0.12019404602050782, 0.12016851043701172, 0.11977823638916016, 0.11987865447998047, 0.11973426818847656, 0.12008038330078125, 0.11942092895507812, 0.11997491455078126, 0.11880242919921875, 0.11979673767089843, 0.1197127685546875, 0.11939532470703125, 0.11976703643798828, 0.11997494506835937, 0.12019094085693359, 0.12064665222167968, 0.12001996612548828, 0.12060364532470703, 0.12040294647216797, 0.11997388458251954, 0.11948646545410156, 0.11998822021484375, 0.11993907165527344, 0.11955097961425781, 0.11953561401367188, 0.11927756500244141, 0.11897036743164062, 0.11999334716796875, 0.11905126190185547, 0.11846553802490234, 0.11841843414306641, 0.11854847717285157, 0.11835596466064453, 0.11830995178222656, 0.11828428649902344, 0.1203987808227539, 0.11818905639648437, 0.1193707504272461, 0.11814604949951171, 0.11847782135009766, 0.1183078384399414, 0.11799350738525391, 0.11804771423339844, 0.11817372894287109, 0.11848700714111328, 0.11852082824707032, 0.11849932861328125, 0.12020735931396484, 0.120089599609375, 0.12004761505126953, 0.12012134552001953, 0.11869286346435547, 0.11855974578857421, 0.11863142395019531, 0.24572006225585938, 0.1183477783203125, 0.11846041870117187, 0.11811737823486328, 0.1181470718383789, 0.11830067443847657, 0.11928985595703125, 0.11911065673828125, 0.1185771484375, 0.11835187530517578, 0.12018994903564453, 0.12008550262451172, 0.11990534210205078, 0.1200219497680664, 0.12004659271240234, 0.11994931030273437, 0.12035686492919923, 0.12041522979736329, 0.12053298950195312, 0.12031999969482422, 0.12058419036865234, 0.12013158416748047, 0.12075929260253906, 0.11833036804199219, 0.1206794204711914, 0.12085862731933594, 0.12161433410644532, 0.11919667053222656, 0.1187041244506836, 0.11844300842285156, 0.11833650970458984, 0.11833856201171875, 0.1186355209350586, 0.11820236968994141, 0.12034559631347656, 0.12019712066650391, 0.12016947174072265, 0.11850035095214843, 0.12005375671386719, 0.12051763153076171, 0.11887615966796874, 0.11807539367675782, 0.11823411560058594, 0.11905741119384766, 0.12032307434082032, 0.1189775390625, 0.12029132843017579, 0.12084019470214843, 0.12052992248535156, 0.1207357406616211, 0.12054220581054688, 0.12078284454345703, 0.12055244445800781, 0.12059852600097656, 0.12064358520507812, 0.12082486724853515, 0.1207326431274414, 0.12082278442382813, 0.12053196716308594, 0.12073065948486328, 0.12090262603759766, 0.12064870452880859, 0.12058112335205078, 0.24601292419433593, 0.11817881774902343, 0.11808255767822265, 0.11800780487060547, 0.118255615234375, 0.11830067443847657, 0.11829145812988281, 0.11941478729248046, 0.12005375671386719, 0.1198202896118164, 0.12005785369873047, 0.11806412506103516, 0.11867750549316407, 0.11829145812988281, 0.11833548736572265, 0.11830989074707031, 0.11806822204589844, 0.11833241271972657, 0.11837747192382812, 0.11845120239257813, 0.11834572601318359, 0.11841228485107422, 0.11827609252929687, 0.11805286407470703, 0.11896729278564454, 0.11836313629150391, 0.11981619262695313, 0.12392044830322266, 0.1205656967163086, 0.12015821075439453, 0.11990016174316406, 0.12003533172607422, 0.11934003448486329, 0.1199974365234375, 0.1193154525756836, 0.11959500885009766, 0.11863040161132812, 0.1188106231689453, 0.12067327880859376, 0.11914444732666016, 0.11830477142333984, 0.11853721618652344, 0.1185269775390625, 0.11941990661621094, 0.11858329772949219, 0.1186693115234375, 0.12036608123779297, 0.12004557037353515, 0.1200558090209961, 0.12050841522216797, 0.12022681427001954, 0.12039065551757812, 0.12011212921142578, 0.12010495758056641, 0.12036608123779297, 0.1201817626953125, 0.12043673706054688, 0.12033843231201172, 0.12007526397705078, 0.12000972747802735, 0.12006505584716796, 0.11996975708007812, 0.11998822021484375, 0.2494791717529297, 0.12030156707763671, 0.11901952362060547, 0.11824127960205078, 0.1192816619873047, 0.11887308502197266, 0.11826175689697266, 0.11869388580322265, 0.11826687622070313, 0.11836211395263672, 0.11810918426513672, 0.11887104034423829, 0.11972198486328126, 0.11960012817382812, 0.1192468490600586, 0.118002685546875, 0.11875942230224609, 0.11941478729248046, 0.11890585327148437, 0.11843382263183594, 0.1191526107788086, 0.11959500885009766, 0.11970668792724609, 0.11957036590576171, 0.11933695983886719, 0.11909327697753906, 0.11842966461181641, 0.11905228424072266, 0.11911888122558593, 0.11894780731201172, 0.11848089599609375, 0.1181890869140625, 0.11840406036376953, 0.1190297622680664, 0.11982848358154297, 0.11972096252441407, 0.11864371490478516, 0.11924582672119141, 0.11929804992675781, 0.11825459289550781, 0.11951718139648437, 0.11916287994384765, 0.11947417449951171, 0.11800064086914062, 0.11845017242431641, 0.11818089294433594, 0.11930210876464843, 0.1199810562133789, 0.11853517150878906, 0.11870310211181641, 0.11945164489746093, 0.11844096374511719, 0.1184194564819336, 0.118508544921875, 0.11926220703125, 0.12017356872558593, 0.11892838287353516, 0.1193021469116211, 0.11871849822998047, 0.11844911956787109, 0.11832115173339844, 0.11867545318603516, 0.11844403076171875, 0.24766259765625, 0.11916802978515625, 0.12123133087158203, 0.12028313446044922, 0.11920588684082031, 0.11884953308105468, 0.11862528228759765, 0.11994931030273437, 0.11886182403564453, 0.11885363006591797, 0.12013670349121094, 0.11913011169433593, 0.11992473602294922, 0.11888435363769531, 0.11946598052978516, 0.11882707214355469, 0.1207070083618164, 0.11865702056884765, 0.11839590454101563, 0.11839078521728516, 0.11947724914550781, 0.11919769287109375, 0.11886080169677735, 0.11848601531982422, 0.11876966094970703, 0.11810099029541016, 0.11880960083007812, 0.11817369842529298, 0.11927142333984375, 0.11965235137939453, 0.118761474609375, 0.11863654327392578, 0.11901439666748047, 0.11827814483642578, 0.11898368072509766, 0.11854233551025391, 0.11844915008544922, 0.11848397064208985, 0.11839078521728516, 0.11844608306884766, 0.11859455871582031, 0.11865190124511718, 0.11820441436767579, 0.11907686614990234, 0.11970867156982422, 0.11921715545654298, 0.11824230194091796, 0.11921817779541016, 0.11916390228271484, 0.11929804992675781, 0.11828326416015625, 0.11838361358642578, 0.11848499298095704, 0.11860991668701172, 0.11808153533935548, 0.11826588439941406, 0.11823715209960937, 0.11888639831542969, 0.11884031677246094, 0.11812454223632812, 0.11833446502685546, 0.11847577667236328, 0.11946905517578126, 0.24866712951660155, 0.11849318695068359, 0.11980188751220704, 0.11915773010253906, 0.11846041870117187, 0.1203589096069336, 0.1205381088256836, 0.11978342437744141, 0.11976601409912109, 0.11947417449951171, 0.11962163543701172, 0.11930316925048828, 0.11854541015625, 0.11897344207763672, 0.11885158538818359, 0.11875020599365234, 0.1181839370727539, 0.11931443023681641, 0.11823616027832032, 0.11900313568115234, 0.11816345977783203, 0.11854847717285157, 0.11989708709716797, 0.11902668762207032, 0.11941683197021484, 0.11941580963134765, 0.11852185821533204, 0.11982540893554687, 0.11974553680419922, 0.11874918365478515, 0.11918950653076171, 0.11970355224609375, 0.11946189117431641, 0.11857100677490234, 0.11914035034179687, 0.11940966033935548, 0.1185689926147461, 0.11831394958496094, 0.11820441436767579, 0.11945574188232422, 0.11997388458251954, 0.1203240966796875, 0.11903180694580077, 0.11818495941162109, 0.11810201263427735, 0.11788288116455078, 0.11807539367675782, 0.11814604949951171, 0.11880038452148438, 0.11919872283935547, 0.11881983947753906, 0.11975987243652343, 0.11975373077392579, 0.11953568267822266, 0.12003424072265625, 0.11991961669921875, 0.12044185638427735, 0.11880550384521485, 0.11889356994628907, 0.11842253112792969, 0.11843379211425781, 0.11815936279296875, 0.11811634826660156, 0.24846543884277345, 0.11927139282226562, 0.11798732757568359, 0.11792998504638671, 0.11840512084960937, 0.11863859558105469, 0.11815321350097656, 0.11805184173583984, 0.11871437072753906, 0.11836109161376954, 0.1180794906616211, 0.11815424346923828, 0.1181317138671875, 0.11853004455566406, 0.11842150115966797, 0.11837542724609375, 0.11854847717285157, 0.11832217407226563, 0.11862220764160156, 0.11864985656738282, 0.11943526458740235, 0.11942912292480469, 0.11993395233154297, 0.12021145629882812, 0.12013568115234376, 0.12069990539550782, 0.12175667572021484, 0.12003942108154297, 0.11987763214111329, 0.11907788848876953, 0.1194076156616211, 0.11910553741455078, 0.11873177337646484, 0.11934413146972657, 0.11961958312988281, 0.11883519744873047, 0.11877273559570313, 0.11876761627197266, 0.12034662628173828, 0.12008140563964843, 0.11964415740966797, 0.12007014465332032, 0.12065280151367187, 0.11880754852294922, 0.11892736053466797, 0.11855974578857421, 0.11855052947998047, 0.11903590393066406, 0.11845938873291016, 0.11849523162841796, 0.1184716796875, 0.11835289764404297, 0.11853619384765625, 0.11834368133544922, 0.11937382507324219, 0.11912806701660156, 0.11848806762695313, 0.11961753845214844, 0.1190113296508789, 0.11965132904052735, 0.11860281372070312, 0.11934611511230468, 0.11991145324707031]",tokens/s,8.254324575433342,,,,, -4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-34B,,cuda,0,42,,,,,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,main,False,False,MB,1743.122432,22129.672192,0.0,21483.225088,20799.036928,s,10,27.909502197265624,2.7909502197265623,0.0030225550178342967,2.7914050292968753,2.79439267578125,2.794928149414062,2.7953565283203123,"[2.794273681640625, 2.795463623046875, 2.7873525390625, 2.789448486328125, 2.787631103515625, 2.78628466796875, 2.790373291015625, 2.792436767578125, 2.79304931640625, 2.793188720703125]",tokens/s,91.72503263962948,kWh,3.290233454770512e-05,1.803171549396211e-05,0.0001583880155991868,0.00020932206564085407,tokens/kWh,1222995.765956342,MB,1743.720448,22129.672192,0.0,21483.225088,20902.142976,s,10,1661.799328125,166.1799328125,0.024199823300055656,166.1772734375,166.2117125,166.21745,166.22204,"[166.17496875, 166.167359375, 166.14703125, 166.2104375, 166.2231875, 166.181296875, 166.179578125, 166.204078125, 166.159546875, 166.15184375]",tokens/s,0.37910714569299164,kWh,0.0019616428730719624,0.001075153456676562,0.009382404450362024,0.012419200780110549,tokens/kWh,5072.790199261051,,s,629,1684.2762666015624,2.677704716377683,0.33156351074934576,2.63752294921875,2.6395720703125,2.6405790039062498,5.428127246093751,"[2.63648046875, 2.637477783203125, 2.636610595703125, 2.63773388671875, 2.63674365234375, 2.6374482421875, 2.63684716796875, 2.637740966796875, 2.636241943359375, 2.6376162109375, 2.636966796875, 2.6365615234375, 2.639180908203125, 2.637003662109375, 2.637414306640625, 2.638074951171875, 2.638488525390625, 2.637603759765625, 2.6381865234375, 2.6367314453125, 2.63737646484375, 2.636360595703125, 2.63697314453125, 2.6366044921875, 2.6370693359375, 2.636568603515625, 2.636980224609375, 2.636353515625, 2.63706103515625, 2.636314697265625, 2.637646728515625, 2.63864208984375, 2.637822998046875, 2.63769091796875, 2.636580810546875, 2.636631103515625, 2.6366474609375, 2.636380126953125, 2.635778076171875, 2.639795166015625, 2.637285400390625, 2.636735595703125, 2.63680419921875, 2.636505126953125, 2.636515380859375, 2.637959228515625, 2.6366884765625, 2.638927001953125, 2.6396015625, 2.6393056640625, 2.640069580078125, 2.638898193359375, 2.6396181640625, 2.639129638671875, 2.6404248046875, 2.63950439453125, 2.6398291015625, 2.639203369140625, 2.639447021484375, 2.638824462890625, 2.638972900390625, 2.637701171875, 5.43399951171875, 2.63682861328125, 2.637526123046875, 2.637077392578125, 2.636991455078125, 2.636851318359375, 2.63690966796875, 2.636982177734375, 2.638088134765625, 2.63716455078125, 2.63853662109375, 2.638309326171875, 2.63889111328125, 2.64060107421875, 2.637055908203125, 2.63606884765625, 2.63802880859375, 2.637928466796875, 2.6390927734375, 2.6384384765625, 2.637506591796875, 2.63756298828125, 2.63665869140625, 2.636583984375, 2.63687060546875, 2.6370908203125, 2.637446044921875, 2.63802685546875, 2.6375556640625, 2.637516845703125, 2.63621728515625, 2.637552734375, 2.637285400390625, 2.637663330078125, 2.6371767578125, 2.637763671875, 2.63699560546875, 2.63773583984375, 2.637076416015625, 2.637444091796875, 2.636642333984375, 2.636317626953125, 2.63705078125, 2.636695556640625, 2.637413330078125, 2.63615478515625, 2.6434365234375, 2.637322265625, 2.638057373046875, 2.63786181640625, 2.638676025390625, 2.637177734375, 2.63716455078125, 2.637470703125, 2.63781689453125, 2.636496826171875, 2.63722802734375, 2.637602783203125, 2.638507080078125, 2.637663330078125, 2.636621826171875, 2.63727001953125, 2.6375107421875, 5.42925830078125, 2.638592041015625, 2.6379560546875, 2.637390869140625, 2.636768310546875, 2.63889501953125, 2.638235595703125, 2.638288818359375, 2.637874267578125, 2.638148681640625, 2.637728759765625, 2.637327392578125, 2.63690869140625, 2.637347900390625, 2.6369453125, 2.637433837890625, 2.63707958984375, 2.636822509765625, 2.637656982421875, 2.63617529296875, 2.636454833984375, 2.6358466796875, 2.636423095703125, 2.63856640625, 2.63604833984375, 2.63657470703125, 2.637000732421875, 2.637095947265625, 2.636664794921875, 2.6381435546875, 2.640372802734375, 2.638454833984375, 2.636282958984375, 2.636675048828125, 2.635629638671875, 2.636507080078125, 2.63680615234375, 2.63701611328125, 2.636884033203125, 2.63752294921875, 2.6366669921875, 2.63699462890625, 2.63780859375, 2.6374072265625, 2.636739501953125, 2.6374296875, 2.637625244140625, 2.639678466796875, 2.63699658203125, 2.63729052734375, 2.637656982421875, 2.636547119140625, 2.6364794921875, 2.636135498046875, 2.636613525390625, 2.636599365234375, 2.63747265625, 2.6357861328125, 2.63600439453125, 2.6359716796875, 2.636907470703125, 2.636198974609375, 2.639195068359375, 5.42974560546875, 2.637918212890625, 2.638065673828125, 2.6382080078125, 2.639459228515625, 2.638834716796875, 2.640337890625, 2.639097900390625, 2.638284912109375, 2.639035400390625, 2.636672119140625, 2.63703955078125, 2.63772265625, 2.639283203125, 2.6380830078125, 2.639520751953125, 2.63948291015625, 2.637526123046875, 2.638147705078125, 2.637602783203125, 2.63681640625, 2.63762841796875, 2.638529541015625, 2.63696484375, 2.63714599609375, 2.637178955078125, 2.6376796875, 2.63855810546875, 2.641314697265625, 2.637876220703125, 2.63853369140625, 2.63912548828125, 2.637345703125, 2.636560302734375, 2.636971923828125, 2.636669921875, 2.636619873046875, 2.6392392578125, 2.63902099609375, 2.638465087890625, 2.63775341796875, 2.638636962890625, 2.637673583984375, 2.6372373046875, 2.64167431640625, 2.6387353515625, 2.637962158203125, 2.639416259765625, 2.637681640625, 2.639564697265625, 2.6377236328125, 2.638668701171875, 2.638095458984375, 2.63796728515625, 2.637655029296875, 2.637210693359375, 2.63780859375, 2.638158935546875, 2.637846435546875, 2.63988232421875, 2.641334228515625, 2.636745849609375, 2.63619482421875, 5.42609521484375, 2.63847412109375, 2.64098095703125, 2.6405458984375, 2.64087451171875, 2.64148388671875, 2.64074755859375, 2.64026318359375, 2.640604248046875, 2.640962646484375, 2.638622802734375, 2.63895458984375, 2.636958740234375, 2.636030029296875, 2.636072021484375, 2.637117431640625, 2.636577880859375, 2.642241455078125, 2.64078857421875, 2.641033203125, 2.641383544921875, 2.64137841796875, 2.639701904296875, 2.640848876953125, 2.640034912109375, 2.6397451171875, 2.640384033203125, 2.637106201171875, 2.636432373046875, 2.636267578125, 2.637424560546875, 2.636958740234375, 2.63680712890625, 2.637274169921875, 2.638011474609375, 2.63804931640625, 2.63707861328125, 2.636590087890625, 2.63739794921875, 2.63689013671875, 2.63980029296875, 2.639002685546875, 2.638836669921875, 2.63720849609375, 2.637477783203125, 2.637814697265625, 2.637075439453125, 2.637104248046875, 2.638521240234375, 2.63874755859375, 2.63686962890625, 2.63682763671875, 2.637802490234375, 2.63747998046875, 2.6362255859375, 2.63657568359375, 2.636940185546875, 2.642130859375, 2.6372998046875, 2.63684912109375, 2.63803271484375, 2.63796533203125, 2.637073486328125, 5.4244033203125, 2.636745849609375, 2.6364814453125, 2.6371123046875, 2.637408203125, 2.636239990234375, 2.63697607421875, 2.638215087890625, 2.639066162109375, 2.63918994140625, 2.639048583984375, 2.63975634765625, 2.639510498046875, 2.63817626953125, 2.637822021484375, 2.6385029296875, 2.637655029296875, 2.637551513671875, 2.636876708984375, 2.63750048828125, 2.63714599609375, 2.63798779296875, 2.63727099609375, 2.637442138671875, 2.636712890625, 2.637854736328125, 2.638200927734375, 2.6394736328125, 2.63754541015625, 2.636971923828125, 2.637844482421875, 2.63727197265625, 2.6361865234375, 2.636801025390625, 2.637619140625, 2.637947998046875, 2.63809228515625, 2.637560791015625, 2.6417490234375, 2.63729052734375, 2.63712158203125, 2.63689111328125, 2.63716259765625, 2.635864990234375, 2.6373251953125, 2.63663623046875, 2.63742578125, 2.637765625, 2.63775341796875, 2.638487548828125, 2.638257080078125, 2.63924951171875, 2.639295654296875, 2.639500244140625, 2.63644970703125, 2.637189208984375, 2.637624267578125, 2.639413330078125, 2.639233154296875, 2.63777587890625, 2.637401123046875, 2.637765625, 2.63680712890625, 5.42891748046875, 2.637751220703125, 2.6369423828125, 2.637287353515625, 2.6384208984375, 2.6376162109375, 2.637705078125, 2.6376396484375, 2.636984375, 2.636735595703125, 2.637454345703125, 2.63906005859375, 2.641005615234375, 2.639869873046875, 2.638488525390625, 2.638159912109375, 2.638043212890625, 2.6367958984375, 2.6363330078125, 2.637013916015625, 2.63684716796875, 2.63693115234375, 2.636872802734375, 2.64043212890625, 2.636421142578125, 2.63861962890625, 2.63849169921875, 2.638710693359375, 2.637824951171875, 2.6379111328125, 2.636610595703125, 2.6375966796875, 2.637263916015625, 2.63708984375, 2.637641845703125, 2.637759521484375, 2.6372197265625, 2.63781591796875, 2.63773583984375, 2.638950439453125, 2.637347900390625, 2.63666796875, 2.638234619140625, 2.638180419921875, 2.637719482421875, 2.63872216796875, 2.639690673828125, 2.63703662109375, 2.636780517578125, 2.636748779296875, 2.637224853515625, 2.636233642578125, 2.636378173828125, 2.636863525390625, 2.63923193359375, 2.638593994140625, 2.638180419921875, 2.63895556640625, 2.638899169921875, 2.637263916015625, 2.63600537109375, 2.637727783203125, 2.63825, 5.43155078125, 2.637918212890625, 2.63883056640625, 2.63889208984375, 2.63790185546875, 2.637695068359375, 2.636559326171875, 2.636966796875, 2.63714208984375, 2.639033447265625, 2.6404404296875, 2.640530517578125, 2.64198046875, 2.640649169921875, 2.637720458984375, 2.637758544921875, 2.639655029296875, 2.63736328125, 2.63708251953125, 2.636771240234375, 2.636508056640625, 2.63598583984375, 2.6370068359375, 2.636777587890625, 2.636590087890625, 2.63606884765625, 2.63714111328125, 2.63913671875, 2.637537353515625, 2.63705712890625, 2.636992431640625, 2.638615478515625, 2.636080078125, 2.637097900390625, 2.636854248046875, 2.640509033203125, 2.63925146484375, 2.6389052734375, 2.6383544921875, 2.63872705078125, 2.638085205078125, 2.6382275390625, 2.637551513671875, 2.637571044921875, 2.637200439453125, 2.63806982421875, 2.637642822265625, 2.637718505859375, 2.6373642578125, 2.6366044921875, 2.638297119140625, 2.63918798828125, 2.63943896484375, 2.6392646484375, 2.639478759765625, 2.637856689453125, 2.638085205078125, 2.638904296875, 2.6383564453125, 2.639803466796875, 2.640150634765625, 2.637856689453125, 2.640256103515625, 5.4306796875, 2.637127685546875, 2.6382705078125, 2.640280517578125, 2.639151123046875, 2.639075439453125, 2.63817529296875, 2.638035888671875, 2.638784423828125, 2.639287353515625, 2.637106201171875, 2.637173828125, 2.6368388671875, 2.6376591796875, 2.635720703125, 2.63617333984375, 2.636442626953125, 2.63904150390625, 2.6368818359375, 2.635509765625, 2.636674072265625, 2.6366484375, 2.637701171875, 2.638306396484375, 2.64076806640625, 2.63686865234375, 2.6378896484375, 2.639107177734375, 2.6401259765625, 2.6394111328125, 2.63695166015625, 2.63650927734375, 2.636439453125, 2.63674267578125, 2.637043701171875, 2.637275146484375, 2.637382568359375, 2.63693115234375, 2.636669921875, 2.637486083984375, 2.636729248046875, 2.636916748046875, 2.6372802734375, 2.638066650390625, 2.637075439453125, 2.637048828125, 2.636240966796875, 2.637787109375, 2.636506103515625, 2.636660888671875, 2.63712548828125, 2.637421630859375, 2.637157470703125, 2.636517333984375, 2.637552734375, 2.636777587890625, 2.63666064453125, 2.6360966796875, 2.637487060546875, 2.636655517578125, 2.6372158203125, 2.637021240234375, 2.63826025390625, 5.43210595703125, 2.637992919921875, 2.6383369140625, 2.636921875, 2.637557861328125, 2.637305908203125, 2.63912646484375, 2.63752099609375, 2.6370693359375, 2.638551025390625, 2.6377646484375, 2.638067626953125, 2.637068359375, 2.63798876953125, 2.6374892578125, 2.63644775390625, 2.636535888671875, 2.637727783203125, 2.63798876953125, 2.637382568359375, 2.637177734375, 2.637806640625, 2.63628173828125, 2.6376171875, 2.63735302734375, 2.63743896484375, 2.6397236328125, 2.637740966796875, 2.637970458984375, 2.6363720703125, 2.636199951171875, 2.6363935546875, 2.638331787109375, 2.636755859375, 2.636908447265625, 2.636675048828125, 2.63809033203125, 2.636525634765625, 2.63657373046875, 2.63596435546875, 2.63809130859375, 2.6373671875, 2.636396484375, 2.637010009765625, 2.637177734375, 2.636538818359375, 2.637276123046875, 2.638035888671875, 2.636796875, 2.636739501953125, 2.636539794921875, 2.63828076171875, 2.63714404296875, 2.637462646484375, 2.637360107421875, 2.638277587890625, 2.636430419921875, 2.6370283203125, 2.6365234375, 2.638316650390625, 2.636391357421875, 2.636579833984375, 2.636466064453125]",tokens/s,0.3734541728532224,,,,, -4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-125m,,cuda,0,42,,,,,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,main,False,False,MB,1273.409536,921.174016,0.0,274.726912,220.646912,s,10,0.326115104675293,0.032611510467529296,0.0012484582231843272,0.03218409538269043,0.032784641265869136,0.034564048767089836,0.035987574768066403,"[0.036343456268310544, 0.03226867294311524, 0.03213779067993164, 0.032141822814941406, 0.03203395080566406, 0.03214742279052735, 0.032220767974853515, 0.03238921737670898, 0.03208832168579102, 0.032343681335449216]",tokens/s,7849.989047728858,kWh,3.834838485734976e-07,2.1012961289151069e-07,8.25654520552228e-07,1.419267982017236e-06,tokens/kWh,180374674.2994524,MB,1273.409536,921.174016,0.0,274.726912,250.722304,s,10,20.085640136718748,2.008564013671875,0.006065788108395492,2.006720520019531,2.0181213745117184,2.0185785705566404,2.018944327392578,"[2.018019775390625, 2.006443115234375, 2.0069979248046876, 2.0103148193359375, 2.0040799560546874, 2.00180322265625, 2.00428125, 2.0133179931640623, 2.0013463134765623, 2.0190357666015624]",tokens/s,31.36569189290069,kWh,2.3731767884623614e-05,1.3005590929520364e-05,4.874369069043466e-05,8.548104950457861e-05,tokens/kWh,737005.457526882,,s,629,20.34066634368894,0.03233810229521298,0.0038819833531807996,0.03179007911682129,0.0323544822692871,0.032856268310546874,0.06427897552490235,"[0.03209625625610352, 0.031882240295410154, 0.031893503189086916, 0.032045055389404296, 0.03220787048339844, 0.031922176361083986, 0.031958015441894534, 0.03196518325805664, 0.031784959793090824, 0.03223756790161133, 0.031974399566650394, 0.031955968856811526, 0.03202969741821289, 0.03302809524536133, 0.03192835235595703, 0.031831008911132816, 0.03254272079467774, 0.033226783752441404, 0.032270305633544924, 0.032140289306640625, 0.031940607070922854, 0.032451583862304685, 0.03203788757324219, 0.03188531112670898, 0.032231422424316404, 0.031936511993408204, 0.03298099136352539, 0.03213516616821289, 0.03167948722839355, 0.031526912689208986, 0.031693824768066405, 0.03172966384887695, 0.03170918464660644, 0.03180953598022461, 0.03179929542541504, 0.031628288269042966, 0.031508480072021484, 0.03155660820007324, 0.03158937644958496, 0.03198259162902832, 0.03151974487304687, 0.03146342468261719, 0.031752191543579104, 0.03203481674194336, 0.03182796859741211, 0.03228160095214844, 0.03201331329345703, 0.03220377731323242, 0.03181056022644043, 0.031865856170654294, 0.031473663330078124, 0.031682559967041016, 0.031628288269042966, 0.03277414321899414, 0.03190790367126465, 0.0320777587890625, 0.03253350448608398, 0.032473087310791016, 0.032159744262695314, 0.03166924858093262, 0.03159654426574707, 0.031492095947265625, 0.06451507568359376, 0.032, 0.03342745590209961, 0.032282623291015625, 0.03176243209838867, 0.0318525447845459, 0.03180646324157715, 0.03143065643310547, 0.03154841613769531, 0.03168870353698731, 0.031514623641967776, 0.03150643157958984, 0.03214438247680664, 0.033160190582275394, 0.03213312149047851, 0.03156684875488281, 0.031958015441894534, 0.03179110336303711, 0.03184332847595215, 0.03142963218688965, 0.03191705513000488, 0.031768575668334964, 0.03164057540893555, 0.0317573127746582, 0.03179827117919922, 0.031731712341308595, 0.03301375961303711, 0.031920127868652344, 0.03172352027893066, 0.03289395141601562, 0.031927295684814457, 0.03162521553039551, 0.03168358421325684, 0.03179110336303711, 0.031666175842285156, 0.03177369689941406, 0.031678464889526366, 0.03179315185546875, 0.031765504837036135, 0.031699968338012696, 0.03142860794067383, 0.03162419128417969, 0.031615999221801756, 0.03140505599975586, 0.03147776031494141, 0.031458303451538085, 0.03143987274169922, 0.03174399948120117, 0.0317255687713623, 0.03142758369445801, 0.03156991958618164, 0.031783935546875, 0.03183616065979004, 0.0317890567779541, 0.03189248085021973, 0.03191910362243652, 0.031932416915893554, 0.032277503967285154, 0.032382976531982424, 0.0317706241607666, 0.03181260871887207, 0.031749120712280275, 0.03174502372741699, 0.06404198455810547, 0.03170611190795898, 0.031704063415527346, 0.031628288269042966, 0.03138764762878418, 0.031698944091796875, 0.03154022407531738, 0.03150540733337402, 0.03196416091918945, 0.03177881622314453, 0.032020481109619144, 0.03214233779907227, 0.03199692726135254, 0.031780864715576174, 0.03226009750366211, 0.032115711212158206, 0.03199180793762207, 0.03156787109375, 0.031453184127807614, 0.03144908714294434, 0.0314839038848877, 0.03146649551391602, 0.031458303451538085, 0.03172147178649903, 0.031893503189086916, 0.0328007698059082, 0.03156787109375, 0.031731712341308595, 0.031849472045898435, 0.03165388870239258, 0.03178700828552246, 0.03168259239196777, 0.03179311943054199, 0.031783935546875, 0.03212083053588867, 0.03178188705444336, 0.03179724884033203, 0.033549312591552735, 0.03284889602661133, 0.031834112167358396, 0.03212492752075195, 0.03198259162902832, 0.03174297523498535, 0.031731712341308595, 0.031753215789794925, 0.03182387161254883, 0.031473663330078124, 0.03165388870239258, 0.03157196807861328, 0.031453184127807614, 0.031734783172607424, 0.03165695953369141, 0.03180953598022461, 0.03184537506103516, 0.03207372665405273, 0.032285694122314454, 0.032556095123291016, 0.0321484146118164, 0.0324956169128418, 0.03182592010498047, 0.03148287963867188, 0.03179110336303711, 0.03181056022644043, 0.06443007659912109, 0.03183001518249512, 0.03196928024291992, 0.03182694435119629, 0.03170099258422852, 0.03181056022644043, 0.0315043830871582, 0.03177881622314453, 0.03177574348449707, 0.03191910362243652, 0.0321710090637207, 0.03204403305053711, 0.03181875228881836, 0.031703039169311525, 0.03174604797363281, 0.03175526428222656, 0.031764480590820314, 0.031731712341308595, 0.03183923149108887, 0.03340390396118164, 0.032922622680664065, 0.032161792755126956, 0.03223756790161133, 0.031734783172607424, 0.03291340637207031, 0.032074752807617186, 0.03202150344848633, 0.031716352462768556, 0.03155046463012695, 0.031692800521850584, 0.03226828765869141, 0.03171737670898438, 0.03140505599975586, 0.03134976005554199, 0.03197952079772949, 0.03200307083129883, 0.031696895599365234, 0.03169484710693359, 0.03176755142211914, 0.0321710090637207, 0.03153919982910156, 0.03143987274169922, 0.03146444892883301, 0.03160371208190918, 0.03245568084716797, 0.0314337272644043, 0.03163545608520508, 0.03231846237182617, 0.03222323226928711, 0.031942655563354495, 0.03174399948120117, 0.03162112045288086, 0.03139788818359375, 0.032302078247070314, 0.03218841552734375, 0.03275161743164062, 0.0322529296875, 0.03198361587524414, 0.03178803253173828, 0.031665151596069335, 0.03129241561889649, 0.03186073684692383, 0.031955968856811526, 0.06424063873291015, 0.03157811164855957, 0.03165695953369141, 0.03165081596374512, 0.03146854400634766, 0.031693824768066405, 0.03204095840454101, 0.03246182250976563, 0.03183103942871094, 0.03218636703491211, 0.03204095840454101, 0.031784959793090824, 0.03218431854248047, 0.03177779197692871, 0.031698944091796875, 0.03156684875488281, 0.031731712341308595, 0.03239424133300781, 0.03142758369445801, 0.03172966384887695, 0.03203379058837891, 0.03207167816162109, 0.03187302398681641, 0.033006591796875, 0.03175526428222656, 0.0316180477142334, 0.031959039688110355, 0.031514623641967776, 0.031884288787841795, 0.03161497688293457, 0.031440895080566404, 0.031442943572998046, 0.03174092864990234, 0.03172966384887695, 0.03197747230529785, 0.03176755142211914, 0.03205222320556641, 0.03165695953369141, 0.03194777679443359, 0.03171327972412109, 0.031562751770019534, 0.03168972778320313, 0.031731712341308595, 0.03172966384887695, 0.03186483192443847, 0.032143360137939454, 0.03146240043640137, 0.03147776031494141, 0.03138355255126953, 0.0313753604888916, 0.03180646324157715, 0.0317255687713623, 0.03186380767822266, 0.03160678482055664, 0.032494590759277346, 0.03222528076171875, 0.032045055389404296, 0.03183206367492676, 0.03183206367492676, 0.03180851173400879, 0.03134464073181152, 0.031470592498779294, 0.031731712341308595, 0.06429388427734375, 0.03167436790466309, 0.031923200607299806, 0.03159756851196289, 0.03151667213439941, 0.031647743225097655, 0.031802400588989255, 0.03235427093505859, 0.03198259162902832, 0.03172147178649903, 0.031888383865356446, 0.03179417610168457, 0.031507455825805664, 0.03172761535644531, 0.031579135894775394, 0.031692800521850584, 0.0315156478881836, 0.031440895080566404, 0.03173990440368652, 0.03155356788635254, 0.031624160766601565, 0.03153510475158691, 0.033051647186279294, 0.03177369689941406, 0.031783935546875, 0.03135897636413574, 0.03179007911682129, 0.031714303970336914, 0.031848447799682614, 0.031476736068725586, 0.03139993667602539, 0.03133440017700195, 0.03155763244628906, 0.03146956825256347, 0.03141024017333984, 0.031344575881958006, 0.03177574348449707, 0.03194675254821777, 0.03219046401977539, 0.03170918464660644, 0.031678464889526366, 0.03172249603271484, 0.03172147178649903, 0.031513599395751955, 0.03161497688293457, 0.03177779197692871, 0.03177574348449707, 0.03137228775024414, 0.03147369575500488, 0.031712223052978515, 0.03170918464660644, 0.03205222320556641, 0.03178803253173828, 0.03221299362182617, 0.03220889663696289, 0.03131391906738281, 0.03146956825256347, 0.03166720008850098, 0.031661056518554685, 0.03207376098632812, 0.03320111846923828, 0.032717823028564456, 0.032519168853759765, 0.06434508514404297, 0.031736831665039066, 0.031752191543579104, 0.031851520538330076, 0.03142963218688965, 0.03157196807861328, 0.032121856689453124, 0.03311513519287109, 0.03222534561157227, 0.03178079986572266, 0.03179827117919922, 0.03183206367492676, 0.03181465530395508, 0.03183206367492676, 0.03177267265319824, 0.03184025573730469, 0.031542272567749025, 0.03150540733337402, 0.03178598403930664, 0.03185663986206055, 0.03176038360595703, 0.03312639999389649, 0.03152179145812988, 0.032102401733398435, 0.03188019180297851, 0.03193036842346191, 0.031324159622192385, 0.031851520538330076, 0.03158016014099121, 0.031389696121215824, 0.03189555168151856, 0.03170099258422852, 0.03173785591125488, 0.031285247802734374, 0.03177068710327149, 0.031356864929199216, 0.03158527946472168, 0.031526912689208986, 0.03160985565185547, 0.03185868835449219, 0.031848447799682614, 0.031437824249267575, 0.0317388801574707, 0.03174297523498535, 0.0319866886138916, 0.03175014305114746, 0.031922176361083986, 0.031817728042602536, 0.03180544090270996, 0.03139072036743164, 0.03181260871887207, 0.03197952079772949, 0.03224166488647461, 0.03209011077880859, 0.03164057540893555, 0.03165184020996094, 0.031438848495483396, 0.03207372665405273, 0.03194675254821777, 0.03155148887634277, 0.03187302398681641, 0.03263488006591797, 0.03184025573730469, 0.06429801940917969, 0.03187299156188965, 0.03203276824951172, 0.032094207763671875, 0.03197849655151367, 0.031958015441894534, 0.031926271438598636, 0.031834112167358396, 0.03189452743530274, 0.03218841552734375, 0.032353279113769534, 0.0318791675567627, 0.03179315185546875, 0.031887359619140625, 0.03185868835449219, 0.03156480026245117, 0.03182796859741211, 0.0322426872253418, 0.032043006896972655, 0.03181670379638672, 0.03299020767211914, 0.032031742095947266, 0.03186892890930176, 0.032304126739501955, 0.03237171173095703, 0.0319866886138916, 0.031454208374023435, 0.031525888442993165, 0.03151155281066895, 0.031492095947265625, 0.03167334365844727, 0.031730688095092774, 0.03897753524780274, 0.032933887481689454, 0.031889408111572266, 0.03211980819702148, 0.031719423294067385, 0.031352832794189454, 0.03140505599975586, 0.031687711715698244, 0.031402975082397463, 0.03169177627563476, 0.031749120712280275, 0.03167948722839355, 0.03165695953369141, 0.0316753921508789, 0.03172659111022949, 0.03139072036743164, 0.03140812873840332, 0.03175526428222656, 0.03156684875488281, 0.03263590240478516, 0.03202252960205078, 0.03183923149108887, 0.0319682559967041, 0.031649791717529296, 0.031784959793090824, 0.03172352027893066, 0.03167129516601563, 0.0314337272644043, 0.03163443183898926, 0.03186483192443847, 0.03146137619018555, 0.06431948852539063, 0.03137843132019043, 0.03158016014099121, 0.031425535202026365, 0.031645696640014646, 0.031730688095092774, 0.03186486434936524, 0.031797216415405276, 0.0318156795501709, 0.03175628852844238, 0.031699968338012696, 0.031410175323486327, 0.03138355255126953, 0.03278540802001953, 0.03300454330444336, 0.032178176879882815, 0.0318156795501709, 0.031855615615844726, 0.031870975494384765, 0.03283967971801758, 0.0314204158782959, 0.0314081916809082, 0.031652799606323244, 0.03211775970458984, 0.031666175842285156, 0.03154022407531738, 0.031373311996459964, 0.03173785591125488, 0.031546367645263675, 0.0316753921508789, 0.0315043830871582, 0.031529983520507815, 0.031764480590820314, 0.03139481544494629, 0.03133440017700195, 0.03138559913635254, 0.03137843132019043, 0.031735807418823245, 0.031543296813964845, 0.03143065643310547, 0.03159859275817871, 0.031936511993408204, 0.031735807418823245, 0.03193548774719238, 0.031458303451538085, 0.031456256866455076, 0.031716352462768556, 0.0317388801574707, 0.032037952423095706, 0.031414207458496095, 0.0326901741027832, 0.03216793441772461, 0.031585344314575194, 0.03156883239746094, 0.03191398429870605, 0.031631359100341795, 0.031903743743896484, 0.03156076812744141, 0.03184736061096191, 0.03181363105773926, 0.03224371337890625, 0.03225702285766602, 0.031955968856811526, 0.06444134521484375, 0.03154431915283203, 0.031665151596069335, 0.03288780975341797, 0.031886335372924804, 0.03204403305053711, 0.03172659111022949, 0.031509504318237305, 0.03173785591125488, 0.032, 0.032320510864257815, 0.03182387161254883, 0.03230003356933594, 0.03187404823303223, 0.032198688507080075, 0.03247203063964844, 0.03235532760620117, 0.032702465057373044, 0.03311718368530273, 0.03155148887634277, 0.03170201683044434, 0.03225094223022461, 0.03220576095581055, 0.03237580871582031, 0.032230400085449216, 0.03218431854248047, 0.03240447998046875, 0.032449535369873043, 0.03225600051879883, 0.03243622589111328, 0.0319498233795166, 0.03165081596374512, 0.031528959274291994, 0.03161702346801758, 0.0321003532409668, 0.03190784072875977, 0.03203788757324219, 0.032328704833984374, 0.032467967987060545, 0.03184435272216797, 0.031665151596069335, 0.03177369689941406, 0.03174604797363281, 0.03233894348144531, 0.031704063415527346, 0.03172454452514648, 0.032039936065673826, 0.03196211242675781, 0.031926271438598636, 0.0328611831665039, 0.03221200180053711, 0.03217712020874024, 0.03181056022644043, 0.03184332847595215, 0.032064510345458985, 0.03207372665405273, 0.03184339141845703, 0.03191596794128418, 0.031817760467529294, 0.031796192169189455, 0.03153408050537109, 0.032161792755126956, 0.03232460784912109]",tokens/s,30.92327406447812,,,,, -4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-6B,,cuda,0,42,,,,,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,main,False,False,MB,1608.957952,5448.925184,0.0,4802.47808,4489.12128,s,10,5.066485656738281,0.5066485656738281,0.0011948895817140703,0.5068220977783202,0.5079840667724609,0.5080279983520508,0.5080631436157227,"[0.5075723266601563, 0.5061219482421875, 0.504515380859375, 0.5053372497558594, 0.5061441345214843, 0.5055963745117188, 0.5075000610351562, 0.5076519470214844, 0.5080719299316406, 0.5079743041992187]",tokens/s,505.2812093912223,kWh,5.964234835571714e-06,3.2679267622643234e-06,2.7223952334698433e-05,3.6456113932534466e-05,tokens/kWh,7022141.758547072,MB,1610.637312,5448.925184,0.0,4802.47808,4557.793792,s,10,299.1010234375,29.910102343749998,0.01023537902845777,29.913578125,29.921135546875,29.9219037109375,29.9225182421875,"[29.899591796875, 29.901322265625, 29.913013671875, 29.890712890625, 29.920396484375, 29.922671875, 29.92096484375, 29.9152421875, 29.90296484375, 29.914142578125]",tokens/s,2.106311749654192,kWh,0.00035296389131082433,0.00019345485010443554,0.0015722178549955057,0.0021186365964107655,tokens/kWh,29736.104864198915,,s,629,303.18425177001956,0.4820099392210167,0.06023141623810366,0.4746004333496094,0.4757639221191406,0.4761423828125,0.98077423828125,"[0.474365966796875, 0.47431475830078124, 0.4748533630371094, 0.47396148681640626, 0.47435162353515625, 0.47414068603515624, 0.47463320922851565, 0.47518923950195313, 0.47532952880859375, 0.4745502624511719, 0.4740556945800781, 0.47391131591796876, 0.47444989013671873, 0.47430245971679685, 0.4739420166015625, 0.4737402954101563, 0.473987060546875, 0.47486770629882813, 0.4739993591308594, 0.475114501953125, 0.47402597045898437, 0.4752353210449219, 0.47403109741210936, 0.4739563598632813, 0.4750960693359375, 0.4755916748046875, 0.4742686767578125, 0.47484210205078126, 0.4754851989746094, 0.4751022033691406, 0.4755619812011719, 0.4745973815917969, 0.47523736572265624, 0.474829833984375, 0.474850341796875, 0.47421026611328126, 0.4759234619140625, 0.47534796142578123, 0.4750888977050781, 0.4741591186523437, 0.4745441284179687, 0.473744384765625, 0.4741099548339844, 0.473807861328125, 0.47429119873046877, 0.4740597839355469, 0.4742881164550781, 0.47400347900390627, 0.4744253234863281, 0.47619073486328123, 0.4757626953125, 0.4742860717773438, 0.47489535522460935, 0.47489950561523436, 0.47460345458984377, 0.4742686767578125, 0.4739921875, 0.4739420166015625, 0.47433624267578123, 0.4748810119628906, 0.4744110107421875, 0.47424615478515625, 0.980806640625, 0.4736993408203125, 0.47429733276367186, 0.4741478271484375, 0.4741099548339844, 0.47420416259765624, 0.4742000732421875, 0.47553228759765626, 0.4755077209472656, 0.4747817077636719, 0.47485235595703124, 0.4747202453613281, 0.4752384033203125, 0.4742318115234375, 0.4740137023925781, 0.4751011962890625, 0.47453387451171875, 0.4759981994628906, 0.4747898864746094, 0.4742266845703125, 0.47406185913085935, 0.4743239440917969, 0.4744366149902344, 0.47420416259765624, 0.47499981689453125, 0.47433114624023437, 0.47446322631835935, 0.47578317260742187, 0.47489739990234375, 0.4754503784179688, 0.4746977233886719, 0.47437005615234373, 0.4743680114746094, 0.47590194702148436, 0.47595211791992187, 0.47569818115234375, 0.474777587890625, 0.474113037109375, 0.4756346740722656, 0.4741652526855469, 0.47452056884765625, 0.4742778930664063, 0.474176513671875, 0.47388363647460935, 0.473754638671875, 0.47433624267578123, 0.474029052734375, 0.47412841796875, 0.4740433654785156, 0.47558758544921875, 0.4756910400390625, 0.4747693786621094, 0.47467111206054685, 0.4742625427246094, 0.4742799377441406, 0.473807861328125, 0.474387451171875, 0.47398092651367185, 0.4740495300292969, 0.47483187866210935, 0.4742236022949219, 0.4746342468261719, 0.4739717102050781, 0.98069091796875, 0.4739993591308594, 0.4743341979980469, 0.4740321350097656, 0.4741048278808594, 0.47378021240234375, 0.4756019287109375, 0.4752998352050781, 0.47476327514648436, 0.47476531982421877, 0.47458303833007814, 0.474603515625, 0.4745164794921875, 0.47447552490234374, 0.4745482177734375, 0.47400344848632814, 0.4754913330078125, 0.47445709228515626, 0.4743076171875, 0.4742041320800781, 0.47417855834960937, 0.4741171264648438, 0.47425332641601564, 0.47636785888671873, 0.4741949462890625, 0.4747110290527344, 0.4743659973144531, 0.47549435424804687, 0.47543499755859375, 0.4749854736328125, 0.4751646728515625, 0.4754985046386719, 0.4756643981933594, 0.475114501953125, 0.4750315551757813, 0.47576882934570314, 0.47484927368164065, 0.47597467041015623, 0.4747806701660156, 0.47479296875, 0.47538177490234373, 0.4749906005859375, 0.47488409423828126, 0.4746322021484375, 0.4749168701171875, 0.47477862548828126, 0.4759582824707031, 0.47688909912109373, 0.47617330932617186, 0.47571865844726563, 0.4744898681640625, 0.4741591186523437, 0.47446939086914064, 0.4745513000488281, 0.4750899047851562, 0.4740403137207031, 0.474461181640625, 0.4741929016113281, 0.47546676635742186, 0.47404339599609374, 0.474176513671875, 0.47427685546875, 0.47412841796875, 0.979821533203125, 0.4742574157714844, 0.4741734313964844, 0.47405874633789064, 0.47411508178710937, 0.47479910278320314, 0.476015625, 0.47458816528320313, 0.4742225952148437, 0.47418679809570313, 0.4743782043457031, 0.47417138671875, 0.4748810119628906, 0.47444174194335936, 0.4738058166503906, 0.4744478759765625, 0.47413861083984377, 0.47446221923828125, 0.47416116333007813, 0.47388262939453124, 0.47408843994140626, 0.47393484497070315, 0.4740771789550781, 0.47400857543945313, 0.47406695556640627, 0.4738518981933594, 0.4751022033691406, 0.4763402099609375, 0.47449191284179687, 0.4742512512207031, 0.4740157470703125, 0.4743352355957031, 0.47394509887695313, 0.473849853515625, 0.47430963134765625, 0.4738713684082031, 0.47423590087890627, 0.4743987121582031, 0.47423590087890627, 0.4738887634277344, 0.4744407043457031, 0.4742297668457031, 0.4740741271972656, 0.47423590087890627, 0.47649996948242185, 0.4742901611328125, 0.4741949462890625, 0.47464346313476563, 0.476073974609375, 0.47489022827148436, 0.4758292541503906, 0.47497113037109373, 0.47431890869140625, 0.4741651916503906, 0.4740843505859375, 0.47435162353515625, 0.47437310791015624, 0.4743291015625, 0.47532852172851564, 0.474313720703125, 0.474640380859375, 0.47469473266601564, 0.4747908630371094, 0.9814691772460937, 0.4747683715820312, 0.4747376708984375, 0.4743690185546875, 0.4749127807617187, 0.4760391540527344, 0.4749434814453125, 0.4753387451171875, 0.4744488830566406, 0.4743935852050781, 0.4743935852050781, 0.47374334716796873, 0.47430349731445315, 0.47411404418945313, 0.47499264526367185, 0.47437823486328123, 0.4740597839355469, 0.4755292053222656, 0.474429443359375, 0.4746004333496094, 0.47424716186523436, 0.4757176208496094, 0.474365966796875, 0.4744038391113281, 0.47409765625, 0.47461581420898435, 0.47587124633789063, 0.47467724609375, 0.47449700927734373, 0.4752076721191406, 0.47526400756835935, 0.4758815002441406, 0.4752086791992188, 0.4745318298339844, 0.4740362548828125, 0.47401980590820314, 0.474777587890625, 0.47447760009765627, 0.47450723266601563, 0.4745646057128906, 0.4748257141113281, 0.4746987609863281, 0.4746670227050781, 0.4746844177246094, 0.4750899047851562, 0.474967041015625, 0.4748011474609375, 0.47588555908203123, 0.47625112915039064, 0.4750325622558594, 0.4751790161132812, 0.47611801147460936, 0.4751523742675781, 0.4750223388671875, 0.4746065979003906, 0.4746875, 0.47461068725585936, 0.47841998291015625, 0.4753489990234375, 0.47590911865234375, 0.4750417785644531, 0.474998779296875, 0.47562240600585937, 0.9817487182617187, 0.47520870971679685, 0.4753449096679688, 0.47469158935546873, 0.4756121520996094, 0.47580160522460935, 0.47535000610351563, 0.47563983154296874, 0.47536431884765623, 0.4756459655761719, 0.4758149108886719, 0.47476327514648436, 0.47486669921875, 0.4744960021972656, 0.4750315551757813, 0.4745902099609375, 0.4752404479980469, 0.4754586181640625, 0.47481234741210937, 0.47426150512695314, 0.4745635986328125, 0.4747796630859375, 0.4742758483886719, 0.4749291381835937, 0.4753070068359375, 0.4757319641113281, 0.47559988403320314, 0.4744366149902344, 0.4746649475097656, 0.47449087524414063, 0.47499365234375, 0.47450009155273437, 0.47397579956054686, 0.47451034545898435, 0.47416116333007813, 0.4750878601074219, 0.47417752075195313, 0.47466085815429687, 0.4771686401367187, 0.4746567687988281, 0.47440997314453126, 0.4744816589355469, 0.4743618469238281, 0.47448269653320313, 0.47479705810546874, 0.47491070556640624, 0.47615179443359373, 0.4757442626953125, 0.4753827819824219, 0.4749906005859375, 0.4746219482421875, 0.4746925964355469, 0.4743485412597656, 0.4746967163085937, 0.47494552612304686, 0.4743270263671875, 0.47532647705078124, 0.47462603759765626, 0.4751585388183594, 0.4746270751953125, 0.47493734741210936, 0.47487387084960936, 0.4749823913574219, 0.9828372192382813, 0.4750807189941406, 0.4744366149902344, 0.4755558471679687, 0.47547698974609376, 0.47533465576171874, 0.4759132080078125, 0.4759879760742188, 0.47586407470703124, 0.4761640930175781, 0.4759255065917969, 0.4748451843261719, 0.474919921875, 0.476189697265625, 0.4745257263183594, 0.4750560913085938, 0.475325439453125, 0.474986572265625, 0.4742686157226563, 0.4745994262695313, 0.4747817077636719, 0.4749609069824219, 0.4750120849609375, 0.47644467163085935, 0.47719833374023435, 0.47524453735351563, 0.4750530700683594, 0.4744356384277344, 0.47469769287109376, 0.4745891418457031, 0.47447756958007814, 0.4750274658203125, 0.47433831787109376, 0.47436083984375, 0.4751431579589844, 0.4747120666503906, 0.474977294921875, 0.47543295288085935, 0.4742215576171875, 0.47395327758789063, 0.474292236328125, 0.473849853515625, 0.4742840270996094, 0.4745062255859375, 0.47460455322265627, 0.475863037109375, 0.4749680786132813, 0.4751790161132812, 0.4745482177734375, 0.47489227294921876, 0.4745994873046875, 0.47445599365234375, 0.47442022705078124, 0.47415603637695314, 0.47442739868164063, 0.474777587890625, 0.47418060302734377, 0.4745369567871094, 0.47485952758789063, 0.4748001708984375, 0.474187744140625, 0.47446221923828125, 0.47530703735351565, 0.9833133544921875, 0.4750899047851562, 0.4764405822753906, 0.47526705932617186, 0.4746055603027344, 0.47418368530273436, 0.4742369384765625, 0.4743475341796875, 0.474524658203125, 0.474323974609375, 0.4742778930664063, 0.47433010864257813, 0.4754462585449219, 0.47468850708007815, 0.474829833984375, 0.4747304992675781, 0.47638424682617186, 0.4761231994628906, 0.47679379272460937, 0.4750878601074219, 0.4745164794921875, 0.47426150512695314, 0.4747745361328125, 0.47573504638671876, 0.4750520324707031, 0.474893310546875, 0.4743792724609375, 0.4743096923828125, 0.474462158203125, 0.47433831787109376, 0.4749609069824219, 0.4743250427246094, 0.4742419738769531, 0.47488204956054686, 0.47455438232421876, 0.4743075866699219, 0.4756408386230469, 0.47438339233398436, 0.4743741760253906, 0.47458706665039063, 0.47492404174804687, 0.47545547485351564, 0.47435980224609375, 0.47457485961914064, 0.4752353210449219, 0.47622964477539065, 0.47467416381835936, 0.47408126831054687, 0.47500799560546875, 0.47463320922851565, 0.4743915405273437, 0.4742758483886719, 0.47516363525390626, 0.47539404296875, 0.4750796813964844, 0.4738518981933594, 0.4742266845703125, 0.477939697265625, 0.474281982421875, 0.47408126831054687, 0.4742758483886719, 0.4744366149902344, 0.4743372802734375, 0.9826785278320312, 0.4751769714355469, 0.47564697265625, 0.47452569580078124, 0.4744765319824219, 0.47413861083984377, 0.47415090942382815, 0.4741754760742187, 0.4749885559082031, 0.47412017822265623, 0.4737423400878906, 0.4744488830566406, 0.4739051513671875, 0.4738478088378906, 0.4745400390625, 0.47486770629882813, 0.47466085815429687, 0.47589376831054686, 0.47535205078125, 0.4755548095703125, 0.4743424072265625, 0.47426663208007813, 0.47488409423828126, 0.47611911010742186, 0.47500384521484373, 0.4744591369628906, 0.47539813232421874, 0.47478475952148436, 0.47490567016601565, 0.47446112060546874, 0.475104248046875, 0.4745164794921875, 0.47489535522460935, 0.47453900146484373, 0.4745594787597656, 0.47434445190429686, 0.47417752075195313, 0.47448269653320313, 0.4744724731445312, 0.4748072509765625, 0.47473458862304685, 0.4740843505859375, 0.4747591552734375, 0.4748912658691406, 0.4760514526367188, 0.47554150390625, 0.4754565124511719, 0.47558041381835936, 0.4751912841796875, 0.47503564453125, 0.47414990234375, 0.4742850646972656, 0.47414886474609375, 0.47481036376953123, 0.47404852294921873, 0.4742584228515625, 0.4739358825683594, 0.4744171447753906, 0.4741263427734375, 0.47414169311523435, 0.4739491882324219, 0.47383755493164065, 0.4742522888183594, 0.983693359375, 0.47566241455078123, 0.47446722412109377, 0.4746977233886719, 0.4751247253417969, 0.47474688720703123, 0.47455032348632814, 0.4745143737792969, 0.474745849609375, 0.4746485900878906, 0.4754575500488281, 0.47418881225585935, 0.474029052734375, 0.47392666625976565, 0.4740771789550781, 0.4740321350097656, 0.4739317626953125, 0.4744591369628906, 0.47454925537109377, 0.47516058349609375, 0.47410687255859374, 0.4743935852050781, 0.4761661376953125, 0.4749885559082031, 0.47419699096679685, 0.47460250854492186, 0.47660543823242185, 0.4745430908203125, 0.4743280639648437, 0.47440179443359376, 0.4743813171386719, 0.47454721069335937, 0.4742799377441406, 0.4744407043457031, 0.4743854064941406, 0.47423898315429686, 0.47517593383789064, 0.4753879089355469, 0.47501516723632814, 0.47492202758789065, 0.4744867858886719, 0.47409762573242187, 0.47553741455078125, 0.4761282653808594, 0.47627365112304687, 0.47592141723632814, 0.474745849609375, 0.47454931640625, 0.47427374267578126, 0.47429937744140627, 0.47435775756835935, 0.47429937744140627, 0.475188232421875, 0.4755568542480469, 0.4755281982421875, 0.4754483337402344, 0.4758005676269531, 0.47536639404296877, 0.4753592224121094, 0.4763494567871094, 0.47446011352539064, 0.4744376220703125, 0.4739686279296875]",tokens/s,2.074646015839662,,,,, -4bit-awq-gemm-flash_attention_2,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-2.7B,,cuda,0,42,,,,,True,True,,float16,True,False,,flash_attention_2,,False,,False,forward,awq,4,gemm,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.0,217063f5c507ed7cc255df7e1f64c4333a0b4dfe,4.40.2,,0.30.1,,,,1.19.2,,,,0.10.0,,,main,False,False,MB,2234.441728,2932.342784,0.0,2285.89568,2082.575872,s,10,2.433214202880859,0.24332142028808593,0.0008617435568583196,0.24325077056884764,0.24459763488769531,0.24462198333740234,0.24464146209716797,"[0.24380015563964844, 0.24204917907714843, 0.24258213806152343, 0.24336614990234376, 0.24313539123535155, 0.24226588439941407, 0.24288755798339845, 0.24388919067382814, 0.24464633178710937, 0.24459222412109374]",tokens/s,1052.1063032465574,kWh,2.856464352872637e-06,1.5652168180297718e-06,1.2995936322662033e-05,1.7417617493564442e-05,tokens/kWh,14697762.199370166,MB,2234.441728,2959.60576,0.0,2313.158656,2180.684288,s,10,139.734884765625,13.9734884765625,0.01146198033243705,13.9710908203125,13.98807626953125,13.992370166015625,13.995805283203126,"[13.9871220703125, 13.9770771484375, 13.9966640625, 13.980609375, 13.95928515625, 13.96146875, 13.9670087890625, 13.9751728515625, 13.965376953125, 13.965099609375]",tokens/s,4.508537728833344,kWh,0.00016483188493384256,9.03412226583766e-05,0.0007461166431891384,0.0010012897507813577,tokens/kWh,62918.85036358144,,s,629,141.7022607421874,0.22528181358058424,0.02898003334479797,0.22168063354492187,0.2224162872314453,0.22265180053710937,0.4647246643066406,"[0.22235647583007812, 0.22139903259277344, 0.22141644287109374, 0.22147891235351563, 0.2213693389892578, 0.22170726013183595, 0.22163967895507813, 0.2214256591796875, 0.22199398803710937, 0.22159564208984375, 0.22243431091308594, 0.22190284729003906, 0.22202674865722657, 0.2210723876953125, 0.22141133117675782, 0.22209945678710938, 0.22192127990722657, 0.2224189453125, 0.22194073486328125, 0.22234317016601562, 0.22204518127441406, 0.22228582763671875, 0.2221793212890625, 0.22186904907226562, 0.222603271484375, 0.22207487487792968, 0.22233599853515626, 0.22205238342285155, 0.22235133361816406, 0.2235494384765625, 0.22263296508789063, 0.22236671447753906, 0.22127410888671875, 0.2215219268798828, 0.2213519287109375, 0.22189056396484375, 0.22161100769042968, 0.22196940612792967, 0.22203187561035156, 0.22196531677246092, 0.22140007019042968, 0.2213396453857422, 0.2219325408935547, 0.2218577880859375, 0.22200933837890624, 0.22186495971679687, 0.22244557189941405, 0.22216499328613282, 0.22224179077148437, 0.221955078125, 0.2217943115234375, 0.2224179229736328, 0.22189776611328124, 0.2225540771484375, 0.22166732788085938, 0.22277632141113282, 0.22537420654296875, 0.22221209716796875, 0.22146047973632813, 0.22176870727539064, 0.22216294860839844, 0.22160383605957032, 0.46420379638671877, 0.2221997833251953, 0.22188236999511718, 0.22198272705078126, 0.2217697296142578, 0.22201139831542968, 0.2215034942626953, 0.22119526672363282, 0.22219879150390626, 0.22249369812011718, 0.22189773559570314, 0.22246092224121095, 0.22249574279785156, 0.22148197937011718, 0.22124134826660155, 0.221802490234375, 0.22128536987304687, 0.2213744659423828, 0.22234214782714845, 0.22216806030273437, 0.22141746520996095, 0.22162637329101562, 0.22157516479492187, 0.22105191040039063, 0.2215782470703125, 0.2218741760253906, 0.2216898498535156, 0.22149427795410156, 0.2215536651611328, 0.2222725067138672, 0.22197760009765624, 0.22207795715332032, 0.2216847381591797, 0.22156083679199218, 0.2215741424560547, 0.2220595245361328, 0.22160179138183594, 0.22255001831054688, 0.22156492614746093, 0.22203904724121093, 0.22160179138183594, 0.22140109252929688, 0.22128640747070313, 0.22133660888671874, 0.2213908233642578, 0.22154035949707032, 0.22508236694335937, 0.22183013916015626, 0.22202879333496095, 0.2222407684326172, 0.2216785888671875, 0.22169293212890626, 0.22160691833496093, 0.22218649291992187, 0.2214686737060547, 0.22141439819335937, 0.222487548828125, 0.2225991668701172, 0.22195301818847657, 0.22147276306152344, 0.22199909973144533, 0.22176666259765626, 0.22219879150390626, 0.4654428100585937, 0.2218987579345703, 0.22210765075683594, 0.2219008026123047, 0.22165504455566407, 0.22182707214355468, 0.22211891174316406, 0.22192640686035156, 0.221770751953125, 0.22130482482910158, 0.22165913391113282, 0.22189263916015625, 0.2217932434082031, 0.22167552185058595, 0.22172467041015625, 0.22150143432617186, 0.22293606567382812, 0.22162535095214844, 0.2228019256591797, 0.22176255798339845, 0.22248243713378907, 0.22172979736328124, 0.22229402160644532, 0.22310092163085937, 0.22198988342285156, 0.22210150146484375, 0.22238616943359374, 0.22228480529785155, 0.22175334167480468, 0.22205337524414062, 0.22236058044433593, 0.2222407684326172, 0.2220369873046875, 0.2219929656982422, 0.2221854705810547, 0.2224127960205078, 0.22233395385742188, 0.22231858825683593, 0.22264627075195312, 0.22218751525878908, 0.22241587829589843, 0.22255718994140625, 0.2218741760253906, 0.22225920104980468, 0.22183935546875, 0.22192332458496095, 0.22207795715332032, 0.223752197265625, 0.22219673156738282, 0.22203289794921874, 0.22212197875976564, 0.2223206329345703, 0.22238616943359374, 0.22253575134277342, 0.22210047912597655, 0.22248442077636718, 0.2219356231689453, 0.2220912628173828, 0.22202572631835937, 0.22224896240234376, 0.22215168762207033, 0.22308761596679688, 0.22262168884277345, 0.46642996215820315, 0.22230323791503906, 0.22220390319824218, 0.22247935485839843, 0.22223155212402343, 0.22242303466796876, 0.22273023986816406, 0.2226411590576172, 0.22237493896484375, 0.22229808044433594, 0.2220185546875, 0.22259507751464844, 0.22249984741210938, 0.22244659423828125, 0.22167654418945312, 0.22273228454589844, 0.2222581787109375, 0.22126182556152343, 0.2230200653076172, 0.22316950988769532, 0.22224179077148437, 0.22140518188476563, 0.22184857177734374, 0.22160281372070312, 0.22178099060058593, 0.22235443115234374, 0.2213928985595703, 0.22145535278320314, 0.22126591491699218, 0.22138368225097657, 0.22140518188476563, 0.22138983154296876, 0.22134886169433593, 0.22174208068847656, 0.22144720458984374, 0.22132118225097655, 0.22146354675292967, 0.2214001007080078, 0.22144508361816406, 0.2214297637939453, 0.22163763427734376, 0.221517822265625, 0.22244044494628906, 0.22324838256835938, 0.22501274108886718, 0.22167861938476563, 0.22164067077636718, 0.2215004119873047, 0.22143283081054688, 0.22134579467773438, 0.22140518188476563, 0.22115122985839844, 0.22140415954589843, 0.2217195587158203, 0.22167449951171875, 0.2215034942626953, 0.22143795776367187, 0.2215854034423828, 0.22149530029296874, 0.22169497680664063, 0.22140931701660158, 0.22172157287597657, 0.22152806091308594, 0.4667914123535156, 0.22145228576660156, 0.2217512969970703, 0.2216468505859375, 0.22147789001464843, 0.221444091796875, 0.22145126342773438, 0.22144717407226563, 0.22157005310058595, 0.22128947448730468, 0.22147071838378907, 0.22156903076171874, 0.22139801025390626, 0.2215116729736328, 0.2211778564453125, 0.22171034240722656, 0.22124748229980468, 0.22265548706054689, 0.22149836730957032, 0.22151065063476563, 0.22136422729492186, 0.22120550537109376, 0.22162431335449218, 0.22184754943847657, 0.2219530487060547, 0.22183625793457032, 0.22151065063476563, 0.22187826538085936, 0.2214799346923828, 0.22132736206054687, 0.22136524963378906, 0.22137344360351563, 0.22129356384277343, 0.22121778869628905, 0.22129254150390626, 0.2211164093017578, 0.22158746337890625, 0.22140415954589843, 0.22128128051757812, 0.22147071838378907, 0.22140313720703125, 0.22121273803710936, 0.22131808471679687, 0.2216048583984375, 0.22186810302734375, 0.22218540954589844, 0.221623291015625, 0.22134477233886718, 0.22152088928222657, 0.22141746520996095, 0.22138368225097657, 0.22134169006347656, 0.22129458618164063, 0.22138983154296876, 0.22134169006347656, 0.2212956085205078, 0.221159423828125, 0.2240184326171875, 0.22193971252441405, 0.2216089630126953, 0.22141952514648439, 0.22154853820800782, 0.2216837158203125, 0.46434201049804685, 0.22140415954589843, 0.22150758361816406, 0.22161817932128905, 0.2220298309326172, 0.22168780517578124, 0.2222530517578125, 0.22144613647460937, 0.2217840576171875, 0.22155264282226564, 0.22125669860839844, 0.22169197082519532, 0.22176358032226562, 0.22181779479980468, 0.22157926940917969, 0.2215782470703125, 0.2214246368408203, 0.22152294921875, 0.22142771911621092, 0.22154547119140625, 0.22132429504394532, 0.22135606384277343, 0.22152394104003906, 0.22135398864746095, 0.22151577758789062, 0.2220185546875, 0.22214349365234376, 0.22169088745117188, 0.2212464599609375, 0.22148403930664062, 0.221370361328125, 0.22125363159179687, 0.22161305236816406, 0.22148197937011718, 0.22166015625, 0.2215977020263672, 0.22134783935546876, 0.2212351989746094, 0.2241535949707031, 0.22163558959960938, 0.2216785888671875, 0.22160691833496093, 0.221549560546875, 0.22167141723632813, 0.22142054748535156, 0.2214297637939453, 0.22139903259277344, 0.2217943115234375, 0.22149119567871095, 0.221876220703125, 0.22163250732421874, 0.221765625, 0.2215004119873047, 0.2217400360107422, 0.22149427795410156, 0.2213939208984375, 0.2211758117675781, 0.22143487548828125, 0.22141644287109374, 0.22124339294433593, 0.22147071838378907, 0.22166630554199218, 0.22171136474609374, 0.46487347412109375, 0.22141133117675782, 0.22127206420898438, 0.221338623046875, 0.2219622344970703, 0.22154342651367187, 0.22202265930175782, 0.22145330810546876, 0.22258995056152345, 0.22142874145507813, 0.22155264282226564, 0.22175334167480468, 0.2215741424560547, 0.22162739562988282, 0.22164378356933595, 0.2218803253173828, 0.22163250732421874, 0.22160076904296874, 0.22158848571777343, 0.2215116729736328, 0.2215679931640625, 0.22120550537109376, 0.22134477233886718, 0.22316543579101564, 0.22167039489746093, 0.22168063354492187, 0.22168678283691406, 0.2215116729736328, 0.2215188751220703, 0.22155363464355468, 0.22205235290527345, 0.22159359741210938, 0.22261351013183595, 0.2219059143066406, 0.22219161987304686, 0.22189773559570314, 0.22165504455566407, 0.22153114318847655, 0.2218076171875, 0.22219570922851561, 0.22192445373535155, 0.22225091552734375, 0.22204415893554688, 0.22189260864257812, 0.22191104125976563, 0.2215352325439453, 0.22143385314941405, 0.22180557250976562, 0.22147584533691406, 0.22139698791503906, 0.22178713989257812, 0.22213325500488282, 0.22149221801757812, 0.22222848510742188, 0.22150860595703126, 0.2214686737060547, 0.22136627197265624, 0.22127001953125, 0.22115020751953124, 0.22117478942871094, 0.22150143432617186, 0.22140211486816405, 0.22120550537109376, 0.46591384887695314, 0.22151065063476563, 0.22132838439941407, 0.22122496032714845, 0.22145433044433593, 0.22153318786621093, 0.2213519287109375, 0.22116044616699218, 0.22144717407226563, 0.22135296630859375, 0.2212833251953125, 0.22198477172851563, 0.22165811157226561, 0.22226739501953124, 0.22169804382324218, 0.22206361389160156, 0.22276095581054686, 0.22171136474609374, 0.22175949096679687, 0.22171034240722656, 0.22183833312988283, 0.2217021484375, 0.22223257446289063, 0.221876220703125, 0.22166323852539063, 0.22185369873046876, 0.22207693481445312, 0.22190386962890624, 0.2217379913330078, 0.22185165405273438, 0.22234214782714845, 0.22208717346191406, 0.22299136352539062, 0.22185061645507811, 0.22382899475097656, 0.22139187622070314, 0.22159461975097655, 0.22185369873046876, 0.2214072265625, 0.22193971252441405, 0.22215577697753905, 0.22185267639160156, 0.22164480590820312, 0.22156594848632813, 0.2215188751220703, 0.22147990417480468, 0.22186189270019532, 0.22172262573242188, 0.22174310302734376, 0.22215577697753905, 0.22219366455078124, 0.2219448699951172, 0.2216007385253906, 0.22184959411621094, 0.22170930480957032, 0.22189164733886718, 0.2221639404296875, 0.22170518493652344, 0.22201344299316406, 0.2217830352783203, 0.22169395446777343, 0.22200831604003907, 0.22163148498535157, 0.46692044067382815, 0.2217400360107422, 0.22165196228027345, 0.22197555541992187, 0.22193458557128906, 0.22203904724121093, 0.22154655456542968, 0.22138873291015626, 0.22145330810546876, 0.22146354675292967, 0.22145024108886718, 0.22175640869140625, 0.22180351257324218, 0.22217727661132813, 0.22153727722167968, 0.221896728515625, 0.2214164123535156, 0.2225797119140625, 0.22176051330566407, 0.22177381896972656, 0.22154649353027345, 0.221233154296875, 0.2214615020751953, 0.22134681701660155, 0.22129356384277343, 0.22127615356445313, 0.22128640747070313, 0.22148095703125, 0.22134169006347656, 0.22124339294433593, 0.22114816284179686, 0.2213191680908203, 0.2215188446044922, 0.22131814575195313, 0.22146969604492187, 0.22117990112304686, 0.22134375, 0.22164889526367187, 0.22181581115722657, 0.222171142578125, 0.22175640869140625, 0.22172979736328124, 0.22134066772460936, 0.22151986694335937, 0.22140313720703125, 0.22141030883789062, 0.22220594787597656, 0.22215887451171876, 0.22227349853515624, 0.22183116149902343, 0.22192640686035156, 0.22173184204101562, 0.2216898498535156, 0.22144717407226563, 0.22147584533691406, 0.22146354675292967, 0.22156903076171874, 0.22176051330566407, 0.22247833251953125, 0.221802490234375, 0.22252543640136718, 0.22219468688964844, 0.22166015625, 0.46707403564453126, 0.22175640869140625, 0.22192536926269532, 0.22187315368652344, 0.22208204650878907, 0.22162535095214844, 0.22142361450195314, 0.22128025817871094, 0.22170008850097656, 0.2215782470703125, 0.22144613647460937, 0.22178201293945313, 0.22167654418945312, 0.22178816223144532, 0.2213816375732422, 0.2216816711425781, 0.22158131408691406, 0.22198066711425782, 0.22218853759765625, 0.221770751953125, 0.22207693481445312, 0.2219069366455078, 0.22215782165527342, 0.2219069366455078, 0.2217830352783203, 0.2216417236328125, 0.2232033233642578, 0.22143283081054688, 0.22127719116210937, 0.22148300170898438, 0.22127104187011717, 0.2214615020751953, 0.22147584533691406, 0.22175538635253905, 0.22177690124511718, 0.2215290832519531, 0.22144717407226563, 0.22135296630859375, 0.22129356384277343, 0.22165196228027345, 0.22160076904296874, 0.22165402221679686, 0.22142771911621092, 0.22134783935546876, 0.22133555603027344, 0.22152088928222657, 0.22161509704589843, 0.2215034942626953, 0.22159257507324218, 0.22131712341308593, 0.22130586242675782, 0.22127410888671875, 0.22141850280761718, 0.22161305236816406, 0.22188954162597657, 0.22163456726074218, 0.2218014678955078, 0.2217902069091797, 0.22162124633789063, 0.22154444885253907, 0.22161305236816406, 0.2218956756591797, 0.22184857177734374]",tokens/s,4.438884720014453,,,,, -4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciCoder-1b,Deci/DeciCoder-1b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-14B,Qwen/Qwen1.5-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): - File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 600, in resolve_trust_remote_code - answer = input( -EOFError: EOF when reading a line - -During handling of the above exception, another exception occurred: - -Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 228, in load_model_with_no_weights - self.create_no_weights_model() - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 210, in create_no_weights_model - meta_model = self.automodel_class.from_config(self.pretrained_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 418, in from_config - trust_remote_code = resolve_trust_remote_code( - File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 613, in resolve_trust_remote_code - raise ValueError( -ValueError: The repository for Deci/DeciCoder-1b contains custom code which must be executed to correctly load the model. You can inspect the repository content at https://hf.co/Deci/DeciCoder-1b. -Please pass the argument `trust_remote_code=True` to allow custom code to be run. + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 -4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciLM-7B,Deci/DeciLM-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen2-beta-14B,Qwen/Qwen2-beta-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): - File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 600, in resolve_trust_remote_code - answer = input( -EOFError: EOF when reading a line - -During handling of the above exception, another exception occurred: - -Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 228, in load_model_with_no_weights - self.create_no_weights_model() - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 210, in create_no_weights_model - meta_model = self.automodel_class.from_config(self.pretrained_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 418, in from_config - trust_remote_code = resolve_trust_remote_code( - File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 613, in resolve_trust_remote_code - raise ValueError( -ValueError: The repository for Deci/DeciLM-7B contains custom code which must be executed to correctly load the model. You can inspect the repository content at https://hf.co/Deci/DeciLM-7B. -Please pass the argument `trust_remote_code=True` to allow custom code to be run. + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 -4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-j-6b,EleutherAI/gpt-j-6b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-30b,facebook/opt-30b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -2592,26 +22287,32 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3544, in from_pretrained - config = cls._autoset_attn_implementation( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1463, in _autoset_attn_implementation - config = cls._check_and_enable_sdpa( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1619, in _check_and_enable_sdpa - raise ValueError( -ValueError: GPTJForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 -4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-1.3B,EleutherAI/gpt-neo-1.3B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-1.3B,EleutherAI/gpt-neo-1.3B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) - File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): - File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -2621,55 +22322,107 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3544, in from_pretrained - config = cls._autoset_attn_implementation( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1463, in _autoset_attn_implementation - config = cls._check_and_enable_sdpa( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1619, in _check_and_enable_sdpa - raise ValueError( -ValueError: GPTNeoForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 -4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neox-20b,EleutherAI/gpt-neox-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,databricks/dbrx-base,databricks/dbrx-base,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/databricks/dbrx-base/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aa51c-07afd52a3e2e6c0365f37f40;45918444-23e9-467a-8d0e-39e5665732fd) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/databricks/dbrx-base/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ - self.load_model_with_no_weights() - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights - self.load_model_from_pretrained() - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained - self.pretrained_model = self.automodel_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained - return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3544, in from_pretrained - config = cls._autoset_attn_implementation( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1463, in _autoset_attn_implementation - config = cls._check_and_enable_sdpa( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1619, in _check_and_enable_sdpa - raise ValueError( -ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like databricks/dbrx-base is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 -4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-34B,01-ai/Yi-34B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,huggyllama/llama-7b,huggyllama/llama-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -2679,13 +22432,13 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained hf_quantizer.postprocess_model(model) File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 436, in post_init_awq_exllama_modules + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules model = exllamav2_post_init( File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init submodule.post_init(scratch_space=model.scratch_spaces[device]) @@ -2693,18 +22446,18 @@ ChildProcessError: Traceback (most recent call last): self.q_handle = exlv2_ext.make_q_matrix( NameError: name 'exlv2_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 -4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-125m,EleutherAI/gpt-neo-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neox-20b,EleutherAI/gpt-neox-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) - File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): - File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -2714,26 +22467,32 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3544, in from_pretrained - config = cls._autoset_attn_implementation( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1463, in _autoset_attn_implementation - config = cls._check_and_enable_sdpa( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1619, in _check_and_enable_sdpa - raise ValueError( -ValueError: GPTNeoForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 -4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-6B,01-ai/Yi-6B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-4B,Qwen/Qwen1.5-4B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -2743,13 +22502,13 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained hf_quantizer.postprocess_model(model) File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 436, in post_init_awq_exllama_modules + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules model = exllamav2_post_init( File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init submodule.post_init(scratch_space=model.scratch_spaces[device]) @@ -2757,18 +22516,18 @@ ChildProcessError: Traceback (most recent call last): self.q_handle = exlv2_ext.make_q_matrix( NameError: name 'exlv2_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 -4bit-awq-exllama-v2-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-2.7B,EleutherAI/gpt-neo-2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-1.8B,Qwen/Qwen1.5-1.8B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -2778,75 +22537,65 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3544, in from_pretrained - config = cls._autoset_attn_implementation( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1463, in _autoset_attn_implementation - config = cls._check_and_enable_sdpa( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1619, in _check_and_enable_sdpa - raise ValueError( -ValueError: GPTNeoForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 -4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciCoder-1b,Deci/DeciCoder-1b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-14B,Qwen/Qwen-14B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 609, in resolve_trust_remote_code + answer = input( +EOFError: EOF when reading a line + +During handling of the above exception, another exception occurred: + +Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 64, in run - report = scenario.run(backend) - File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run - _ = backend.generate(self.inputs, self.config.generate_kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate - return self.pretrained_model.generate(**inputs, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1576, in generate - result = self._greedy_search( - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2494, in _greedy_search - outputs = self( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1211, in forward - outputs = self.model( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1018, in forward - layer_outputs = decoder_layer( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 741, in forward - hidden_states, self_attn_weights, present_key_value = self.self_attn( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) -TypeError: DeciCoderAttention.forward() got an unexpected keyword argument 'cache_position' + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 228, in load_model_with_no_weights + self.create_no_weights_model() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 210, in create_no_weights_model + meta_model = self.automodel_class.from_config(self.pretrained_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 418, in from_config + trust_remote_code = resolve_trust_remote_code( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 622, in resolve_trust_remote_code + raise ValueError( +ValueError: The repository for Qwen/Qwen-14B contains custom code which must be executed to correctly load the model. You can inspect the repository content at https://hf.co/Qwen/Qwen-14B. +Please pass the argument `trust_remote_code=True` to allow custom code to be run. -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciLM-7B,Deci/DeciLM-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-7.5B,facebook/xglm-7.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -2854,28 +22603,34 @@ ChildProcessError: Traceback (most recent call last): self.load_model_from_pretrained() File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained self.pretrained_model = self.automodel_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 558, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3544, in from_pretrained - config = cls._autoset_attn_implementation( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1463, in _autoset_attn_implementation - config = cls._check_and_enable_sdpa( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1619, in _check_and_enable_sdpa - raise ValueError( -ValueError: DeciLMForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-j-6b,EleutherAI/gpt-j-6b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-12b,EleutherAI/pythia-12b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -2885,26 +22640,32 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3544, in from_pretrained - config = cls._autoset_attn_implementation( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1463, in _autoset_attn_implementation - config = cls._check_and_enable_sdpa( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1619, in _check_and_enable_sdpa - raise ValueError( -ValueError: GPTJForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-1.3B,EleutherAI/gpt-neo-1.3B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-7B,Qwen/Qwen1.5-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -2914,26 +22675,32 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3544, in from_pretrained - config = cls._autoset_attn_implementation( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1463, in _autoset_attn_implementation - config = cls._check_and_enable_sdpa( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1619, in _check_and_enable_sdpa - raise ValueError( -ValueError: GPTNeoForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neox-20b,EleutherAI/gpt-neox-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-34B,01-ai/Yi-34B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) - File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): - File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -2943,85 +22710,177 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3544, in from_pretrained - config = cls._autoset_attn_implementation( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1463, in _autoset_attn_implementation - config = cls._check_and_enable_sdpa( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1619, in _check_and_enable_sdpa - raise ValueError( -ValueError: GPTNeoXForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-34B,01-ai/Yi-34B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/gemma-2b,google/gemma-2b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://huggingface.co/google/gemma-2b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 367, in hf_raise_for_status + raise HfHubHTTPError(message, response=response) from e +huggingface_hub.utils._errors.HfHubHTTPError: (Request ID: Root=1-664aa9a3-05d3169600d9db1b46cdaf73;67a4853e-9348-4efc-8225-c1d28cd81aa2) + +403 Forbidden: Authorization error.. +Cannot access content at: https://huggingface.co/google/gemma-2b/resolve/main/config.json. +If you are trying to create or update content,make sure you have a token with the `write` role. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1826, in _raise_on_head_call_error + raise LocalEntryNotFoundError( +huggingface_hub.utils._errors.LocalEntryNotFoundError: An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache. Please check your connection and try again or make sure your Internet connection is on. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 64, in run - report = scenario.run(backend) - File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run - _ = backend.generate(self.inputs, self.config.generate_kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate - return self.pretrained_model.generate(**inputs, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1576, in generate - result = self._greedy_search( - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2494, in _greedy_search - outputs = self( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1211, in forward - outputs = self.model( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1018, in forward - layer_outputs = decoder_layer( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 741, in forward - hidden_states, self_attn_weights, present_key_value = self.self_attn( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 636, in forward - query_states = self.q_proj(hidden_states) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward - assert AWQ_INSTALLED, ( -AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 442, in cached_file + raise EnvironmentError( +OSError: We couldn't connect to 'https://huggingface.co' to load this file, couldn't find it in the cached files and it looks like google/gemma-2b is not the path to a directory containing a file named config.json. +Checkout your internet connection or see how to run the library in offline mode at 'https://huggingface.co/docs/transformers/installation#offline-mode'. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/polyglot-ko-12.8b,EleutherAI/polyglot-ko-12.8b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined + +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-125m,EleutherAI/gpt-neo-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-125m,EleutherAI/gpt-neo-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-13b,facebook/opt-13b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -3031,85 +22890,102 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3544, in from_pretrained - config = cls._autoset_attn_implementation( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1463, in _autoset_attn_implementation - config = cls._check_and_enable_sdpa( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1619, in _check_and_enable_sdpa - raise ValueError( -ValueError: GPTNeoForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-6B,01-ai/Yi-6B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,google/recurrentgemma-7b,google/recurrentgemma-7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 304, in hf_raise_for_status + response.raise_for_status() + File ""/usr/local/lib/python3.10/dist-packages/requests/models.py"", line 1021, in raise_for_status + raise HTTPError(http_error_msg, response=self) +requests.exceptions.HTTPError: 404 Client Error: Not Found for url: https://huggingface.co/google/recurrentgemma-7b/resolve/main/config.json + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 399, in cached_file + resolved_file = hf_hub_download( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1221, in hf_hub_download + return _hf_hub_download_to_cache_dir( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1325, in _hf_hub_download_to_cache_dir + _raise_on_head_call_error(head_call_error, force_download, local_files_only) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1823, in _raise_on_head_call_error + raise head_call_error + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1722, in _get_metadata_or_catch_error + metadata = get_hf_file_metadata(url=url, proxies=proxies, timeout=etag_timeout, headers=headers) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_validators.py"", line 114, in _inner_fn + return fn(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 1645, in get_hf_file_metadata + r = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 372, in _request_wrapper + response = _request_wrapper( + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/file_download.py"", line 396, in _request_wrapper + hf_raise_for_status(response) + File ""/usr/local/lib/python3.10/dist-packages/huggingface_hub/utils/_errors.py"", line 352, in hf_raise_for_status + raise RepositoryNotFoundError(message, response) from e +huggingface_hub.utils._errors.RepositoryNotFoundError: 404 Client Error. (Request ID: Root=1-664aaaa4-4dac6ffc22548609380f1682;3e0c4a0f-09dd-4297-8ab0-3edf4ea2f509) + +Repository Not Found for url: https://huggingface.co/google/recurrentgemma-7b/resolve/main/config.json. +Please make sure you specified the correct `repo_id` and `repo_type`. +If you are trying to access a private or gated repo, make sure you are authenticated. + +The above exception was the direct cause of the following exception: + +Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 64, in run - report = scenario.run(backend) - File ""/workspace/optimum_benchmark/scenarios/inference/scenario.py"", line 117, in run - _ = backend.generate(self.inputs, self.config.generate_kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 369, in generate - return self.pretrained_model.generate(**inputs, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 1576, in generate - result = self._greedy_search( - File ""/usr/local/lib/python3.10/dist-packages/transformers/generation/utils.py"", line 2494, in _greedy_search - outputs = self( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1211, in forward - outputs = self.model( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 1018, in forward - layer_outputs = decoder_layer( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 741, in forward - hidden_states, self_attn_weights, present_key_value = self.self_attn( - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/llama/modeling_llama.py"", line 636, in forward - query_states = self.q_proj(hidden_states) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1532, in _wrapped_call_impl - return self._call_impl(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1541, in _call_impl - return forward_call(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py"", line 115, in decorate_context - return func(*args, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/gemv.py"", line 162, in forward - assert AWQ_INSTALLED, ( -AssertionError: AWQ kernels could not be loaded. Please install them from https://github.com/casper-hansen/AutoAWQ_kernels + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 37, in __init__ + super().__init__(config) + File ""/workspace/optimum_benchmark/backends/base.py"", line 62, in __init__ + self.pretrained_config = get_transformers_pretrained_config(self.config.model, **self.config.model_kwargs) + File ""/workspace/optimum_benchmark/backends/transformers_utils.py"", line 22, in get_transformers_pretrained_config + return AutoConfig.from_pretrained(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/configuration_auto.py"", line 934, in from_pretrained + config_dict, unused_kwargs = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 632, in get_config_dict + config_dict, kwargs = cls._get_config_dict(pretrained_model_name_or_path, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/configuration_utils.py"", line 689, in _get_config_dict + resolved_config_file = cached_file( + File ""/usr/local/lib/python3.10/dist-packages/transformers/utils/hub.py"", line 422, in cached_file + raise EnvironmentError( +OSError: google/recurrentgemma-7b is not a local folder and is not a valid model identifier listed on 'https://huggingface.co/models' +If this is a private repository, make sure to pass a token having permission to this repo either by logging in with `huggingface-cli login` or by passing `token=` -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-gemv-sdpa,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-2.7B,EleutherAI/gpt-neo-2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,sdpa,,False,,False,forward,awq,4,gemv,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-410m,EleutherAI/pythia-410m,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -3119,92 +22995,102 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3544, in from_pretrained - config = cls._autoset_attn_implementation( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1463, in _autoset_attn_implementation - config = cls._check_and_enable_sdpa( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 1619, in _check_and_enable_sdpa - raise ValueError( -ValueError: GPTNeoForCausalLM does not support an attention implementation through torch.nn.functional.scaled_dot_product_attention yet. Please request the support for this architecture: https://github.com/huggingface/transformers/issues/28005. If you believe this error is a bug, please open an issue in Transformers GitHub repository and load your model with the argument `attn_implementation=""eager""` meanwhile. Example: `model = AutoModel.from_pretrained(""openai/whisper-tiny"", attn_implementation=""eager"")` + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, -4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciCoder-1b,Deci/DeciCoder-1b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen1.5-32B,Qwen/Qwen1.5-32B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): - File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 600, in resolve_trust_remote_code - answer = input( -EOFError: EOF when reading a line - -During handling of the above exception, another exception occurred: - -Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 228, in load_model_with_no_weights - self.create_no_weights_model() - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 210, in create_no_weights_model - meta_model = self.automodel_class.from_config(self.pretrained_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 418, in from_config - trust_remote_code = resolve_trust_remote_code( - File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 613, in resolve_trust_remote_code - raise ValueError( -ValueError: The repository for Deci/DeciCoder-1b contains custom code which must be executed to correctly load the model. You can inspect the repository content at https://hf.co/Deci/DeciCoder-1b. -Please pass the argument `trust_remote_code=True` to allow custom code to be run. + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 -4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Deci/DeciLM-7B,Deci/DeciLM-7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-4.5B,facebook/xglm-4.5B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): - File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 600, in resolve_trust_remote_code - answer = input( -EOFError: EOF when reading a line - -During handling of the above exception, another exception occurred: - -Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 228, in load_model_with_no_weights - self.create_no_weights_model() - File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 210, in create_no_weights_model - meta_model = self.automodel_class.from_config(self.pretrained_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 418, in from_config - trust_remote_code = resolve_trust_remote_code( - File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 613, in resolve_trust_remote_code - raise ValueError( -ValueError: The repository for Deci/DeciLM-7B contains custom code which must be executed to correctly load the model. You can inspect the repository content at https://hf.co/Deci/DeciLM-7B. -Please pass the argument `trust_remote_code=True` to allow custom code to be run. + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained + hf_quantizer.postprocess_model(model) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model + return self._process_model_after_weight_loading(model, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading + model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules + model = exllamav2_post_init( + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init + submodule.post_init(scratch_space=model.scratch_spaces[device]) + File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 81, in post_init + self.q_handle = exlv2_ext.make_q_matrix( +NameError: name 'exlv2_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 -4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-j-6b,EleutherAI/gpt-j-6b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,TencentARC/Mistral_Pro_8B_v0.1,TencentARC/Mistral_Pro_8B_v0.1,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -3214,13 +23100,13 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained hf_quantizer.postprocess_model(model) File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 436, in post_init_awq_exllama_modules + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules model = exllamav2_post_init( File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init submodule.post_init(scratch_space=model.scratch_spaces[device]) @@ -3228,18 +23114,18 @@ ChildProcessError: Traceback (most recent call last): self.q_handle = exlv2_ext.make_q_matrix( NameError: name 'exlv2_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 -4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-1.3B,EleutherAI/gpt-neo-1.3B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/xglm-564M,facebook/xglm-564M,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -3249,13 +23135,13 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained hf_quantizer.postprocess_model(model) File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 436, in post_init_awq_exllama_modules + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules model = exllamav2_post_init( File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init submodule.post_init(scratch_space=model.scratch_spaces[device]) @@ -3263,18 +23149,18 @@ ChildProcessError: Traceback (most recent call last): self.q_handle = exlv2_ext.make_q_matrix( NameError: name 'exlv2_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 -4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neox-20b,EleutherAI/gpt-neox-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-350m,facebook/opt-350m,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -3284,13 +23170,13 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained hf_quantizer.postprocess_model(model) File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 436, in post_init_awq_exllama_modules + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules model = exllamav2_post_init( File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init submodule.post_init(scratch_space=model.scratch_spaces[device]) @@ -3298,18 +23184,18 @@ ChildProcessError: Traceback (most recent call last): self.q_handle = exlv2_ext.make_q_matrix( NameError: name 'exlv2_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 -4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-34B,01-ai/Yi-34B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-6.7b,facebook/opt-6.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -3319,13 +23205,13 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained hf_quantizer.postprocess_model(model) File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 436, in post_init_awq_exllama_modules + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules model = exllamav2_post_init( File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init submodule.post_init(scratch_space=model.scratch_spaces[device]) @@ -3333,18 +23219,51 @@ ChildProcessError: Traceback (most recent call last): self.q_handle = exlv2_ext.make_q_matrix( NameError: name 'exlv2_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 -4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-125m,EleutherAI/gpt-neo-125m,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen-72B,Qwen/Qwen-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 609, in resolve_trust_remote_code + answer = input( +EOFError: EOF when reading a line + +During handling of the above exception, another exception occurred: + +Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 228, in load_model_with_no_weights + self.create_no_weights_model() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 210, in create_no_weights_model + meta_model = self.automodel_class.from_config(self.pretrained_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 418, in from_config + trust_remote_code = resolve_trust_remote_code( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 622, in resolve_trust_remote_code + raise ValueError( +ValueError: The repository for Qwen/Qwen-72B contains custom code which must be executed to correctly load the model. You can inspect the repository content at https://hf.co/Qwen/Qwen-72B. +Please pass the argument `trust_remote_code=True` to allow custom code to be run. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-6B,01-ai/Yi-6B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -3354,13 +23273,13 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained hf_quantizer.postprocess_model(model) File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 436, in post_init_awq_exllama_modules + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules model = exllamav2_post_init( File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init submodule.post_init(scratch_space=model.scratch_spaces[device]) @@ -3368,18 +23287,60 @@ ChildProcessError: Traceback (most recent call last): self.q_handle = exlv2_ext.make_q_matrix( NameError: name 'exlv2_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 -4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,01-ai/Yi-6B,01-ai/Yi-6B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,Qwen/Qwen2-beta-72B,Qwen/Qwen2-beta-72B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 96.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-2.7B,EleutherAI/gpt-neo-2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 47, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 66, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 102, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 60, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -3389,13 +23350,13 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained hf_quantizer.postprocess_model(model) File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 436, in post_init_awq_exllama_modules + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules model = exllamav2_post_init( File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init submodule.post_init(scratch_space=model.scratch_spaces[device]) @@ -3403,18 +23364,93 @@ ChildProcessError: Traceback (most recent call last): self.q_handle = exlv2_ext.make_q_matrix( NameError: name 'exlv2_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 -4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/gpt-neo-2.7B,EleutherAI/gpt-neo-2.7B,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.40.2,,0.30.1,,,,1.19.2,,,,0.11.0,,"Traceback (most recent call last): +",False,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,facebook/opt-66b,facebook/opt-66b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 255, in load_model_with_no_weights + self.load_model_from_pretrained() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 169, in load_model_from_pretrained + self.pretrained_model = self.automodel_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained + return model_class.from_pretrained( + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3820, in from_pretrained + dispatch_model(model, **device_map_kwargs) + File ""/usr/local/lib/python3.10/dist-packages/accelerate/big_modeling.py"", line 488, in dispatch_model + model.to(device) + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 2724, in to + return super().to(*args, **kwargs) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1173, in to + return self._apply(convert) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 779, in _apply + module._apply(fn) + [Previous line repeated 2 more times] + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 853, in _apply + self._buffers[key] = fn(buf) + File ""/usr/local/lib/python3.10/dist-packages/torch/nn/modules/module.py"", line 1159, in convert + return t.to( +torch.cuda.OutOfMemoryError: CUDA out of memory. Tried to allocate 162.00 MiB. GPU + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,internlm/internlm2-20b,internlm/internlm2-20b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): + File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch + benchmark_report = Benchmark.launch(benchmark_config) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch + report = launcher.launch(worker=cls.run, worker_args=[config]) + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch + raise ChildProcessError(response[""traceback""]) +ChildProcessError: Traceback (most recent call last): + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 609, in resolve_trust_remote_code + answer = input( +EOFError: EOF when reading a line + +During handling of the above exception, another exception occurred: + +Traceback (most recent call last): + File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target + report = worker(*worker_args) + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run + backend: Backend = backend_factory(backend_config) + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ + self.load_model_with_no_weights() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 228, in load_model_with_no_weights + self.create_no_weights_model() + File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 210, in create_no_weights_model + meta_model = self.automodel_class.from_config(self.pretrained_config) + File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 418, in from_config + trust_remote_code = resolve_trust_remote_code( + File ""/usr/local/lib/python3.10/dist-packages/transformers/dynamic_module_utils.py"", line 622, in resolve_trust_remote_code + raise ValueError( +ValueError: The repository for internlm/internlm2-20b contains custom code which must be executed to correctly load the model. You can inspect the repository content at https://hf.co/internlm/internlm2-20b. +Please pass the argument `trust_remote_code=True` to allow custom code to be run. + +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +4bit-awq-exllama-v2-eager,pytorch,2.3.0+cu121,optimum_benchmark.backends.pytorch.backend.PyTorchBackend,text-generation,transformers,EleutherAI/pythia-6.7b,EleutherAI/pythia-6.7b,cuda,0,42,,,True,True,True,True,,float16,True,False,,eager,,False,,False,forward,awq,4,exllama,False,,inference,optimum_benchmark.scenarios.inference.scenario.InferenceScenario,10,10,10,1,2,256,,True,True,True,64,64,process,optimum_benchmark.launchers.process.launcher.ProcessLauncher,True,kill,spawn, AMD EPYC 7R32,16,66697.29792,Linux,x86_64,Linux-5.10.215-203.850.amzn2.x86_64-x86_64-with-glibc2.35,x86_64,3.10.12,['NVIDIA A10G'],1,24146608128,0.2.1,,4.41.0,,0.30.1,,,,1.19.2,,,,0.11.1,,"Traceback (most recent call last): File ""/workspace/llm_perf/update_llm_perf_cuda_pytorch.py"", line 148, in benchmark_cuda_pytorch benchmark_report = Benchmark.launch(benchmark_config) - File ""/workspace/optimum_benchmark/base.py"", line 43, in launch + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 43, in launch report = launcher.launch(worker=cls.run, worker_args=[config]) File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 54, in launch raise ChildProcessError(response[""traceback""]) ChildProcessError: Traceback (most recent call last): File ""/workspace/optimum_benchmark/launchers/process/launcher.py"", line 87, in target report = worker(*worker_args) - File ""/workspace/optimum_benchmark/base.py"", line 56, in run + File ""/workspace/optimum_benchmark/benchmark/base.py"", line 56, in run backend: Backend = backend_factory(backend_config) File ""/workspace/optimum_benchmark/backends/pytorch/backend.py"", line 79, in __init__ self.load_model_with_no_weights() @@ -3424,13 +23460,13 @@ ChildProcessError: Traceback (most recent call last): self.pretrained_model = self.automodel_class.from_pretrained( File ""/usr/local/lib/python3.10/dist-packages/transformers/models/auto/auto_factory.py"", line 563, in from_pretrained return model_class.from_pretrained( - File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3738, in from_pretrained + File ""/usr/local/lib/python3.10/dist-packages/transformers/modeling_utils.py"", line 3823, in from_pretrained hf_quantizer.postprocess_model(model) File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/base.py"", line 195, in postprocess_model return self._process_model_after_weight_loading(model, **kwargs) - File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 105, in _process_model_after_weight_loading + File ""/usr/local/lib/python3.10/dist-packages/transformers/quantizers/quantizer_awq.py"", line 107, in _process_model_after_weight_loading model = post_init_awq_exllama_modules(model, self.quantization_config.exllama_config) - File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 436, in post_init_awq_exllama_modules + File ""/usr/local/lib/python3.10/dist-packages/transformers/integrations/awq.py"", line 465, in post_init_awq_exllama_modules model = exllamav2_post_init( File ""/usr/local/lib/python3.10/dist-packages/awq/modules/linear/exllamav2.py"", line 198, in exllamav2_post_init submodule.post_init(scratch_space=model.scratch_spaces[device]) @@ -3438,4 +23474,4 @@ ChildProcessError: Traceback (most recent call last): self.q_handle = exlv2_ext.make_q_matrix( NameError: name 'exlv2_ext' is not defined -",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1 +",,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,2,64,1