Create README.md

README.md

@@ -0,0 +1,286 @@
+---
+license: mit
+tags:
+- deepseek
+- int4
+- vllm
+- llmcompressor
+base_model: deepseek-ai/DeepSeek-R1-Distill-Llama-70B
+library_name: transformers
+---
+
+# DeepSeek-R1-Distill-Llama-70B-quantized.w4a16
+
+## Model Overview
+- **Model Architecture:** LlamaForCausalLM
+  - **Input:** Text
+  - **Output:** Text
+- **Model Optimizations:**
+  - **Weight quantization:** INT4
+- **Release Date:** 2/7/2025
+- **Version:** 1.0
+- **Model Developers:** Neural Magic
+
+Quantized version of [DeepSeek-R1-Distill-Llama-70B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Llama-70B).
+
+
+### Model Optimizations
+
+This model was obtained by quantizing the weights of [DeepSeek-R1-Distill-Llama-70B](https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Llama-70B) to INT4 data type.
+This optimization reduces the number of bits per parameter from 16 to 4, reducing the disk size and GPU memory requirements by approximately 75%.
+
+
+Only the weights of the linear operators within transformers blocks are quantized.
+Weights are quantized using a symmetric per-group scheme, with group size 128.
+The [GPTQ](https://arxiv.org/abs/2210.17323) algorithm is applied for quantization, as implemented in the [llm-compressor](https://github.com/vllm-project/llm-compressor) library.
+
+
+## Use with vLLM
+
+This model can be deployed efficiently using the [vLLM](https://docs.vllm.ai/en/latest/) backend, as shown in the example below.
+
+```python
+from transformers import AutoTokenizer
+from vllm import LLM, SamplingParams
+
+number_gpus = 1
+model_name = "neuralmagic/DeepSeek-R1-Distill-Llama-70B-quantized.w4a16"
+
+tokenizer = AutoTokenizer.from_pretrained(model_name)
+sampling_params = SamplingParams(temperature=0.6, max_tokens=256, stop_token_ids=[tokenizer.eos_token_id])
+llm = LLM(model=model_name, tensor_parallel_size=number_gpus, trust_remote_code=True)
+
+messages_list = [
+    [{"role": "user", "content": "Who are you? Please respond in pirate speak!"}],
+]
+
+prompt_token_ids = [tokenizer.apply_chat_template(messages, add_generation_prompt=True) for messages in messages_list]
+
+outputs = llm.generate(prompt_token_ids=prompt_token_ids, sampling_params=sampling_params)
+
+generated_text = [output.outputs[0].text for output in outputs]
+print(generated_text)
+```
+
+vLLM also supports OpenAI-compatible serving. See the [documentation](https://docs.vllm.ai/en/latest/) for more details.
+
+## Creation
+
+This model was created with [llm-compressor](https://github.com/vllm-project/llm-compressor) by running the code snippet below. 
+
+
+```python
+from transformers import AutoModelForCausalLM, AutoTokenizer
+from llmcompressor.modifiers.quantization import QuantizationModifier
+from llmcompressor.modifiers.smoothquant import SmoothQuantModifier
+from llmcompressor.transformers import oneshot
+from llmcompressor.transformers.compression.helpers import calculate_offload_device_map
+
+# Load model
+model_stub = "deepseek-ai/DeepSeek-R1-Distill-Llama-70B"
+model_name = model_stub.split("/")[-1]
+
+num_samples = 3072
+max_seq_len = 8192
+
+tokenizer = AutoTokenizer.from_pretrained(model_stub)
+
+device_map = calculate_offload_device_map(
+    model_stub,
+    reserve_for_hessians=True,
+    num_gpus=2,
+    torch_dtype="auto",
+)
+
+model = AutoModelForCausalLM.from_pretrained(
+    model_stub,
+    device_map=device_map,
+    torch_dtype="auto",
+)
+
+def preprocess_fn(example):
+  return {"text": tokenizer.apply_chat_template(example["messages"], add_generation_prompt=False, tokenize=False)}
+
+ds = load_dataset("neuralmagic/LLM_compression_calibration", split="train")
+ds = ds.map(preprocess_fn)
+
+# Configure the quantization algorithm and scheme
+recipe = QuantizationModifier(
+    targets="Linear",
+    scheme="W4A16",
+    ignore=["lm_head"],
+    dampening_frac=0.1,
+)
+
+# Apply quantization
+oneshot(
+    model=model,
+    dataset=ds, 
+    recipe=recipe,
+    max_seq_length=max_seq_len,
+    num_calibration_samples=num_samples,
+)
+
+# Save to disk in compressed-tensors format
+save_path = model_name + "-quantized.w4a16
+model.save_pretrained(save_path)
+tokenizer.save_pretrained(save_path)
+print(f"Model and tokenizer saved to: {save_path}")
+```
+
+## Evaluation
+
+The model was evaluated on OpenLLM Leaderboard [V1](https://huggingface.co/spaces/open-llm-leaderboard-old/open_llm_leaderboard) and [V2](https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard#/), using the following commands:
+
+OpenLLM Leaderboard V1:
+```
+lm_eval \
+  --model vllm \
+  --model_args pretrained="neuralmagic/DeepSeek-R1-Distill-Llama-70B-quantized.w4a16",dtype=auto,max_model_len=4096,tensor_parallel_size=1,enable_chunked_prefill=True \
+  --tasks openllm \
+  --write_out \
+  --batch_size auto \
+  --output_path output_dir \
+  --show_config
+```
+
+OpenLLM Leaderboard V2:
+```
+lm_eval \
+  --model vllm \
+  --model_args pretrained="neuralmagic/DeepSeek-R1-Distill-Llama-70B-quantized.w4a16",dtype=auto,max_model_len=4096,tensor_parallel_size=1,enable_chunked_prefill=True \
+  --apply_chat_template \
+  --fewshot_as_multiturn \
+  --tasks leaderboard \
+  --write_out \
+  --batch_size auto \
+  --output_path output_dir \
+  --show_config
+```
+
+### Accuracy
+
+<table>
+  <thead>
+    <tr>
+      <th>Category</th>
+      <th>Metric</th>
+      <th>deepseek-ai/DeepSeek-R1-Distill-Llama-70B</th>
+      <th>neuralmagic/DeepSeek-R1-Distill-Llama-70B-quantized.w4a16</th>
+      <th>Recovery</th>
+    </tr>
+  </thead>
+  <tbody>
+    <tr>
+      <td rowspan="7"><b>OpenLLM V1</b></td>
+      <td>ARC-Challenge (Acc-Norm, 25-shot)</td>
+      <td>63.65</td>
+      <td>63.31</td>
+      <td>99.5%</td>
+    </tr>
+    <tr>
+      <td>GSM8K (Strict-Match, 5-shot)</td>
+      <td>93.03</td>
+      <td>93.03</td>
+      <td>100.0%</td>
+    </tr>
+    <tr>
+      <td>HellaSwag (Acc-Norm, 10-shot)</td>
+      <td>84.85</td>
+      <td>84.43</td>
+      <td>99.5%</td>
+    </tr>
+    <tr>
+      <td>MMLU (Acc, 5-shot)</td>
+      <td>78.04</td>
+      <td>77.15</td>
+      <td>98.9%</td>
+    </tr>
+    <tr>
+      <td>TruthfulQA (MC2, 0-shot)</td>
+      <td>56.67</td>
+      <td>57.79</td>
+      <td>102.0%</td>
+    </tr>
+    <tr>
+      <td>Winogrande (Acc, 5-shot)</td>
+      <td>78.22</td>
+      <td>79.48</td>
+      <td>101.6%</td>
+    </tr>
+    <tr>
+      <td><b>Average Score</b></td>
+      <td><b>75.74</b></td>
+      <td><b>75.86</b></td>
+      <td><b>100.2%</b></td>
+    </tr>
+    <tr>
+      <td rowspan="7"><b>OpenLLM V2</b></td>
+      <td>IFEval (Inst Level Strict Acc, 0-shot)</td>
+      <td>43.15</td>
+      <td>42.08</td>
+      <td>97.5%</td>
+    </tr>
+    <tr>
+      <td>BBH (Acc-Norm, 3-shot)</td>
+      <td>64.32</td>
+      <td>63.91</td>
+      <td>99.4%</td>
+    </tr>
+    <tr>
+      <td>Math-Hard (Exact-Match, 4-shot)</td>
+      <td>35.04</td>
+      <td>37.81</td>
+      <td>107.9%</td>
+    </tr>
+    <tr>
+      <td>GPQA (Acc-Norm, 0-shot)</td>
+      <td>37.15</td>
+      <td>36.64</td>
+      <td>98.6%</td>
+    </tr>
+    <tr>
+      <td>MUSR (Acc-Norm, 0-shot)</td>
+      <td>42.89</td>
+      <td>42.49</td>
+      <td>99.1%</td>
+    </tr>
+    <tr>
+      <td>MMLU-Pro (Acc, 5-shot)</td>
+      <td>47.22</td>
+      <td>45.78</td>
+      <td>96.9%</td>
+    </tr>
+    <tr>
+      <td><b>Average Score</b></td>
+      <td><b>44.96</b></td>
+      <td><b>44.78</b></td>
+      <td><b>99.6%</b></td>
+    </tr>
+    <tr>
+      <td rowspan="4"><b>Coding</b></td>
+      <td>HumanEval (pass@1)</td>
+      <td>81.10</td>
+      <td>80.20</td>
+      <td><b>98.9%</b></td>
+    </tr>
+    <tr>
+      <td>HumanEval (pass@10)</td>
+      <td>87.60</td>
+      <td>89.30</td>
+      <td>101.9%</td>
+    </tr>
+    <tr>
+      <td>HumanEval+ (pass@10)</td>
+      <td>75.20</td>
+      <td>73.00</td>
+      <td>97.1%</td>
+    </tr>
+    <tr>
+      <td>HumanEval+ (pass@10)</td>
+      <td>83.10</td>
+      <td>83.70</td>
+      <td>100.7%</td>
+    </tr>
+  </tbody>
+</table>