Hugging Face's logo

optimum-internal-testing
/
neuron-testing-cache

Model card Files Community
dacorvo HF Staff commited on
Commit
1febc4e
·
verified ·
1 Parent(s): f23cc9c

Synchronizing local compiler cache.

Browse files
Files changed (16) hide show
  1. .gitattributes +4 -0
  2. neuronxcc-2.17.194.0+d312836f/0_REGISTRY/0.2.0.dev7/llama/meta-llama/Llama-3.1-8B-Instruct/d04674b7e4428d46a038.json +77 -0
  3. neuronxcc-2.17.194.0+d312836f/0_REGISTRY/0.2.0.dev7/llama/unsloth/Llama-3.2-1B-Instruct/a01ee1e6bc19cfb5be32.json +78 -0
  4. neuronxcc-2.17.194.0+d312836f/MODULE_a30822a1b2d1fbcee30c+bfe5714b/compile_flags.json +1 -0
  5. neuronxcc-2.17.194.0+d312836f/MODULE_a30822a1b2d1fbcee30c+bfe5714b/model.done +0 -0
  6. neuronxcc-2.17.194.0+d312836f/MODULE_a30822a1b2d1fbcee30c+bfe5714b/model.hlo_module.pb +3 -0
  7. neuronxcc-2.17.194.0+d312836f/MODULE_a30822a1b2d1fbcee30c+bfe5714b/model.neff +3 -0
  8. neuronxcc-2.17.194.0+d312836f/MODULE_a849442b615562b13ba6+431f5505/compile_flags.json +1 -0
  9. neuronxcc-2.17.194.0+d312836f/MODULE_a849442b615562b13ba6+431f5505/model.done +0 -0
  10. neuronxcc-2.17.194.0+d312836f/MODULE_a849442b615562b13ba6+431f5505/model.hlo_module.pb +3 -0
  11. neuronxcc-2.17.194.0+d312836f/MODULE_a849442b615562b13ba6+431f5505/model.neff +3 -0
  12. neuronxcc-2.17.194.0+d312836f/MODULE_ccfcad18cf4ad9b6e4bc+165e9558/compile_flags.json +1 -0
  13. neuronxcc-2.17.194.0+d312836f/MODULE_ccfcad18cf4ad9b6e4bc+165e9558/model.done +0 -0
  14. neuronxcc-2.17.194.0+d312836f/MODULE_ccfcad18cf4ad9b6e4bc+165e9558/model.hlo_module.pb +3 -0
  15. neuronxcc-2.17.194.0+d312836f/MODULE_ccfcad18cf4ad9b6e4bc+165e9558/model.neff +3 -0
  16. neuronxcc-2.17.194.0+d312836f/MODULE_ccfcad18cf4ad9b6e4bc+165e9558/wrapped_neff.hlo +3 -0
.gitattributes CHANGED
@@ -2704,3 +2704,7 @@ neuronxcc-2.17.194.0+d312836f/MODULE_f2426656f5f2ceb6f048+165e9558/wrapped_neff.
2704
  neuronxcc-2.17.194.0+d312836f/MODULE_f2bc87b9df15b7784add+bfe5714b/model.neff filter=lfs diff=lfs merge=lfs -text
2705
  neuronxcc-2.17.194.0+d312836f/MODULE_f781291bf31bf79730b8+431f5505/model.neff filter=lfs diff=lfs merge=lfs -text
2706
  neuronxcc-2.17.194.0+d312836f/MODULE_fdcbd41f965fdff93616+431f5505/model.neff filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
2704
  neuronxcc-2.17.194.0+d312836f/MODULE_f2bc87b9df15b7784add+bfe5714b/model.neff filter=lfs diff=lfs merge=lfs -text
2705
  neuronxcc-2.17.194.0+d312836f/MODULE_f781291bf31bf79730b8+431f5505/model.neff filter=lfs diff=lfs merge=lfs -text
2706
  neuronxcc-2.17.194.0+d312836f/MODULE_fdcbd41f965fdff93616+431f5505/model.neff filter=lfs diff=lfs merge=lfs -text
2707
+ neuronxcc-2.17.194.0+d312836f/MODULE_a30822a1b2d1fbcee30c+bfe5714b/model.neff filter=lfs diff=lfs merge=lfs -text
2708
+ neuronxcc-2.17.194.0+d312836f/MODULE_a849442b615562b13ba6+431f5505/model.neff filter=lfs diff=lfs merge=lfs -text
2709
+ neuronxcc-2.17.194.0+d312836f/MODULE_ccfcad18cf4ad9b6e4bc+165e9558/model.neff filter=lfs diff=lfs merge=lfs -text
2710
+ neuronxcc-2.17.194.0+d312836f/MODULE_ccfcad18cf4ad9b6e4bc+165e9558/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.17.194.0+d312836f/0_REGISTRY/0.2.0.dev7/llama/meta-llama/Llama-3.1-8B-Instruct/d04674b7e4428d46a038.json ADDED
@@ -0,0 +1,77 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "meta-llama/Llama-3.1-8B-Instruct",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 4096,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 14336,
15
+ "max_position_embeddings": 131072,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "neuron": {
19
+ "_serialized_key": "NxDNeuronConfig",
20
+ "async_mode": false,
21
+ "attn_kernel_enabled": false,
22
+ "batch_size": 32,
23
+ "capacity_factor": null,
24
+ "cc_pipeline_tiling_factor": 2,
25
+ "checkpoint_id": "meta-llama/Llama-3.1-8B-Instruct",
26
+ "checkpoint_revision": "0e9e39f249a16976918f6564b8830bc894c89659",
27
+ "continuous_batching": true,
28
+ "enable_bucketing": false,
29
+ "ep_degree": 1,
30
+ "flash_decoding_enabled": false,
31
+ "fused_qkv": true,
32
+ "glu_mlp": true,
33
+ "is_chunked_prefill": false,
34
+ "local_ranks_size": 8,
35
+ "logical_nc_config": 1,
36
+ "max_batch_size": 32,
37
+ "max_context_length": 4096,
38
+ "max_topk": 256,
39
+ "mlp_kernel_enabled": false,
40
+ "mlp_kernel_fuse_residual_add": false,
41
+ "n_active_tokens": 4096,
42
+ "neuronxcc_version": "2.17.194.0+d312836f",
43
+ "num_cores_per_group": 1,
44
+ "on_device_sampling": true,
45
+ "optimum_neuron_version": "0.2.0.dev7",
46
+ "output_logits": false,
47
+ "padding_side": "right",
48
+ "pp_degree": 1,
49
+ "qk_layernorm": false,
50
+ "qkv_kernel_enabled": false,
51
+ "rpl_reduce_dtype": "bfloat16",
52
+ "sequence_length": 4096,
53
+ "sequence_parallel_enabled": false,
54
+ "speculation_length": 0,
55
+ "start_rank_id": 0,
56
+ "target": null,
57
+ "torch_dtype": "bfloat16",
58
+ "tp_degree": 8,
59
+ "vocab_parallel": false
60
+ },
61
+ "num_attention_heads": 32,
62
+ "num_hidden_layers": 32,
63
+ "num_key_value_heads": 8,
64
+ "pretraining_tp": 1,
65
+ "rms_norm_eps": 1e-05,
66
+ "rope_scaling": {
67
+ "factor": 8.0,
68
+ "high_freq_factor": 4.0,
69
+ "low_freq_factor": 1.0,
70
+ "original_max_position_embeddings": 8192,
71
+ "rope_type": "llama3"
72
+ },
73
+ "rope_theta": 500000.0,
74
+ "tie_word_embeddings": false,
75
+ "use_cache": true,
76
+ "vocab_size": 128256
77
+ }
neuronxcc-2.17.194.0+d312836f/0_REGISTRY/0.2.0.dev7/llama/unsloth/Llama-3.2-1B-Instruct/a01ee1e6bc19cfb5be32.json ADDED
@@ -0,0 +1,78 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "unsloth/Llama-3.2-1B-Instruct",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "head_dim": 64,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 2048,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 8192,
15
+ "max_position_embeddings": 131072,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "neuron": {
19
+ "_serialized_key": "NxDNeuronConfig",
20
+ "async_mode": false,
21
+ "attn_kernel_enabled": false,
22
+ "batch_size": 4,
23
+ "capacity_factor": null,
24
+ "cc_pipeline_tiling_factor": 2,
25
+ "checkpoint_id": "unsloth/Llama-3.2-1B-Instruct",
26
+ "checkpoint_revision": "5a8abab4a5d6f164389b1079fb721cfab8d7126c",
27
+ "continuous_batching": true,
28
+ "enable_bucketing": false,
29
+ "ep_degree": 1,
30
+ "flash_decoding_enabled": false,
31
+ "fused_qkv": true,
32
+ "glu_mlp": true,
33
+ "is_chunked_prefill": false,
34
+ "local_ranks_size": 2,
35
+ "logical_nc_config": 1,
36
+ "max_batch_size": 4,
37
+ "max_context_length": 4096,
38
+ "max_topk": 256,
39
+ "mlp_kernel_enabled": false,
40
+ "mlp_kernel_fuse_residual_add": false,
41
+ "n_active_tokens": 4096,
42
+ "neuronxcc_version": "2.17.194.0+d312836f",
43
+ "num_cores_per_group": 1,
44
+ "on_device_sampling": false,
45
+ "optimum_neuron_version": "0.2.0.dev7",
46
+ "output_logits": false,
47
+ "padding_side": "right",
48
+ "pp_degree": 1,
49
+ "qk_layernorm": false,
50
+ "qkv_kernel_enabled": false,
51
+ "rpl_reduce_dtype": "float16",
52
+ "sequence_length": 4096,
53
+ "sequence_parallel_enabled": false,
54
+ "speculation_length": 0,
55
+ "start_rank_id": 0,
56
+ "target": null,
57
+ "torch_dtype": "float16",
58
+ "tp_degree": 2,
59
+ "vocab_parallel": false
60
+ },
61
+ "num_attention_heads": 32,
62
+ "num_hidden_layers": 16,
63
+ "num_key_value_heads": 8,
64
+ "pretraining_tp": 1,
65
+ "rms_norm_eps": 1e-05,
66
+ "rope_scaling": {
67
+ "factor": 32.0,
68
+ "high_freq_factor": 4.0,
69
+ "low_freq_factor": 1.0,
70
+ "original_max_position_embeddings": 8192,
71
+ "rope_type": "llama3"
72
+ },
73
+ "rope_theta": 500000.0,
74
+ "tie_word_embeddings": true,
75
+ "unsloth_fixed": true,
76
+ "use_cache": true,
77
+ "vocab_size": 128256
78
+ }
neuronxcc-2.17.194.0+d312836f/MODULE_a30822a1b2d1fbcee30c+bfe5714b/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ "--auto-cast=none --model-type=transformer --tensorizer-options='--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-dge-dma --vectorize-strided-dma ' -O2 --internal-num-neuroncores-per-sengine=1 --logfile=/tmp/nxd_model/context_encoding_model/_tp0_bk0/log-neuron-cc.txt"
neuronxcc-2.17.194.0+d312836f/MODULE_a30822a1b2d1fbcee30c+bfe5714b/model.done ADDED
File without changes
neuronxcc-2.17.194.0+d312836f/MODULE_a30822a1b2d1fbcee30c+bfe5714b/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:974014e95c3bc47c8e83b01095f6407ed002d68ff08e93827c4a9589db1f3295
3
+ size 919863
neuronxcc-2.17.194.0+d312836f/MODULE_a30822a1b2d1fbcee30c+bfe5714b/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5aab7ca70c91652f4d87340f48b339c5ffe22d95c368408c0979f30b58e10fb0
3
+ size 32646144
neuronxcc-2.17.194.0+d312836f/MODULE_a849442b615562b13ba6+431f5505/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ "--model-type=transformer -O1 --lnc=1 --internal-hlo2tensorizer-options=--experimental-unsafe-fp8e4m3fn-as-fp8e4m3 --logfile=/tmp/nxd_model/layout_opt/log-neuron-cc.txt"
neuronxcc-2.17.194.0+d312836f/MODULE_a849442b615562b13ba6+431f5505/model.done ADDED
File without changes
neuronxcc-2.17.194.0+d312836f/MODULE_a849442b615562b13ba6+431f5505/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:df45ba39345e926a837e7279b93c0aa0b4d0d0a6300ce0cc94e0825594278d87
3
+ size 136016
neuronxcc-2.17.194.0+d312836f/MODULE_a849442b615562b13ba6+431f5505/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:69e1534acc92cfbdc06bb5206a788c3368ab7c56ccd791e49381ad50da3f04a8
3
+ size 2202624
neuronxcc-2.17.194.0+d312836f/MODULE_ccfcad18cf4ad9b6e4bc+165e9558/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ "--auto-cast=none --model-type=transformer --tensorizer-options='--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-dge-dma --vectorize-strided-dma ' -O2 --internal-num-neuroncores-per-sengine=1 --logfile=/tmp/nxd_model/token_generation_model/_tp0_bk0/log-neuron-cc.txt --enable-internal-neff-wrapper"
neuronxcc-2.17.194.0+d312836f/MODULE_ccfcad18cf4ad9b6e4bc+165e9558/model.done ADDED
File without changes
neuronxcc-2.17.194.0+d312836f/MODULE_ccfcad18cf4ad9b6e4bc+165e9558/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8f026ef2394f1ab68b86b8f5a76efcd1159ed906b0eff506aaec9da272b8db28
3
+ size 777185
neuronxcc-2.17.194.0+d312836f/MODULE_ccfcad18cf4ad9b6e4bc+165e9558/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dbb584faccdc5c239df99820750fe6464ba47dfb5627956c255247ae77b781ee
3
+ size 6994944
neuronxcc-2.17.194.0+d312836f/MODULE_ccfcad18cf4ad9b6e4bc+165e9558/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d06ab7ee1fbb9965f4d0d637e9fbba48f1b6083991440a25f56f510444431256
3
+ size 7132888