Hugging Face's logo

aws-neuron
/
optimum-neuron-cache

Model card Files Community
474
dacorvo HF Staff commited on
Commit
dbb0edb
·
verified ·
1 Parent(s): b623de4

Synchronizing local compiler cache.

Browse files
Files changed (27) hide show
  1. .gitattributes +7 -0
  2. neuronxcc-2.17.194.0+d312836f/0_REGISTRY/0.3.0.dev0/qwen2/deepseek-ai/DeepSeek-R1-Distill-Qwen-7B/be9ce236e618e23d37e3.json +71 -0
  3. neuronxcc-2.17.194.0+d312836f/MODULE_10f5c14a0a8444e83fb4+bfe5714b/compile_flags.json +1 -0
  4. neuronxcc-2.17.194.0+d312836f/MODULE_10f5c14a0a8444e83fb4+bfe5714b/model.done +0 -0
  5. neuronxcc-2.17.194.0+d312836f/MODULE_10f5c14a0a8444e83fb4+bfe5714b/model.hlo_module.pb +3 -0
  6. neuronxcc-2.17.194.0+d312836f/MODULE_10f5c14a0a8444e83fb4+bfe5714b/model.neff +3 -0
  7. neuronxcc-2.17.194.0+d312836f/MODULE_5492c571f19a0b5a01ee+bfe5714b/compile_flags.json +1 -0
  8. neuronxcc-2.17.194.0+d312836f/MODULE_5492c571f19a0b5a01ee+bfe5714b/model.hlo_module.pb +3 -0
  9. neuronxcc-2.17.194.0+d312836f/MODULE_6d8044cb45e8291d5efb+165e9558/compile_flags.json +1 -0
  10. neuronxcc-2.17.194.0+d312836f/MODULE_6d8044cb45e8291d5efb+165e9558/model.done +0 -0
  11. neuronxcc-2.17.194.0+d312836f/MODULE_6d8044cb45e8291d5efb+165e9558/model.hlo_module.pb +3 -0
  12. neuronxcc-2.17.194.0+d312836f/MODULE_6d8044cb45e8291d5efb+165e9558/model.neff +3 -0
  13. neuronxcc-2.17.194.0+d312836f/MODULE_6d8044cb45e8291d5efb+165e9558/wrapped_neff.hlo +3 -0
  14. neuronxcc-2.17.194.0+d312836f/MODULE_8341f89f99e9d3ed7317+165e9558/compile_flags.json +1 -0
  15. neuronxcc-2.17.194.0+d312836f/MODULE_8341f89f99e9d3ed7317+165e9558/model.done +0 -0
  16. neuronxcc-2.17.194.0+d312836f/MODULE_8341f89f99e9d3ed7317+165e9558/model.hlo_module.pb +3 -0
  17. neuronxcc-2.17.194.0+d312836f/MODULE_8341f89f99e9d3ed7317+165e9558/model.neff +3 -0
  18. neuronxcc-2.17.194.0+d312836f/MODULE_8341f89f99e9d3ed7317+165e9558/wrapped_neff.hlo +3 -0
  19. neuronxcc-2.17.194.0+d312836f/MODULE_b377cd22814ba489f609+165e9558/compile_flags.json +1 -0
  20. neuronxcc-2.17.194.0+d312836f/MODULE_b377cd22814ba489f609+165e9558/model.hlo_module.pb +3 -0
  21. neuronxcc-2.17.194.0+d312836f/MODULE_d3dacc450634f2463c59+165e9558/compile_flags.json +1 -0
  22. neuronxcc-2.17.194.0+d312836f/MODULE_d3dacc450634f2463c59+165e9558/model.done +0 -0
  23. neuronxcc-2.17.194.0+d312836f/MODULE_d3dacc450634f2463c59+165e9558/model.hlo_module.pb +3 -0
  24. neuronxcc-2.17.194.0+d312836f/MODULE_d3dacc450634f2463c59+165e9558/model.neff +3 -0
  25. neuronxcc-2.17.194.0+d312836f/MODULE_d3dacc450634f2463c59+165e9558/wrapped_neff.hlo +3 -0
  26. neuronxcc-2.17.194.0+d312836f/MODULE_e8864eafb9123cca22ae+bfe5714b/compile_flags.json +1 -0
  27. neuronxcc-2.17.194.0+d312836f/MODULE_e8864eafb9123cca22ae+bfe5714b/model.hlo_module.pb +3 -0
.gitattributes CHANGED
@@ -8455,3 +8455,10 @@ neuronxcc-2.17.194.0+d312836f/MODULE_185dab62e9cbd950a909+bfe5714b/model.neff fi
8455
  neuronxcc-2.17.194.0+d312836f/MODULE_2f42eb97be7f825d9062+431f5505/model.neff filter=lfs diff=lfs merge=lfs -text
8456
  neuronxcc-2.17.194.0+d312836f/MODULE_340b27ffe6c2d35777a3+165e9558/model.neff filter=lfs diff=lfs merge=lfs -text
8457
  neuronxcc-2.17.194.0+d312836f/MODULE_340b27ffe6c2d35777a3+165e9558/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
 
 
 
8455
  neuronxcc-2.17.194.0+d312836f/MODULE_2f42eb97be7f825d9062+431f5505/model.neff filter=lfs diff=lfs merge=lfs -text
8456
  neuronxcc-2.17.194.0+d312836f/MODULE_340b27ffe6c2d35777a3+165e9558/model.neff filter=lfs diff=lfs merge=lfs -text
8457
  neuronxcc-2.17.194.0+d312836f/MODULE_340b27ffe6c2d35777a3+165e9558/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
8458
+ neuronxcc-2.17.194.0+d312836f/MODULE_10f5c14a0a8444e83fb4+bfe5714b/model.neff filter=lfs diff=lfs merge=lfs -text
8459
+ neuronxcc-2.17.194.0+d312836f/MODULE_6d8044cb45e8291d5efb+165e9558/model.neff filter=lfs diff=lfs merge=lfs -text
8460
+ neuronxcc-2.17.194.0+d312836f/MODULE_6d8044cb45e8291d5efb+165e9558/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
8461
+ neuronxcc-2.17.194.0+d312836f/MODULE_8341f89f99e9d3ed7317+165e9558/model.neff filter=lfs diff=lfs merge=lfs -text
8462
+ neuronxcc-2.17.194.0+d312836f/MODULE_8341f89f99e9d3ed7317+165e9558/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
8463
+ neuronxcc-2.17.194.0+d312836f/MODULE_d3dacc450634f2463c59+165e9558/model.neff filter=lfs diff=lfs merge=lfs -text
8464
+ neuronxcc-2.17.194.0+d312836f/MODULE_d3dacc450634f2463c59+165e9558/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.17.194.0+d312836f/0_REGISTRY/0.3.0.dev0/qwen2/deepseek-ai/DeepSeek-R1-Distill-Qwen-7B/be9ce236e618e23d37e3.json ADDED
@@ -0,0 +1,71 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "deepseek-ai/DeepSeek-R1-Distill-Qwen-7B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "Qwen2ForCausalLM"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "hidden_act": "silu",
10
+ "hidden_size": 3584,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 18944,
13
+ "max_position_embeddings": 131072,
14
+ "max_window_layers": 28,
15
+ "model_type": "qwen2",
16
+ "neuron": {
17
+ "_serialized_key": "NxDNeuronConfig",
18
+ "async_mode": false,
19
+ "attn_kernel_enabled": false,
20
+ "batch_size": 1,
21
+ "capacity_factor": null,
22
+ "cc_pipeline_tiling_factor": 2,
23
+ "checkpoint_id": "deepseek-ai/DeepSeek-R1-Distill-Qwen-7B",
24
+ "checkpoint_revision": "916b56a44061fd5cd7d6a8fb632557ed4f724f60",
25
+ "continuous_batching": false,
26
+ "enable_bucketing": false,
27
+ "ep_degree": 1,
28
+ "flash_decoding_enabled": false,
29
+ "fused_qkv": false,
30
+ "glu_mlp": true,
31
+ "is_chunked_prefill": false,
32
+ "local_ranks_size": 8,
33
+ "logical_nc_config": 1,
34
+ "max_batch_size": 1,
35
+ "max_context_length": 4096,
36
+ "max_topk": 256,
37
+ "mlp_kernel_enabled": false,
38
+ "mlp_kernel_fuse_residual_add": false,
39
+ "n_active_tokens": 4096,
40
+ "neuronxcc_version": "2.17.194.0+d312836f",
41
+ "num_cores_per_group": 1,
42
+ "on_device_sampling": true,
43
+ "optimum_neuron_version": "0.3.0.dev0",
44
+ "output_logits": false,
45
+ "padding_side": "right",
46
+ "pp_degree": 1,
47
+ "qk_layernorm": false,
48
+ "qkv_kernel_enabled": false,
49
+ "rpl_reduce_dtype": "bfloat16",
50
+ "sequence_length": 4096,
51
+ "sequence_parallel_enabled": false,
52
+ "speculation_length": 0,
53
+ "start_rank_id": 0,
54
+ "target": null,
55
+ "torch_dtype": "bfloat16",
56
+ "tp_degree": 8,
57
+ "vocab_parallel": false
58
+ },
59
+ "num_attention_heads": 28,
60
+ "num_hidden_layers": 28,
61
+ "num_key_value_heads": 4,
62
+ "rms_norm_eps": 1e-06,
63
+ "rope_scaling": null,
64
+ "rope_theta": 10000,
65
+ "sliding_window": 4096,
66
+ "tie_word_embeddings": false,
67
+ "use_cache": true,
68
+ "use_mrope": false,
69
+ "use_sliding_window": false,
70
+ "vocab_size": 152064
71
+ }
neuronxcc-2.17.194.0+d312836f/MODULE_10f5c14a0a8444e83fb4+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_10f5c14a0a8444e83fb4+bfe5714b/model.done ADDED
File without changes
neuronxcc-2.17.194.0+d312836f/MODULE_10f5c14a0a8444e83fb4+bfe5714b/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f18842e2a1284a9092194bff7aae25c7ed5b1ebb405a74a19f6946b44fd15aca
3
+ size 880540
neuronxcc-2.17.194.0+d312836f/MODULE_10f5c14a0a8444e83fb4+bfe5714b/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:72e8181484797c42eb20f5debbe6a6cb8693f7ebc1541e8447a903154117731b
3
+ size 30096384
neuronxcc-2.17.194.0+d312836f/MODULE_5492c571f19a0b5a01ee+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_5492c571f19a0b5a01ee+bfe5714b/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b8f942d1a97b9852b39e8b10f7cd9d14de6e745bb624b085ee71461089c9e60b
3
+ size 921363
neuronxcc-2.17.194.0+d312836f/MODULE_6d8044cb45e8291d5efb+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_6d8044cb45e8291d5efb+165e9558/model.done ADDED
File without changes
neuronxcc-2.17.194.0+d312836f/MODULE_6d8044cb45e8291d5efb+165e9558/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d6eacb487408eb1861d51ce316bdf14c1e44113ae051523b0dc4d3e1f44c4a2d
3
+ size 780477
neuronxcc-2.17.194.0+d312836f/MODULE_6d8044cb45e8291d5efb+165e9558/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:01159befb7db5b0895a4da0b5bb630145705fbde76cbc94b97341444f352466c
3
+ size 9841664
neuronxcc-2.17.194.0+d312836f/MODULE_6d8044cb45e8291d5efb+165e9558/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0ba44d958d575a772d69d5b929a6111f0e840c1444ed3b010dc3560c163d5606
3
+ size 10036243
neuronxcc-2.17.194.0+d312836f/MODULE_8341f89f99e9d3ed7317+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_8341f89f99e9d3ed7317+165e9558/model.done ADDED
File without changes
neuronxcc-2.17.194.0+d312836f/MODULE_8341f89f99e9d3ed7317+165e9558/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7d69771cc9809f9f6c987c4102911688cb57898df2700f6cfd03ad417100f741
3
+ size 812531
neuronxcc-2.17.194.0+d312836f/MODULE_8341f89f99e9d3ed7317+165e9558/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b5b1f03635bc93b6652d256dac9c21e910395f467ac5b930701f883cccfcee48
3
+ size 6288384
neuronxcc-2.17.194.0+d312836f/MODULE_8341f89f99e9d3ed7317+165e9558/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2bb37a77a7245a6d8cbedaf4470f8a62816ae820b3d4ef1835be24a4242337f6
3
+ size 6482951
neuronxcc-2.17.194.0+d312836f/MODULE_b377cd22814ba489f609+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_b377cd22814ba489f609+165e9558/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c1d96d4ae8f3326636f0a07b2d2560b3a45ea18d41f2b5977021af8c14f12b08
3
+ size 780477
neuronxcc-2.17.194.0+d312836f/MODULE_d3dacc450634f2463c59+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_d3dacc450634f2463c59+165e9558/model.done ADDED
File without changes
neuronxcc-2.17.194.0+d312836f/MODULE_d3dacc450634f2463c59+165e9558/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:55ed94ad23233174498018688a41091e6b27e012696d01f76d815995c4762efa
3
+ size 798755
neuronxcc-2.17.194.0+d312836f/MODULE_d3dacc450634f2463c59+165e9558/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ef2c9747890975dc4b462d495288083b295a23b9889fcbb74872cce3afad068d
3
+ size 2182144
neuronxcc-2.17.194.0+d312836f/MODULE_d3dacc450634f2463c59+165e9558/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cd08468644fb2e09c3e5946ecdb38095f95f58a1e4466751338e57e2e146f7d3
3
+ size 2351604
neuronxcc-2.17.194.0+d312836f/MODULE_e8864eafb9123cca22ae+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_e8864eafb9123cca22ae+bfe5714b/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:01bb5c764e4a93c76da3b751d21927ee3894e7706bb4b7eb5e96818cd37a9a5f
3
+ size 897476