Hugging Face's logo

aws-neuron
/
optimum-neuron-cache

Model card Files
xet
 Community
709
dacorvo HF Staff commited on
Commit
a4b55d1
·
verified ·
1 Parent(s): 8c375aa

Synchronizing local compiler cache.

Browse files
Files changed (11) hide show
  1. .gitattributes +3 -0
  2. neuronxcc-2.26.6360.0+6f180f47/0_REGISTRY/0.4.7.dev0/c033ab36fa35321d96cda66c61c1b610e3f28f44b24f28439782fbe605dc5a48/730c60e40b2012bfa8bb.json +96 -0
  3. neuronxcc-2.26.6360.0+6f180f47/MODULE_24f0bb26b2ca667fdcf7+84337dd9/compile_flags.json +1 -0
  4. neuronxcc-2.26.6360.0+6f180f47/MODULE_24f0bb26b2ca667fdcf7+84337dd9/model.done +0 -0
  5. neuronxcc-2.26.6360.0+6f180f47/MODULE_24f0bb26b2ca667fdcf7+84337dd9/model.hlo_module.pb +3 -0
  6. neuronxcc-2.26.6360.0+6f180f47/MODULE_24f0bb26b2ca667fdcf7+84337dd9/model.neff +3 -0
  7. neuronxcc-2.26.6360.0+6f180f47/MODULE_24f0bb26b2ca667fdcf7+84337dd9/wrapped_neff.hlo +3 -0
  8. neuronxcc-2.26.6360.0+6f180f47/MODULE_f16325aea956a251ad69+12dd9119/compile_flags.json +1 -0
  9. neuronxcc-2.26.6360.0+6f180f47/MODULE_f16325aea956a251ad69+12dd9119/model.done +0 -0
  10. neuronxcc-2.26.6360.0+6f180f47/MODULE_f16325aea956a251ad69+12dd9119/model.hlo_module.pb +3 -0
  11. neuronxcc-2.26.6360.0+6f180f47/MODULE_f16325aea956a251ad69+12dd9119/model.neff +3 -0
.gitattributes CHANGED
@@ -17837,3 +17837,6 @@ neuronxcc-2.26.6360.0+6f180f47/MODULE_db92952a9c1e0a17922c+12dd9119/model.neff f
17837
  neuronxcc-2.26.6360.0+6f180f47/MODULE_996074a4d504caa17cc2+12dd9119/model.neff filter=lfs diff=lfs merge=lfs -text
17838
  neuronxcc-2.26.6360.0+6f180f47/MODULE_bc66647825c61293a609+84337dd9/model.neff filter=lfs diff=lfs merge=lfs -text
17839
  neuronxcc-2.26.6360.0+6f180f47/MODULE_bc66647825c61293a609+84337dd9/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
 
 
 
 
17837
  neuronxcc-2.26.6360.0+6f180f47/MODULE_996074a4d504caa17cc2+12dd9119/model.neff filter=lfs diff=lfs merge=lfs -text
17838
  neuronxcc-2.26.6360.0+6f180f47/MODULE_bc66647825c61293a609+84337dd9/model.neff filter=lfs diff=lfs merge=lfs -text
17839
  neuronxcc-2.26.6360.0+6f180f47/MODULE_bc66647825c61293a609+84337dd9/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
17840
+ neuronxcc-2.26.6360.0+6f180f47/MODULE_24f0bb26b2ca667fdcf7+84337dd9/model.neff filter=lfs diff=lfs merge=lfs -text
17841
+ neuronxcc-2.26.6360.0+6f180f47/MODULE_24f0bb26b2ca667fdcf7+84337dd9/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
17842
+ neuronxcc-2.26.6360.0+6f180f47/MODULE_f16325aea956a251ad69+12dd9119/model.neff filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.26.6360.0+6f180f47/0_REGISTRY/0.4.7.dev0/c033ab36fa35321d96cda66c61c1b610e3f28f44b24f28439782fbe605dc5a48/730c60e40b2012bfa8bb.json ADDED
@@ -0,0 +1,96 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "Qwen/Qwen3-8B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "Qwen3ForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "dtype": "bfloat16",
11
+ "head_dim": 128,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 4096,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 12288,
16
+ "layer_types": [
17
+ "full_attention",
18
+ "full_attention",
19
+ "full_attention",
20
+ "full_attention",
21
+ "full_attention",
22
+ "full_attention",
23
+ "full_attention",
24
+ "full_attention",
25
+ "full_attention",
26
+ "full_attention",
27
+ "full_attention",
28
+ "full_attention",
29
+ "full_attention",
30
+ "full_attention",
31
+ "full_attention",
32
+ "full_attention",
33
+ "full_attention",
34
+ "full_attention",
35
+ "full_attention",
36
+ "full_attention",
37
+ "full_attention",
38
+ "full_attention",
39
+ "full_attention",
40
+ "full_attention",
41
+ "full_attention",
42
+ "full_attention",
43
+ "full_attention",
44
+ "full_attention",
45
+ "full_attention",
46
+ "full_attention",
47
+ "full_attention",
48
+ "full_attention",
49
+ "full_attention",
50
+ "full_attention",
51
+ "full_attention",
52
+ "full_attention"
53
+ ],
54
+ "max_position_embeddings": 40960,
55
+ "max_window_layers": 36,
56
+ "model_type": "qwen3",
57
+ "neuron": {
58
+ "_serialized_key": "NxDNeuronConfig",
59
+ "batch_size": 8,
60
+ "capacity_factor": null,
61
+ "checkpoint_id": "Qwen/Qwen3-8B",
62
+ "checkpoint_revision": "b968826d9c46dd6066d109eabc6255188de91218",
63
+ "continuous_batching": true,
64
+ "ep_degree": 1,
65
+ "fused_qkv": true,
66
+ "glu_mlp": true,
67
+ "local_ranks_size": 2,
68
+ "max_batch_size": 8,
69
+ "max_context_length": 4096,
70
+ "max_topk": 256,
71
+ "n_active_tokens": 4096,
72
+ "neuronxcc_version": "2.26.6360.0+6f180f47",
73
+ "on_device_sampling": false,
74
+ "optimum_neuron_version": "0.4.7.dev0",
75
+ "output_logits": false,
76
+ "pp_degree": 1,
77
+ "prefill_chunk_size": 1024,
78
+ "sequence_length": 4096,
79
+ "speculation_length": 0,
80
+ "start_rank_id": 0,
81
+ "target": "trn1",
82
+ "torch_dtype": "bfloat16",
83
+ "tp_degree": 2
84
+ },
85
+ "num_attention_heads": 32,
86
+ "num_hidden_layers": 36,
87
+ "num_key_value_heads": 8,
88
+ "rms_norm_eps": 1e-06,
89
+ "rope_scaling": null,
90
+ "rope_theta": 1000000,
91
+ "sliding_window": null,
92
+ "tie_word_embeddings": false,
93
+ "use_cache": true,
94
+ "use_sliding_window": false,
95
+ "vocab_size": 151936
96
+ }
neuronxcc-2.26.6360.0+6f180f47/MODULE_24f0bb26b2ca667fdcf7+84337dd9/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn1", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=1", "--verbose=35", "--logfile=/tmp/nxd_model/token_generation/_tp0_bk0/log-neuron-cc.txt", "--enable-internal-neff-wrapper"]
neuronxcc-2.26.6360.0+6f180f47/MODULE_24f0bb26b2ca667fdcf7+84337dd9/model.done ADDED
File without changes
neuronxcc-2.26.6360.0+6f180f47/MODULE_24f0bb26b2ca667fdcf7+84337dd9/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f4bafa5b4906588b9ddfabd3225a16f349d17ab23311b14503315ccf9150de28
3
+ size 3684395
neuronxcc-2.26.6360.0+6f180f47/MODULE_24f0bb26b2ca667fdcf7+84337dd9/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a39b42fa7493d408bdbbecf7cab16be5c68210fd7da4b9eb4f1691119d4c4e9c
3
+ size 13394944
neuronxcc-2.26.6360.0+6f180f47/MODULE_24f0bb26b2ca667fdcf7+84337dd9/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d968215bc32c550da5bba13ee5f4e45dc767e2622b835c8360a5bdbdeda00bc7
3
+ size 13569901
neuronxcc-2.26.6360.0+6f180f47/MODULE_f16325aea956a251ad69+12dd9119/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn1", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=1", "--verbose=35", "--logfile=/tmp/nxd_model/chunked_prefill/_tp0_bk0/log-neuron-cc.txt"]
neuronxcc-2.26.6360.0+6f180f47/MODULE_f16325aea956a251ad69+12dd9119/model.done ADDED
File without changes
neuronxcc-2.26.6360.0+6f180f47/MODULE_f16325aea956a251ad69+12dd9119/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cecf06617a6c2068ede09cf28dfedb01b908957fa06052b55331458f49bbbf35
3
+ size 5163648
neuronxcc-2.26.6360.0+6f180f47/MODULE_f16325aea956a251ad69+12dd9119/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2e48ac7e6e3b14c7358291ac7ba2c09a7851c69d5e40b8ec3f315acba41e7e96
3
+ size 5438464