Hugging Face's logo

aws-neuron
/
optimum-neuron-cache

Model card Files
xet
Community
628
dacorvo HF Staff commited on
Commit
475dc4d
·
verified ·
1 Parent(s): 9b4b4a4

Synchronizing local compiler cache.

Browse files
Files changed (11) hide show
  1. .gitattributes +3 -0
  2. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.3.dev2/qwen3/Qwen/Qwen3-8B/d6a3058cf50b03b1d2bf.json +95 -0
  3. neuronxcc-2.21.33363.0+82129205/MODULE_0d5453b7886405e7efd0+24129607/compile_flags.json +1 -0
  4. neuronxcc-2.21.33363.0+82129205/MODULE_0d5453b7886405e7efd0+24129607/model.done +0 -0
  5. neuronxcc-2.21.33363.0+82129205/MODULE_0d5453b7886405e7efd0+24129607/model.hlo_module.pb +3 -0
  6. neuronxcc-2.21.33363.0+82129205/MODULE_0d5453b7886405e7efd0+24129607/model.neff +3 -0
  7. neuronxcc-2.21.33363.0+82129205/MODULE_a96e8592c2ef49ab0738+a02c3a36/compile_flags.json +1 -0
  8. neuronxcc-2.21.33363.0+82129205/MODULE_a96e8592c2ef49ab0738+a02c3a36/model.done +0 -0
  9. neuronxcc-2.21.33363.0+82129205/MODULE_a96e8592c2ef49ab0738+a02c3a36/model.hlo_module.pb +3 -0
  10. neuronxcc-2.21.33363.0+82129205/MODULE_a96e8592c2ef49ab0738+a02c3a36/model.neff +3 -0
  11. neuronxcc-2.21.33363.0+82129205/MODULE_a96e8592c2ef49ab0738+a02c3a36/wrapped_neff.hlo +3 -0
.gitattributes CHANGED
@@ -13745,3 +13745,6 @@ neuronxcc-2.21.33363.0+82129205/MODULE_b527ea6afa2339e878d2+24129607/model.neff
13745
  neuronxcc-2.21.33363.0+82129205/MODULE_373136c34475445511e5+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
13746
  neuronxcc-2.21.33363.0+82129205/MODULE_373136c34475445511e5+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
13747
  neuronxcc-2.21.33363.0+82129205/MODULE_bb1b872fa136444a4e81+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
 
 
 
 
13745
  neuronxcc-2.21.33363.0+82129205/MODULE_373136c34475445511e5+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
13746
  neuronxcc-2.21.33363.0+82129205/MODULE_373136c34475445511e5+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
13747
  neuronxcc-2.21.33363.0+82129205/MODULE_bb1b872fa136444a4e81+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
13748
+ neuronxcc-2.21.33363.0+82129205/MODULE_0d5453b7886405e7efd0+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
13749
+ neuronxcc-2.21.33363.0+82129205/MODULE_a96e8592c2ef49ab0738+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
13750
+ neuronxcc-2.21.33363.0+82129205/MODULE_a96e8592c2ef49ab0738+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.3.dev2/qwen3/Qwen/Qwen3-8B/d6a3058cf50b03b1d2bf.json ADDED
@@ -0,0 +1,95 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": 1,
60
+ "capacity_factor": null,
61
+ "checkpoint_id": "Qwen/Qwen3-8B",
62
+ "checkpoint_revision": "b968826d9c46dd6066d109eabc6255188de91218",
63
+ "continuous_batching": false,
64
+ "ep_degree": 1,
65
+ "fused_qkv": true,
66
+ "glu_mlp": true,
67
+ "local_ranks_size": 2,
68
+ "max_batch_size": 1,
69
+ "max_context_length": 4096,
70
+ "max_topk": 256,
71
+ "n_active_tokens": 4096,
72
+ "neuronxcc_version": "2.21.33363.0+82129205",
73
+ "on_device_sampling": true,
74
+ "optimum_neuron_version": "0.4.3.dev2",
75
+ "output_logits": false,
76
+ "pp_degree": 1,
77
+ "sequence_length": 4096,
78
+ "speculation_length": 0,
79
+ "start_rank_id": 0,
80
+ "target": "trn1",
81
+ "torch_dtype": "bfloat16",
82
+ "tp_degree": 2
83
+ },
84
+ "num_attention_heads": 32,
85
+ "num_hidden_layers": 36,
86
+ "num_key_value_heads": 8,
87
+ "rms_norm_eps": 1e-06,
88
+ "rope_scaling": null,
89
+ "rope_theta": 1000000,
90
+ "sliding_window": null,
91
+ "tie_word_embeddings": false,
92
+ "use_cache": true,
93
+ "use_sliding_window": false,
94
+ "vocab_size": 151936
95
+ }
neuronxcc-2.21.33363.0+82129205/MODULE_0d5453b7886405e7efd0+24129607/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", "--logfile=/tmp/nxd_model/context_encoding/_tp0_bk0/log-neuron-cc.txt"]
neuronxcc-2.21.33363.0+82129205/MODULE_0d5453b7886405e7efd0+24129607/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_0d5453b7886405e7efd0+24129607/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6a6b690c844a7d1ac3ba3a14fd615b1be14abdbe4f767533ff260aa3c2548e9c
3
+ size 1085355
neuronxcc-2.21.33363.0+82129205/MODULE_0d5453b7886405e7efd0+24129607/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b9a4454c19925f9fb64a855c71c79da81fc1ed106755c4429b0a622767f5ba15
3
+ size 6964224
neuronxcc-2.21.33363.0+82129205/MODULE_a96e8592c2ef49ab0738+a02c3a36/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", "--logfile=/tmp/nxd_model/token_generation/_tp0_bk0/log-neuron-cc.txt", "--enable-internal-neff-wrapper"]
neuronxcc-2.21.33363.0+82129205/MODULE_a96e8592c2ef49ab0738+a02c3a36/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_a96e8592c2ef49ab0738+a02c3a36/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5a69e826c689c0049d797431c23c3cd5df47b7ec81604d414df26335478fa0a6
3
+ size 943563
neuronxcc-2.21.33363.0+82129205/MODULE_a96e8592c2ef49ab0738+a02c3a36/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fceb04d649c6264451ab4f7614d232e6e5f5ab7866cff01313368744c68d7132
3
+ size 10302464
neuronxcc-2.21.33363.0+82129205/MODULE_a96e8592c2ef49ab0738+a02c3a36/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4a01f84a3900b57cd6a43b6cb48302c5097471853c3831deec99c705ecc4c763
3
+ size 10507391