Hugging Face's logo

aws-neuron
/
optimum-neuron-cache

Model card Files
xet
Community
654
dacorvo HF Staff commited on
Commit
4a62b90
·
verified ·
1 Parent(s): 7925bfc

Synchronizing local compiler cache.

Browse files
Files changed (11) hide show
  1. .gitattributes +3 -0
  2. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/qwen3/Qwen/Qwen3-8B/dcd74ada1295d4fd053d.json +95 -0
  3. neuronxcc-2.21.33363.0+82129205/MODULE_47553fb37e8438019a32+24129607/compile_flags.json +1 -0
  4. neuronxcc-2.21.33363.0+82129205/MODULE_47553fb37e8438019a32+24129607/model.done +0 -0
  5. neuronxcc-2.21.33363.0+82129205/MODULE_47553fb37e8438019a32+24129607/model.hlo_module.pb +3 -0
  6. neuronxcc-2.21.33363.0+82129205/MODULE_47553fb37e8438019a32+24129607/model.neff +3 -0
  7. neuronxcc-2.21.33363.0+82129205/MODULE_686171fdb53c3e472a17+a02c3a36/compile_flags.json +1 -0
  8. neuronxcc-2.21.33363.0+82129205/MODULE_686171fdb53c3e472a17+a02c3a36/model.done +0 -0
  9. neuronxcc-2.21.33363.0+82129205/MODULE_686171fdb53c3e472a17+a02c3a36/model.hlo_module.pb +3 -0
  10. neuronxcc-2.21.33363.0+82129205/MODULE_686171fdb53c3e472a17+a02c3a36/model.neff +3 -0
  11. neuronxcc-2.21.33363.0+82129205/MODULE_686171fdb53c3e472a17+a02c3a36/wrapped_neff.hlo +3 -0
.gitattributes CHANGED
@@ -15283,3 +15283,6 @@ neuronxcc-2.21.33363.0+82129205/MODULE_f95f635e3a6cb68e4619+e0c92c88/model.neff
15283
  neuronxcc-2.21.33363.0+82129205/MODULE_5238577c83dc087eb2d8+2dde74c7/model.neff filter=lfs diff=lfs merge=lfs -text
15284
  neuronxcc-2.21.33363.0+82129205/MODULE_5238577c83dc087eb2d8+2dde74c7/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
15285
  neuronxcc-2.21.33363.0+82129205/MODULE_c09b33112d0c8495e9d4+283df001/model.neff filter=lfs diff=lfs merge=lfs -text
 
 
 
 
15283
  neuronxcc-2.21.33363.0+82129205/MODULE_5238577c83dc087eb2d8+2dde74c7/model.neff filter=lfs diff=lfs merge=lfs -text
15284
  neuronxcc-2.21.33363.0+82129205/MODULE_5238577c83dc087eb2d8+2dde74c7/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
15285
  neuronxcc-2.21.33363.0+82129205/MODULE_c09b33112d0c8495e9d4+283df001/model.neff filter=lfs diff=lfs merge=lfs -text
15286
+ neuronxcc-2.21.33363.0+82129205/MODULE_47553fb37e8438019a32+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
15287
+ neuronxcc-2.21.33363.0+82129205/MODULE_686171fdb53c3e472a17+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
15288
+ neuronxcc-2.21.33363.0+82129205/MODULE_686171fdb53c3e472a17+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/qwen3/Qwen/Qwen3-8B/dcd74ada1295d4fd053d.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": 32,
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": 8,
68
+ "max_batch_size": 32,
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": "0.4.4.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": 8
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_47553fb37e8438019a32+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_47553fb37e8438019a32+24129607/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_47553fb37e8438019a32+24129607/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:59d420b00808bd3900438c1b9915a0c03ad9690b1dba9d4387b225ec59e3a84e
3
+ size 1142943
neuronxcc-2.21.33363.0+82129205/MODULE_47553fb37e8438019a32+24129607/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2481f50a72aca435458c56e767cfbe2cbd7d221c727d047f2a9accb427c52b6c
3
+ size 45599744
neuronxcc-2.21.33363.0+82129205/MODULE_686171fdb53c3e472a17+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_686171fdb53c3e472a17+a02c3a36/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_686171fdb53c3e472a17+a02c3a36/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e058f00a843fb502cfd8823cf1d6e15bba011d02540653dbd004243db994fa12
3
+ size 942443
neuronxcc-2.21.33363.0+82129205/MODULE_686171fdb53c3e472a17+a02c3a36/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cbe528be24b345922cfb0bef1af217c1737fa0c83c95dbc19adabe94f43e2564
3
+ size 7752704
neuronxcc-2.21.33363.0+82129205/MODULE_686171fdb53c3e472a17+a02c3a36/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:803368641ff0b47af809e04b00231e70a4b6a969efc97a3fbb5b3778710d9f19
3
+ size 7927272