Hugging Face's logo

aws-neuron
/
optimum-neuron-cache

Model card Files
xet
 Community
694
dacorvo HF Staff commited on
Commit
2330e7d
·
verified ·
1 Parent(s): cf7d87c

Synchronizing local compiler cache.

Browse files
Files changed (11) hide show
  1. .gitattributes +3 -0
  2. neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.3.1.dev7/qwen3/Qwen/Qwen3-8B/6ab1e6cd7dcfab9e19fb.json +96 -0
  3. neuronxcc-2.21.18209.0+043b1bf7/MODULE_15e061e7741366c32100+a9d440f5/compile_flags.json +1 -0
  4. neuronxcc-2.21.18209.0+043b1bf7/MODULE_15e061e7741366c32100+a9d440f5/model.done +0 -0
  5. neuronxcc-2.21.18209.0+043b1bf7/MODULE_15e061e7741366c32100+a9d440f5/model.hlo_module.pb +3 -0
  6. neuronxcc-2.21.18209.0+043b1bf7/MODULE_15e061e7741366c32100+a9d440f5/model.neff +3 -0
  7. neuronxcc-2.21.18209.0+043b1bf7/MODULE_15e061e7741366c32100+a9d440f5/wrapped_neff.hlo +3 -0
  8. neuronxcc-2.21.18209.0+043b1bf7/MODULE_e198d0f9c0782f06ee42+ed72d204/compile_flags.json +1 -0
  9. neuronxcc-2.21.18209.0+043b1bf7/MODULE_e198d0f9c0782f06ee42+ed72d204/model.done +0 -0
  10. neuronxcc-2.21.18209.0+043b1bf7/MODULE_e198d0f9c0782f06ee42+ed72d204/model.hlo_module.pb +3 -0
  11. neuronxcc-2.21.18209.0+043b1bf7/MODULE_e198d0f9c0782f06ee42+ed72d204/model.neff +3 -0
.gitattributes CHANGED
@@ -11251,3 +11251,6 @@ neuronxcc-2.21.18209.0+043b1bf7/MODULE_de408d74aa957dd45204+a9d440f5/wrapped_nef
11251
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_222cde7d531b84898767+ed72d204/model.neff filter=lfs diff=lfs merge=lfs -text
11252
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_97e03eb20bdafac9522a+a9d440f5/model.neff filter=lfs diff=lfs merge=lfs -text
11253
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_97e03eb20bdafac9522a+a9d440f5/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
 
 
 
 
11251
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_222cde7d531b84898767+ed72d204/model.neff filter=lfs diff=lfs merge=lfs -text
11252
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_97e03eb20bdafac9522a+a9d440f5/model.neff filter=lfs diff=lfs merge=lfs -text
11253
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_97e03eb20bdafac9522a+a9d440f5/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
11254
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_15e061e7741366c32100+a9d440f5/model.neff filter=lfs diff=lfs merge=lfs -text
11255
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_15e061e7741366c32100+a9d440f5/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
11256
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_e198d0f9c0782f06ee42+ed72d204/model.neff filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.3.1.dev7/qwen3/Qwen/Qwen3-8B/6ab1e6cd7dcfab9e19fb.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
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 4096,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 12288,
15
+ "layer_types": [
16
+ "full_attention",
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
+ ],
53
+ "max_position_embeddings": 40960,
54
+ "max_window_layers": 36,
55
+ "model_type": "qwen3",
56
+ "neuron": {
57
+ "_serialized_key": "NxDNeuronConfig",
58
+ "batch_size": 8,
59
+ "capacity_factor": null,
60
+ "checkpoint_id": "Qwen/Qwen3-8B",
61
+ "checkpoint_revision": "b968826d9c46dd6066d109eabc6255188de91218",
62
+ "continuous_batching": true,
63
+ "enable_bucketing": false,
64
+ "ep_degree": 1,
65
+ "fused_qkv": true,
66
+ "glu_mlp": true,
67
+ "local_ranks_size": 2,
68
+ "logical_nc_config": 1,
69
+ "max_batch_size": 8,
70
+ "max_context_length": 4096,
71
+ "max_topk": 256,
72
+ "n_active_tokens": 4096,
73
+ "neuronxcc_version": "2.21.18209.0+043b1bf7",
74
+ "on_device_sampling": false,
75
+ "optimum_neuron_version": "0.3.1.dev7",
76
+ "output_logits": false,
77
+ "pp_degree": 1,
78
+ "sequence_length": 4096,
79
+ "speculation_length": 0,
80
+ "start_rank_id": 0,
81
+ "target": null,
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-2.21.18209.0+043b1bf7/MODULE_15e061e7741366c32100+a9d440f5/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_model/_tp0_bk0/log-neuron-cc.txt", "--enable-internal-neff-wrapper"]
neuronxcc-2.21.18209.0+043b1bf7/MODULE_15e061e7741366c32100+a9d440f5/model.done ADDED
File without changes
neuronxcc-2.21.18209.0+043b1bf7/MODULE_15e061e7741366c32100+a9d440f5/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cf70645d7046d072e5ae37760ecbb9ada5c566026789e025fa4248fded35ed98
3
+ size 905395
neuronxcc-2.21.18209.0+043b1bf7/MODULE_15e061e7741366c32100+a9d440f5/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0580ff7712a6e31baa94dd8f7ed40f2701b4873fc387c30653abf2a23d7d6e0e
3
+ size 12821504
neuronxcc-2.21.18209.0+043b1bf7/MODULE_15e061e7741366c32100+a9d440f5/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:474d5d3541a2d4da3509fdd1bd0f088a804e005469615d724bf023d6fe061034
3
+ size 12996604
neuronxcc-2.21.18209.0+043b1bf7/MODULE_e198d0f9c0782f06ee42+ed72d204/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_model/_tp0_bk0/log-neuron-cc.txt"]
neuronxcc-2.21.18209.0+043b1bf7/MODULE_e198d0f9c0782f06ee42+ed72d204/model.done ADDED
File without changes
neuronxcc-2.21.18209.0+043b1bf7/MODULE_e198d0f9c0782f06ee42+ed72d204/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1aa0650612194d065812e64ba0ac5539e7c4645eb95d4f0ee41d24d886a77bc6
3
+ size 1110837
neuronxcc-2.21.18209.0+043b1bf7/MODULE_e198d0f9c0782f06ee42+ed72d204/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2cb9d3c8df75a77d221e2597c3e7d9fcaea3427706eb0b3c700264e074adc736
3
+ size 6810624