Hugging Face's logo

aws-neuron
/
optimum-neuron-cache

Model card Files
xet
 Community
709
dacorvo HF Staff commited on
Commit
af6324a
·
verified ·
1 Parent(s): 4fe9c7b

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.dev0/21e221cd6930845b922595ba5c04130e00e49d73db5ce2ab5a5318379ba7b958/789c0f651316ba8f535b.json +100 -0
  3. neuronxcc-2.26.6360.0+6f180f47/MODULE_636708c2f78aa8b33791+84337dd9/compile_flags.json +1 -0
  4. neuronxcc-2.26.6360.0+6f180f47/MODULE_636708c2f78aa8b33791+84337dd9/model.done +0 -0
  5. neuronxcc-2.26.6360.0+6f180f47/MODULE_636708c2f78aa8b33791+84337dd9/model.hlo_module.pb +3 -0
  6. neuronxcc-2.26.6360.0+6f180f47/MODULE_636708c2f78aa8b33791+84337dd9/model.neff +3 -0
  7. neuronxcc-2.26.6360.0+6f180f47/MODULE_636708c2f78aa8b33791+84337dd9/wrapped_neff.hlo +3 -0
  8. neuronxcc-2.26.6360.0+6f180f47/MODULE_de0eea6382a4a1d40dd2+12dd9119/compile_flags.json +1 -0
  9. neuronxcc-2.26.6360.0+6f180f47/MODULE_de0eea6382a4a1d40dd2+12dd9119/model.done +0 -0
  10. neuronxcc-2.26.6360.0+6f180f47/MODULE_de0eea6382a4a1d40dd2+12dd9119/model.hlo_module.pb +3 -0
  11. neuronxcc-2.26.6360.0+6f180f47/MODULE_de0eea6382a4a1d40dd2+12dd9119/model.neff +3 -0
.gitattributes CHANGED
@@ -17942,3 +17942,6 @@ neuronxcc-2.26.6360.0+6f180f47/MODULE_794cf90f1edd98ba7a65+2a395fce/model.neff f
17942
  neuronxcc-2.26.6360.0+6f180f47/MODULE_16b8232896c165ec0439+84337dd9/model.neff filter=lfs diff=lfs merge=lfs -text
17943
  neuronxcc-2.26.6360.0+6f180f47/MODULE_16b8232896c165ec0439+84337dd9/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
17944
  neuronxcc-2.26.6360.0+6f180f47/MODULE_989c98283f1ed0a0c911+12dd9119/model.neff filter=lfs diff=lfs merge=lfs -text
 
 
 
 
17942
  neuronxcc-2.26.6360.0+6f180f47/MODULE_16b8232896c165ec0439+84337dd9/model.neff filter=lfs diff=lfs merge=lfs -text
17943
  neuronxcc-2.26.6360.0+6f180f47/MODULE_16b8232896c165ec0439+84337dd9/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
17944
  neuronxcc-2.26.6360.0+6f180f47/MODULE_989c98283f1ed0a0c911+12dd9119/model.neff filter=lfs diff=lfs merge=lfs -text
17945
+ neuronxcc-2.26.6360.0+6f180f47/MODULE_636708c2f78aa8b33791+84337dd9/model.neff filter=lfs diff=lfs merge=lfs -text
17946
+ neuronxcc-2.26.6360.0+6f180f47/MODULE_636708c2f78aa8b33791+84337dd9/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
17947
+ neuronxcc-2.26.6360.0+6f180f47/MODULE_de0eea6382a4a1d40dd2+12dd9119/model.neff filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.26.6360.0+6f180f47/0_REGISTRY/0.4.7.dev0/21e221cd6930845b922595ba5c04130e00e49d73db5ce2ab5a5318379ba7b958/789c0f651316ba8f535b.json ADDED
@@ -0,0 +1,100 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "Qwen/Qwen3-14B",
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": 5120,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 17408,
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
+ "full_attention",
54
+ "full_attention",
55
+ "full_attention",
56
+ "full_attention"
57
+ ],
58
+ "max_position_embeddings": 40960,
59
+ "max_window_layers": 40,
60
+ "model_type": "qwen3",
61
+ "neuron": {
62
+ "_serialized_key": "NxDNeuronConfig",
63
+ "batch_size": 16,
64
+ "capacity_factor": null,
65
+ "checkpoint_id": "Qwen/Qwen3-14B",
66
+ "checkpoint_revision": "40c069824f4251a91eefaf281ebe4c544efd3e18",
67
+ "continuous_batching": true,
68
+ "ep_degree": 1,
69
+ "fused_qkv": true,
70
+ "glu_mlp": true,
71
+ "local_ranks_size": 8,
72
+ "max_batch_size": 16,
73
+ "max_context_length": 4096,
74
+ "max_topk": 256,
75
+ "n_active_tokens": 4096,
76
+ "neuronxcc_version": "2.26.6360.0+6f180f47",
77
+ "on_device_sampling": true,
78
+ "optimum_neuron_version": "0.4.7.dev0",
79
+ "output_logits": false,
80
+ "pp_degree": 1,
81
+ "prefill_chunk_size": 1024,
82
+ "sequence_length": 4096,
83
+ "speculation_length": 0,
84
+ "start_rank_id": 0,
85
+ "target": "trn1",
86
+ "torch_dtype": "bfloat16",
87
+ "tp_degree": 8
88
+ },
89
+ "num_attention_heads": 40,
90
+ "num_hidden_layers": 40,
91
+ "num_key_value_heads": 8,
92
+ "rms_norm_eps": 1e-06,
93
+ "rope_scaling": null,
94
+ "rope_theta": 1000000,
95
+ "sliding_window": null,
96
+ "tie_word_embeddings": false,
97
+ "use_cache": true,
98
+ "use_sliding_window": false,
99
+ "vocab_size": 151936
100
+ }
neuronxcc-2.26.6360.0+6f180f47/MODULE_636708c2f78aa8b33791+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_636708c2f78aa8b33791+84337dd9/model.done ADDED
File without changes
neuronxcc-2.26.6360.0+6f180f47/MODULE_636708c2f78aa8b33791+84337dd9/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ac173e1acc576a528fa7c88ced5c8c450c4e781a7d890b2e08ae0aed55348afa
3
+ size 4098410
neuronxcc-2.26.6360.0+6f180f47/MODULE_636708c2f78aa8b33791+84337dd9/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2dd2ad610d4fbce99eaceb8e0cf57b3a019dec7c79f25b5a5450ebca147148d1
3
+ size 7762944
neuronxcc-2.26.6360.0+6f180f47/MODULE_636708c2f78aa8b33791+84337dd9/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:469a203bdec6034c62756c0432cf15fe7809e97070dbdcbd471f6b3ec423468c
3
+ size 7957456
neuronxcc-2.26.6360.0+6f180f47/MODULE_de0eea6382a4a1d40dd2+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_de0eea6382a4a1d40dd2+12dd9119/model.done ADDED
File without changes
neuronxcc-2.26.6360.0+6f180f47/MODULE_de0eea6382a4a1d40dd2+12dd9119/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d0c48f8118daf858a73772ff7b77186adadfe27163703fbb287854bf8dc85fbf
3
+ size 5697639
neuronxcc-2.26.6360.0+6f180f47/MODULE_de0eea6382a4a1d40dd2+12dd9119/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2b850473df0bae85a2b7c0edfd190ba6fe0678a06d7993682e03d7e649af396f
3
+ size 6964224