Hugging Face's logo

aws-neuron
/
optimum-neuron-cache

Model card Files
xet
 Community
709
dacorvo HF Staff commited on
Commit
5f54f8c
·
verified ·
1 Parent(s): 34763a8

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/c033ab36fa35321d96cda66c61c1b610e3f28f44b24f28439782fbe605dc5a48/4d638a28b0b1de188a1e.json +96 -0
  3. neuronxcc-2.26.6360.0+6f180f47/MODULE_7a7387e6587d9cfa4f88+12dd9119/compile_flags.json +1 -0
  4. neuronxcc-2.26.6360.0+6f180f47/MODULE_7a7387e6587d9cfa4f88+12dd9119/model.done +0 -0
  5. neuronxcc-2.26.6360.0+6f180f47/MODULE_7a7387e6587d9cfa4f88+12dd9119/model.hlo_module.pb +3 -0
  6. neuronxcc-2.26.6360.0+6f180f47/MODULE_7a7387e6587d9cfa4f88+12dd9119/model.neff +3 -0
  7. neuronxcc-2.26.6360.0+6f180f47/MODULE_b3595ff85c95e992739a+84337dd9/compile_flags.json +1 -0
  8. neuronxcc-2.26.6360.0+6f180f47/MODULE_b3595ff85c95e992739a+84337dd9/model.done +0 -0
  9. neuronxcc-2.26.6360.0+6f180f47/MODULE_b3595ff85c95e992739a+84337dd9/model.hlo_module.pb +3 -0
  10. neuronxcc-2.26.6360.0+6f180f47/MODULE_b3595ff85c95e992739a+84337dd9/model.neff +3 -0
  11. neuronxcc-2.26.6360.0+6f180f47/MODULE_b3595ff85c95e992739a+84337dd9/wrapped_neff.hlo +3 -0
.gitattributes CHANGED
@@ -17846,3 +17846,6 @@ neuronxcc-2.26.6360.0+6f180f47/MODULE_a5bb4e738c72ff4cc969+84337dd9/wrapped_neff
17846
  neuronxcc-2.26.6360.0+6f180f47/MODULE_b755b28b014b3a94da4b+84337dd9/model.neff filter=lfs diff=lfs merge=lfs -text
17847
  neuronxcc-2.26.6360.0+6f180f47/MODULE_b755b28b014b3a94da4b+84337dd9/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
17848
  neuronxcc-2.26.6360.0+6f180f47/MODULE_dd0b522e5e0219bff205+12dd9119/model.neff filter=lfs diff=lfs merge=lfs -text
 
 
 
 
17846
  neuronxcc-2.26.6360.0+6f180f47/MODULE_b755b28b014b3a94da4b+84337dd9/model.neff filter=lfs diff=lfs merge=lfs -text
17847
  neuronxcc-2.26.6360.0+6f180f47/MODULE_b755b28b014b3a94da4b+84337dd9/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
17848
  neuronxcc-2.26.6360.0+6f180f47/MODULE_dd0b522e5e0219bff205+12dd9119/model.neff filter=lfs diff=lfs merge=lfs -text
17849
+ neuronxcc-2.26.6360.0+6f180f47/MODULE_7a7387e6587d9cfa4f88+12dd9119/model.neff filter=lfs diff=lfs merge=lfs -text
17850
+ neuronxcc-2.26.6360.0+6f180f47/MODULE_b3595ff85c95e992739a+84337dd9/model.neff filter=lfs diff=lfs merge=lfs -text
17851
+ neuronxcc-2.26.6360.0+6f180f47/MODULE_b3595ff85c95e992739a+84337dd9/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.26.6360.0+6f180f47/0_REGISTRY/0.4.7.dev0/c033ab36fa35321d96cda66c61c1b610e3f28f44b24f28439782fbe605dc5a48/4d638a28b0b1de188a1e.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": 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": 8,
68
+ "max_batch_size": 1,
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": true,
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": 8
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_7a7387e6587d9cfa4f88+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_7a7387e6587d9cfa4f88+12dd9119/model.done ADDED
File without changes
neuronxcc-2.26.6360.0+6f180f47/MODULE_7a7387e6587d9cfa4f88+12dd9119/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:baba443aef0a55b46a595b3d727c6484ba85c71deed046b1020740461e5b4be3
3
+ size 4860864
neuronxcc-2.26.6360.0+6f180f47/MODULE_7a7387e6587d9cfa4f88+12dd9119/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0818825c1886205fbea211ac74492c6c2be711c9378a87f212c64a0115196819
3
+ size 6759424
neuronxcc-2.26.6360.0+6f180f47/MODULE_b3595ff85c95e992739a+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_b3595ff85c95e992739a+84337dd9/model.done ADDED
File without changes
neuronxcc-2.26.6360.0+6f180f47/MODULE_b3595ff85c95e992739a+84337dd9/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4b065e10171bc764bb54e9d181a1900fc2e49d723e9f7d7639522d259170819e
3
+ size 3695290
neuronxcc-2.26.6360.0+6f180f47/MODULE_b3595ff85c95e992739a+84337dd9/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ec9997eab6cd9d8d3ba48521d5774d7fd13e7d2d5002a48758d17953fa72da81
3
+ size 3052544
neuronxcc-2.26.6360.0+6f180f47/MODULE_b3595ff85c95e992739a+84337dd9/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:33de56e4f8cafee4bb25c94e06ab92fdcef3082c347b9989845203e9d8907f01
3
+ size 3227112