Hugging Face's logo

aws-neuron
/
optimum-neuron-cache

Model card Files
xet
Community
647
dacorvo HF Staff commited on
Commit
5be11ff
·
verified ·
1 Parent(s): 613236d

Synchronizing local compiler cache.

Browse files
Files changed (11) hide show
  1. .gitattributes +2 -0
  2. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.5.dev2/e0b6d1e2424243dcd9ff1755e02969dcc312d14df531d876c5c2892f285b2863/9054b94f39e5c374b6b8.json +95 -0
  3. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.5.dev2/e0b6d1e2424243dcd9ff1755e02969dcc312d14df531d876c5c2892f285b2863/d06a9ca97bbded610b72.json +95 -0
  4. neuronxcc-2.21.33363.0+82129205/MODULE_1efaefc590ce7ee07e97+fb4cc044/compile_flags.json +1 -0
  5. neuronxcc-2.21.33363.0+82129205/MODULE_1efaefc590ce7ee07e97+fb4cc044/model.done +0 -0
  6. neuronxcc-2.21.33363.0+82129205/MODULE_1efaefc590ce7ee07e97+fb4cc044/model.hlo_module.pb +3 -0
  7. neuronxcc-2.21.33363.0+82129205/MODULE_1efaefc590ce7ee07e97+fb4cc044/model.neff +3 -0
  8. neuronxcc-2.21.33363.0+82129205/MODULE_5fdd7ad15e00da4fcd3e+fb4cc044/compile_flags.json +1 -0
  9. neuronxcc-2.21.33363.0+82129205/MODULE_5fdd7ad15e00da4fcd3e+fb4cc044/model.done +0 -0
  10. neuronxcc-2.21.33363.0+82129205/MODULE_5fdd7ad15e00da4fcd3e+fb4cc044/model.hlo_module.pb +3 -0
  11. neuronxcc-2.21.33363.0+82129205/MODULE_5fdd7ad15e00da4fcd3e+fb4cc044/model.neff +3 -0
.gitattributes CHANGED
@@ -16136,3 +16136,5 @@ neuronxcc-2.21.33363.0+82129205/MODULE_dd01d7a2913cbd8c1ba1+24129607/model.neff
16136
  neuronxcc-2.21.33363.0+82129205/MODULE_9c01f09fb84f2b339fc6+fb4cc044/model.neff filter=lfs diff=lfs merge=lfs -text
16137
  neuronxcc-2.21.33363.0+82129205/MODULE_14268084bdb93e2af8ea+fb4cc044/model.neff filter=lfs diff=lfs merge=lfs -text
16138
  neuronxcc-2.21.33363.0+82129205/MODULE_93b5314f91b367ce312a+fb4cc044/model.neff filter=lfs diff=lfs merge=lfs -text
 
 
 
16136
  neuronxcc-2.21.33363.0+82129205/MODULE_9c01f09fb84f2b339fc6+fb4cc044/model.neff filter=lfs diff=lfs merge=lfs -text
16137
  neuronxcc-2.21.33363.0+82129205/MODULE_14268084bdb93e2af8ea+fb4cc044/model.neff filter=lfs diff=lfs merge=lfs -text
16138
  neuronxcc-2.21.33363.0+82129205/MODULE_93b5314f91b367ce312a+fb4cc044/model.neff filter=lfs diff=lfs merge=lfs -text
16139
+ neuronxcc-2.21.33363.0+82129205/MODULE_1efaefc590ce7ee07e97+fb4cc044/model.neff filter=lfs diff=lfs merge=lfs -text
16140
+ neuronxcc-2.21.33363.0+82129205/MODULE_5fdd7ad15e00da4fcd3e+fb4cc044/model.neff filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.5.dev2/e0b6d1e2424243dcd9ff1755e02969dcc312d14df531d876c5c2892f285b2863/9054b94f39e5c374b6b8.json ADDED
@@ -0,0 +1,95 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "Qwen/Qwen3-Embedding-4B",
4
+ "_task": "feature-extraction",
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": 2560,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 9728,
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-Embedding-4B",
62
+ "checkpoint_revision": "5cf2132abc99cad020ac570b19d031efec650f2b",
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": 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": false,
74
+ "optimum_neuron_version": "0.4.5.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": true,
92
+ "use_cache": true,
93
+ "use_sliding_window": false,
94
+ "vocab_size": 151665
95
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.5.dev2/e0b6d1e2424243dcd9ff1755e02969dcc312d14df531d876c5c2892f285b2863/d06a9ca97bbded610b72.json ADDED
@@ -0,0 +1,95 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "Qwen/Qwen3-Embedding-4B",
4
+ "_task": "feature-extraction",
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": 2560,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 9728,
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": 4,
60
+ "capacity_factor": null,
61
+ "checkpoint_id": "Qwen/Qwen3-Embedding-4B",
62
+ "checkpoint_revision": "5cf2132abc99cad020ac570b19d031efec650f2b",
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": 4,
69
+ "max_context_length": 1024,
70
+ "max_topk": 256,
71
+ "n_active_tokens": 1024,
72
+ "neuronxcc_version": "2.21.33363.0+82129205",
73
+ "on_device_sampling": false,
74
+ "optimum_neuron_version": "0.4.5.dev2",
75
+ "output_logits": false,
76
+ "pp_degree": 1,
77
+ "sequence_length": 1024,
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": true,
92
+ "use_cache": true,
93
+ "use_sliding_window": false,
94
+ "vocab_size": 151665
95
+ }
neuronxcc-2.21.33363.0+82129205/MODULE_1efaefc590ce7ee07e97+fb4cc044/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/encoding/_tp0_bk0/log-neuron-cc.txt"]
neuronxcc-2.21.33363.0+82129205/MODULE_1efaefc590ce7ee07e97+fb4cc044/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_1efaefc590ce7ee07e97+fb4cc044/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0a917908cdedb3336190bdbfa5dc4c0b006052c9672b3c51ee0aa7c7a9e1e439
3
+ size 628622
neuronxcc-2.21.33363.0+82129205/MODULE_1efaefc590ce7ee07e97+fb4cc044/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5cdc725c9067ddd09df8d0768aa1744294c6717afb9cb332d40ec2fa3d13f86d
3
+ size 2888704
neuronxcc-2.21.33363.0+82129205/MODULE_5fdd7ad15e00da4fcd3e+fb4cc044/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/encoding/_tp0_bk0/log-neuron-cc.txt"]
neuronxcc-2.21.33363.0+82129205/MODULE_5fdd7ad15e00da4fcd3e+fb4cc044/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_5fdd7ad15e00da4fcd3e+fb4cc044/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:232afcad03e2df34f698ed0701ed5e2072121afca9edb009d66a9aec12f1a379
3
+ size 840640
neuronxcc-2.21.33363.0+82129205/MODULE_5fdd7ad15e00da4fcd3e+fb4cc044/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:296c2002b93c609d4a68eb22ac8c2afc5385391ddafe81315c177874ccdea5df
3
+ size 45046784