Hugging Face's logo

aws-neuron
/
optimum-neuron-cache

Model card Files
xet
 Community
709
dacorvo HF Staff commited on
Commit
5eeaa45
·
verified ·
1 Parent(s): 94d7dcd

Synchronizing local compiler cache.

Browse files
Files changed (8) hide show
  1. .gitattributes +2 -0
  2. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/21e221cd6930845b922595ba5c04130e00e49d73db5ce2ab5a5318379ba7b958/800dd7b146b285cae4a0.json +99 -0
  3. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/qwen3/Qwen/Qwen3-14B/800dd7b146b285cae4a0.json +99 -0
  4. neuronxcc-2.21.33363.0+82129205/MODULE_aecd78e409b708d74c95+a02c3a36/compile_flags.json +1 -0
  5. neuronxcc-2.21.33363.0+82129205/MODULE_aecd78e409b708d74c95+a02c3a36/model.done +0 -0
  6. neuronxcc-2.21.33363.0+82129205/MODULE_aecd78e409b708d74c95+a02c3a36/model.hlo_module.pb +3 -0
  7. neuronxcc-2.21.33363.0+82129205/MODULE_aecd78e409b708d74c95+a02c3a36/model.neff +3 -0
  8. neuronxcc-2.21.33363.0+82129205/MODULE_aecd78e409b708d74c95+a02c3a36/wrapped_neff.hlo +3 -0
.gitattributes CHANGED
@@ -16844,3 +16844,5 @@ neuronxcc-2.21.33363.0+82129205/MODULE_9bad88aca4ded96246a4+a02c3a36/model.neff
16844
  neuronxcc-2.21.33363.0+82129205/MODULE_9bad88aca4ded96246a4+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
16845
  neuronxcc-2.21.33363.0+82129205/MODULE_2d26ade7ef0d76088638+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
16846
  neuronxcc-2.21.33363.0+82129205/MODULE_2d26ade7ef0d76088638+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
 
 
 
16844
  neuronxcc-2.21.33363.0+82129205/MODULE_9bad88aca4ded96246a4+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
16845
  neuronxcc-2.21.33363.0+82129205/MODULE_2d26ade7ef0d76088638+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
16846
  neuronxcc-2.21.33363.0+82129205/MODULE_2d26ade7ef0d76088638+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
16847
+ neuronxcc-2.21.33363.0+82129205/MODULE_aecd78e409b708d74c95+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
16848
+ neuronxcc-2.21.33363.0+82129205/MODULE_aecd78e409b708d74c95+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/21e221cd6930845b922595ba5c04130e00e49d73db5ce2ab5a5318379ba7b958/800dd7b146b285cae4a0.json ADDED
@@ -0,0 +1,99 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": 1,
64
+ "capacity_factor": null,
65
+ "checkpoint_id": "Qwen/Qwen3-14B",
66
+ "checkpoint_revision": "40c069824f4251a91eefaf281ebe4c544efd3e18",
67
+ "continuous_batching": false,
68
+ "ep_degree": 1,
69
+ "fused_qkv": true,
70
+ "glu_mlp": true,
71
+ "local_ranks_size": 8,
72
+ "max_batch_size": 1,
73
+ "max_context_length": 4096,
74
+ "max_topk": 256,
75
+ "n_active_tokens": 4096,
76
+ "neuronxcc_version": "2.21.33363.0+82129205",
77
+ "on_device_sampling": true,
78
+ "optimum_neuron_version": "0.4.6.dev2",
79
+ "output_logits": false,
80
+ "pp_degree": 1,
81
+ "sequence_length": 4096,
82
+ "speculation_length": 0,
83
+ "start_rank_id": 0,
84
+ "target": "trn1",
85
+ "torch_dtype": "bfloat16",
86
+ "tp_degree": 8
87
+ },
88
+ "num_attention_heads": 40,
89
+ "num_hidden_layers": 40,
90
+ "num_key_value_heads": 8,
91
+ "rms_norm_eps": 1e-06,
92
+ "rope_scaling": null,
93
+ "rope_theta": 1000000,
94
+ "sliding_window": null,
95
+ "tie_word_embeddings": false,
96
+ "use_cache": true,
97
+ "use_sliding_window": false,
98
+ "vocab_size": 151936
99
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/qwen3/Qwen/Qwen3-14B/800dd7b146b285cae4a0.json ADDED
@@ -0,0 +1,99 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": 1,
64
+ "capacity_factor": null,
65
+ "checkpoint_id": "Qwen/Qwen3-14B",
66
+ "checkpoint_revision": "40c069824f4251a91eefaf281ebe4c544efd3e18",
67
+ "continuous_batching": false,
68
+ "ep_degree": 1,
69
+ "fused_qkv": true,
70
+ "glu_mlp": true,
71
+ "local_ranks_size": 8,
72
+ "max_batch_size": 1,
73
+ "max_context_length": 4096,
74
+ "max_topk": 256,
75
+ "n_active_tokens": 4096,
76
+ "neuronxcc_version": "2.21.33363.0+82129205",
77
+ "on_device_sampling": true,
78
+ "optimum_neuron_version": "0.4.6.dev2",
79
+ "output_logits": false,
80
+ "pp_degree": 1,
81
+ "sequence_length": 4096,
82
+ "speculation_length": 0,
83
+ "start_rank_id": 0,
84
+ "target": "trn1",
85
+ "torch_dtype": "bfloat16",
86
+ "tp_degree": 8
87
+ },
88
+ "num_attention_heads": 40,
89
+ "num_hidden_layers": 40,
90
+ "num_key_value_heads": 8,
91
+ "rms_norm_eps": 1e-06,
92
+ "rope_scaling": null,
93
+ "rope_theta": 1000000,
94
+ "sliding_window": null,
95
+ "tie_word_embeddings": false,
96
+ "use_cache": true,
97
+ "use_sliding_window": false,
98
+ "vocab_size": 151936
99
+ }
neuronxcc-2.21.33363.0+82129205/MODULE_aecd78e409b708d74c95+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_aecd78e409b708d74c95+a02c3a36/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_aecd78e409b708d74c95+a02c3a36/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:48b5b4d107ae14b671c0ba544300cedf905f8377aaeccda80dbd47a834c22b63
3
+ size 1083901
neuronxcc-2.21.33363.0+82129205/MODULE_aecd78e409b708d74c95+a02c3a36/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e6b7d83f35a0ee0a8413039a057ac1a5b8c3587828f59fc658b8d1979458fa43
3
+ size 5243904
neuronxcc-2.21.33363.0+82129205/MODULE_aecd78e409b708d74c95+a02c3a36/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9081e305870507213720b115f799e4c7c3c2141da81c6669e0e75d3defc478ee
3
+ size 5438416