Hugging Face's logo

aws-neuron
/
optimum-neuron-cache

Model card Files
xet
Community
657
dacorvo HF Staff commited on
Commit
3e5c92e
·
verified ·
1 Parent(s): fca9699

Synchronizing local compiler cache.

Browse files
Files changed (11) hide show
  1. .gitattributes +3 -0
  2. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/smollm3/HuggingFaceTB/SmolLM3-3B/b04222fd18da88122b41.json +134 -0
  3. neuronxcc-2.21.33363.0+82129205/MODULE_395ce70a9e28a8354d28+24129607/compile_flags.json +1 -0
  4. neuronxcc-2.21.33363.0+82129205/MODULE_395ce70a9e28a8354d28+24129607/model.done +0 -0
  5. neuronxcc-2.21.33363.0+82129205/MODULE_395ce70a9e28a8354d28+24129607/model.hlo_module.pb +3 -0
  6. neuronxcc-2.21.33363.0+82129205/MODULE_395ce70a9e28a8354d28+24129607/model.neff +3 -0
  7. neuronxcc-2.21.33363.0+82129205/MODULE_87d78206aae887c1e84a+a02c3a36/compile_flags.json +1 -0
  8. neuronxcc-2.21.33363.0+82129205/MODULE_87d78206aae887c1e84a+a02c3a36/model.done +0 -0
  9. neuronxcc-2.21.33363.0+82129205/MODULE_87d78206aae887c1e84a+a02c3a36/model.hlo_module.pb +3 -0
  10. neuronxcc-2.21.33363.0+82129205/MODULE_87d78206aae887c1e84a+a02c3a36/model.neff +3 -0
  11. neuronxcc-2.21.33363.0+82129205/MODULE_87d78206aae887c1e84a+a02c3a36/wrapped_neff.hlo +3 -0
.gitattributes CHANGED
@@ -15208,3 +15208,6 @@ neuronxcc-2.21.33363.0+82129205/MODULE_b38fcd5e784c9adb7a8d+80d05c3f/wrapped_nef
15208
  neuronxcc-2.21.33363.0+82129205/MODULE_1f4353eeadb20cf74586+a32116a7/model.neff filter=lfs diff=lfs merge=lfs -text
15209
  neuronxcc-2.21.33363.0+82129205/MODULE_b9eb314d0be80ce19cd0+ac10809c/model.neff filter=lfs diff=lfs merge=lfs -text
15210
  neuronxcc-2.21.33363.0+82129205/MODULE_b9eb314d0be80ce19cd0+ac10809c/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
 
 
 
 
15208
  neuronxcc-2.21.33363.0+82129205/MODULE_1f4353eeadb20cf74586+a32116a7/model.neff filter=lfs diff=lfs merge=lfs -text
15209
  neuronxcc-2.21.33363.0+82129205/MODULE_b9eb314d0be80ce19cd0+ac10809c/model.neff filter=lfs diff=lfs merge=lfs -text
15210
  neuronxcc-2.21.33363.0+82129205/MODULE_b9eb314d0be80ce19cd0+ac10809c/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
15211
+ neuronxcc-2.21.33363.0+82129205/MODULE_395ce70a9e28a8354d28+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
15212
+ neuronxcc-2.21.33363.0+82129205/MODULE_87d78206aae887c1e84a+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
15213
+ neuronxcc-2.21.33363.0+82129205/MODULE_87d78206aae887c1e84a+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/smollm3/HuggingFaceTB/SmolLM3-3B/b04222fd18da88122b41.json ADDED
@@ -0,0 +1,134 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "HuggingFaceTB/SmolLM3-3B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "SmolLM3ForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "dtype": "bfloat16",
11
+ "hidden_act": "silu",
12
+ "hidden_size": 2048,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 11008,
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": 65536,
54
+ "max_window_layers": 28,
55
+ "mlp_bias": false,
56
+ "model_type": "smollm3",
57
+ "neuron": {
58
+ "_serialized_key": "NxDNeuronConfig",
59
+ "batch_size": 16,
60
+ "capacity_factor": null,
61
+ "checkpoint_id": "HuggingFaceTB/SmolLM3-3B",
62
+ "checkpoint_revision": "a07cc9a04f16550a088caea529712d1d335b0ac1",
63
+ "continuous_batching": true,
64
+ "ep_degree": 1,
65
+ "fused_qkv": true,
66
+ "glu_mlp": true,
67
+ "local_ranks_size": 2,
68
+ "max_batch_size": 16,
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": true,
74
+ "optimum_neuron_version": "0.4.4.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": 2
83
+ },
84
+ "no_rope_layer_interval": 4,
85
+ "no_rope_layers": [
86
+ 1,
87
+ 1,
88
+ 1,
89
+ 0,
90
+ 1,
91
+ 1,
92
+ 1,
93
+ 0,
94
+ 1,
95
+ 1,
96
+ 1,
97
+ 0,
98
+ 1,
99
+ 1,
100
+ 1,
101
+ 0,
102
+ 1,
103
+ 1,
104
+ 1,
105
+ 0,
106
+ 1,
107
+ 1,
108
+ 1,
109
+ 0,
110
+ 1,
111
+ 1,
112
+ 1,
113
+ 0,
114
+ 1,
115
+ 1,
116
+ 1,
117
+ 0,
118
+ 1,
119
+ 1,
120
+ 1,
121
+ 0
122
+ ],
123
+ "num_attention_heads": 16,
124
+ "num_hidden_layers": 36,
125
+ "num_key_value_heads": 4,
126
+ "pretraining_tp": 2,
127
+ "rms_norm_eps": 1e-06,
128
+ "rope_scaling": null,
129
+ "rope_theta": 5000000.0,
130
+ "sliding_window": null,
131
+ "use_cache": false,
132
+ "use_sliding_window": false,
133
+ "vocab_size": 128256
134
+ }
neuronxcc-2.21.33363.0+82129205/MODULE_395ce70a9e28a8354d28+24129607/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/_tp0_bk0/log-neuron-cc.txt"]
neuronxcc-2.21.33363.0+82129205/MODULE_395ce70a9e28a8354d28+24129607/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_395ce70a9e28a8354d28+24129607/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:71bf649e3af9e699670bfd2e826a6c5e4898ec9fc3d03addb07ed63af206a6e5
3
+ size 1061093
neuronxcc-2.21.33363.0+82129205/MODULE_395ce70a9e28a8354d28+24129607/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ec620c5e98d11a5c3b80907e1750e9fe6aa3ad58ef142485ca2208ad46beebf6
3
+ size 9473024
neuronxcc-2.21.33363.0+82129205/MODULE_87d78206aae887c1e84a+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_87d78206aae887c1e84a+a02c3a36/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_87d78206aae887c1e84a+a02c3a36/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cb6384ebe6e3841b1f3de42fad99356a93cf0964931b6a1da961b923c7459500
3
+ size 872513
neuronxcc-2.21.33363.0+82129205/MODULE_87d78206aae887c1e84a+a02c3a36/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ef9ded7a6a03dde42d6205711e87edb694c05cdc823b165caae6254f12e93da6
3
+ size 9391104
neuronxcc-2.21.33363.0+82129205/MODULE_87d78206aae887c1e84a+a02c3a36/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:85a7226fcbbec796edcc1123f30b1255dbb0e1b381b46f51e7827d423447e46f
3
+ size 9557628