Hugging Face's logo

aws-neuron
/
optimum-neuron-cache

Model card Files
xet
Community
611
dacorvo HF Staff commited on
Commit
a10011b
·
verified ·
1 Parent(s): d560d31

Synchronizing local compiler cache.

Browse files
Files changed (11) hide show
  1. .gitattributes +3 -0
  2. neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.3.1.dev7/qwen2/Qwen/Qwen2.5-14B/d520f13b01562c579a72.json +106 -0
  3. neuronxcc-2.21.18209.0+043b1bf7/MODULE_04147eff5c30f01fed2f+a9d440f5/compile_flags.json +1 -0
  4. neuronxcc-2.21.18209.0+043b1bf7/MODULE_04147eff5c30f01fed2f+a9d440f5/model.done +0 -0
  5. neuronxcc-2.21.18209.0+043b1bf7/MODULE_04147eff5c30f01fed2f+a9d440f5/model.hlo_module.pb +3 -0
  6. neuronxcc-2.21.18209.0+043b1bf7/MODULE_04147eff5c30f01fed2f+a9d440f5/model.neff +3 -0
  7. neuronxcc-2.21.18209.0+043b1bf7/MODULE_04147eff5c30f01fed2f+a9d440f5/wrapped_neff.hlo +3 -0
  8. neuronxcc-2.21.18209.0+043b1bf7/MODULE_888475437bc2fdc16e7f+ed72d204/compile_flags.json +1 -0
  9. neuronxcc-2.21.18209.0+043b1bf7/MODULE_888475437bc2fdc16e7f+ed72d204/model.done +0 -0
  10. neuronxcc-2.21.18209.0+043b1bf7/MODULE_888475437bc2fdc16e7f+ed72d204/model.hlo_module.pb +3 -0
  11. neuronxcc-2.21.18209.0+043b1bf7/MODULE_888475437bc2fdc16e7f+ed72d204/model.neff +3 -0
.gitattributes CHANGED
@@ -11428,3 +11428,6 @@ neuronxcc-2.21.18209.0+043b1bf7/MODULE_771fa5f7dd63d5f785da+a9d440f5/wrapped_nef
11428
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_047d941ae224ce0be63f+ed72d204/model.neff filter=lfs diff=lfs merge=lfs -text
11429
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_fed3d6a36a02419a6aed+a9d440f5/model.neff filter=lfs diff=lfs merge=lfs -text
11430
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_fed3d6a36a02419a6aed+a9d440f5/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
 
 
 
 
11428
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_047d941ae224ce0be63f+ed72d204/model.neff filter=lfs diff=lfs merge=lfs -text
11429
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_fed3d6a36a02419a6aed+a9d440f5/model.neff filter=lfs diff=lfs merge=lfs -text
11430
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_fed3d6a36a02419a6aed+a9d440f5/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
11431
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_04147eff5c30f01fed2f+a9d440f5/model.neff filter=lfs diff=lfs merge=lfs -text
11432
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_04147eff5c30f01fed2f+a9d440f5/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
11433
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_888475437bc2fdc16e7f+ed72d204/model.neff filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.3.1.dev7/qwen2/Qwen/Qwen2.5-14B/d520f13b01562c579a72.json ADDED
@@ -0,0 +1,106 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "Qwen/Qwen2.5-14B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "Qwen2ForCausalLM"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "hidden_act": "silu",
10
+ "hidden_size": 5120,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 13824,
13
+ "layer_types": [
14
+ "full_attention",
15
+ "full_attention",
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
+ "full_attention",
53
+ "full_attention",
54
+ "full_attention",
55
+ "full_attention",
56
+ "full_attention",
57
+ "full_attention",
58
+ "full_attention",
59
+ "full_attention",
60
+ "full_attention",
61
+ "full_attention"
62
+ ],
63
+ "max_position_embeddings": 131072,
64
+ "max_window_layers": 48,
65
+ "model_type": "qwen2",
66
+ "neuron": {
67
+ "_serialized_key": "NxDNeuronConfig",
68
+ "batch_size": 1,
69
+ "capacity_factor": null,
70
+ "checkpoint_id": "Qwen/Qwen2.5-14B",
71
+ "checkpoint_revision": "97e1e76335b7017d8f67c08a19d103c0504298c9",
72
+ "continuous_batching": false,
73
+ "enable_bucketing": false,
74
+ "ep_degree": 1,
75
+ "fused_qkv": false,
76
+ "glu_mlp": true,
77
+ "local_ranks_size": 8,
78
+ "logical_nc_config": 1,
79
+ "max_batch_size": 1,
80
+ "max_context_length": 4096,
81
+ "max_topk": 256,
82
+ "n_active_tokens": 4096,
83
+ "neuronxcc_version": "2.21.18209.0+043b1bf7",
84
+ "on_device_sampling": true,
85
+ "optimum_neuron_version": "0.3.1.dev7",
86
+ "output_logits": false,
87
+ "pp_degree": 1,
88
+ "sequence_length": 4096,
89
+ "speculation_length": 0,
90
+ "start_rank_id": 0,
91
+ "target": null,
92
+ "torch_dtype": "bfloat16",
93
+ "tp_degree": 8
94
+ },
95
+ "num_attention_heads": 40,
96
+ "num_hidden_layers": 48,
97
+ "num_key_value_heads": 8,
98
+ "rms_norm_eps": 1e-05,
99
+ "rope_scaling": null,
100
+ "rope_theta": 1000000.0,
101
+ "sliding_window": null,
102
+ "tie_word_embeddings": false,
103
+ "use_cache": true,
104
+ "use_sliding_window": false,
105
+ "vocab_size": 152064
106
+ }
neuronxcc-2.21.18209.0+043b1bf7/MODULE_04147eff5c30f01fed2f+a9d440f5/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_model/_tp0_bk0/log-neuron-cc.txt", "--enable-internal-neff-wrapper"]
neuronxcc-2.21.18209.0+043b1bf7/MODULE_04147eff5c30f01fed2f+a9d440f5/model.done ADDED
File without changes
neuronxcc-2.21.18209.0+043b1bf7/MODULE_04147eff5c30f01fed2f+a9d440f5/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cae54528756108bfa6aae8abe44e09a61866b075ed0c703e3848aa0907642de7
3
+ size 1233606
neuronxcc-2.21.18209.0+043b1bf7/MODULE_04147eff5c30f01fed2f+a9d440f5/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1a288ed7c9aa772539efc06e93873c0627bf637558b9f1f1c367f3f9ed3d3aed
3
+ size 5264384
neuronxcc-2.21.18209.0+043b1bf7/MODULE_04147eff5c30f01fed2f+a9d440f5/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b9519086e450477e006fc902910f32152484717128ab6b6c9d2da1be69096bbd
3
+ size 5580525
neuronxcc-2.21.18209.0+043b1bf7/MODULE_888475437bc2fdc16e7f+ed72d204/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_model/_tp0_bk0/log-neuron-cc.txt"]
neuronxcc-2.21.18209.0+043b1bf7/MODULE_888475437bc2fdc16e7f+ed72d204/model.done ADDED
File without changes
neuronxcc-2.21.18209.0+043b1bf7/MODULE_888475437bc2fdc16e7f+ed72d204/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5fd4393547255665daa4f7cd0271ef574f3ef0e1cee19f6bc7699fb14fb90924
3
+ size 1426478
neuronxcc-2.21.18209.0+043b1bf7/MODULE_888475437bc2fdc16e7f+ed72d204/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ab65af93c77b75e298587c7220e3a904b888830b960f3a6588737b4851674f1f
3
+ size 3124224