Hugging Face's logo

aws-neuron
/
optimum-neuron-cache

Model card Files
xet
 Community
710
tengomucho HF Staff commited on
Commit
bcadd80
·
verified ·
1 Parent(s): d6248b2

Synchronizing local compiler cache.

Browse files
Files changed (12) hide show
  1. .gitattributes +3 -0
  2. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev0/llama/unsloth/Llama-3.2-1B-Instruct/282467d4772e5e1a2fb7.json +63 -0
  3. neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev0/qwen2/Qwen/Qwen2.5-0.5B/b0bb9b0f8b8aa747aa99.json +82 -0
  4. neuronxcc-2.21.33363.0+82129205/MODULE_9d35248a93d4142a3cf1+24129607/compile_flags.json +1 -0
  5. neuronxcc-2.21.33363.0+82129205/MODULE_9d35248a93d4142a3cf1+24129607/model.done +0 -0
  6. neuronxcc-2.21.33363.0+82129205/MODULE_9d35248a93d4142a3cf1+24129607/model.hlo_module.pb +3 -0
  7. neuronxcc-2.21.33363.0+82129205/MODULE_9d35248a93d4142a3cf1+24129607/model.neff +3 -0
  8. neuronxcc-2.21.33363.0+82129205/MODULE_c2b3afa51c57d431a332+a02c3a36/compile_flags.json +1 -0
  9. neuronxcc-2.21.33363.0+82129205/MODULE_c2b3afa51c57d431a332+a02c3a36/model.done +0 -0
  10. neuronxcc-2.21.33363.0+82129205/MODULE_c2b3afa51c57d431a332+a02c3a36/model.hlo_module.pb +3 -0
  11. neuronxcc-2.21.33363.0+82129205/MODULE_c2b3afa51c57d431a332+a02c3a36/model.neff +3 -0
  12. neuronxcc-2.21.33363.0+82129205/MODULE_c2b3afa51c57d431a332+a02c3a36/wrapped_neff.hlo +3 -0
.gitattributes CHANGED
@@ -14442,3 +14442,6 @@ neuronxcc-2.21.33363.0+82129205/MODULE_5d461ca056317ba5e9d4+b4a27fa5/model.neff
14442
  neuronxcc-2.21.33363.0+82129205/MODULE_5d461ca056317ba5e9d4+b4a27fa5/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
14443
  neuronxcc-2.21.33363.0+82129205/MODULE_602977aff4407b4089b8+b6eebe96/model.neff filter=lfs diff=lfs merge=lfs -text
14444
  neuronxcc-2.21.33363.0+82129205/MODULE_602977aff4407b4089b8+b6eebe96/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
 
 
 
 
14442
  neuronxcc-2.21.33363.0+82129205/MODULE_5d461ca056317ba5e9d4+b4a27fa5/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
14443
  neuronxcc-2.21.33363.0+82129205/MODULE_602977aff4407b4089b8+b6eebe96/model.neff filter=lfs diff=lfs merge=lfs -text
14444
  neuronxcc-2.21.33363.0+82129205/MODULE_602977aff4407b4089b8+b6eebe96/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
14445
+ neuronxcc-2.21.33363.0+82129205/MODULE_9d35248a93d4142a3cf1+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
14446
+ neuronxcc-2.21.33363.0+82129205/MODULE_c2b3afa51c57d431a332+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
14447
+ neuronxcc-2.21.33363.0+82129205/MODULE_c2b3afa51c57d431a332+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev0/llama/unsloth/Llama-3.2-1B-Instruct/282467d4772e5e1a2fb7.json ADDED
@@ -0,0 +1,63 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "unsloth/Llama-3.2-1B-Instruct",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "dtype": "bfloat16",
11
+ "head_dim": 64,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 2048,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 8192,
16
+ "max_position_embeddings": 131072,
17
+ "mlp_bias": false,
18
+ "model_type": "llama",
19
+ "neuron": {
20
+ "_serialized_key": "NxDNeuronConfig",
21
+ "batch_size": 4,
22
+ "capacity_factor": null,
23
+ "checkpoint_id": "unsloth/Llama-3.2-1B-Instruct",
24
+ "checkpoint_revision": "5a8abab4a5d6f164389b1079fb721cfab8d7126c",
25
+ "continuous_batching": true,
26
+ "ep_degree": 1,
27
+ "fused_qkv": true,
28
+ "glu_mlp": true,
29
+ "local_ranks_size": 2,
30
+ "max_batch_size": 4,
31
+ "max_context_length": 4096,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 4096,
34
+ "neuronxcc_version": "2.21.33363.0+82129205",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.4.dev0",
37
+ "output_logits": false,
38
+ "pp_degree": 1,
39
+ "sequence_length": 4096,
40
+ "speculation_length": 0,
41
+ "start_rank_id": 0,
42
+ "target": "trn1",
43
+ "torch_dtype": "bfloat16",
44
+ "tp_degree": 2
45
+ },
46
+ "num_attention_heads": 32,
47
+ "num_hidden_layers": 16,
48
+ "num_key_value_heads": 8,
49
+ "pretraining_tp": 1,
50
+ "rms_norm_eps": 1e-05,
51
+ "rope_scaling": {
52
+ "factor": 32.0,
53
+ "high_freq_factor": 4.0,
54
+ "low_freq_factor": 1.0,
55
+ "original_max_position_embeddings": 8192,
56
+ "rope_type": "llama3"
57
+ },
58
+ "rope_theta": 500000.0,
59
+ "tie_word_embeddings": true,
60
+ "unsloth_fixed": true,
61
+ "use_cache": true,
62
+ "vocab_size": 128256
63
+ }
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev0/qwen2/Qwen/Qwen2.5-0.5B/b0bb9b0f8b8aa747aa99.json ADDED
@@ -0,0 +1,82 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "Qwen/Qwen2.5-0.5B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "Qwen2ForCausalLM"
7
+ ],
8
+ "attention_dropout": 0.0,
9
+ "dtype": "bfloat16",
10
+ "hidden_act": "silu",
11
+ "hidden_size": 896,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 4864,
14
+ "layer_types": [
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
+ ],
40
+ "max_position_embeddings": 32768,
41
+ "max_window_layers": 24,
42
+ "model_type": "qwen2",
43
+ "neuron": {
44
+ "_serialized_key": "NxDNeuronConfig",
45
+ "batch_size": 4,
46
+ "capacity_factor": null,
47
+ "checkpoint_id": "Qwen/Qwen2.5-0.5B",
48
+ "checkpoint_revision": "060db6499f32faf8b98477b0a26969ef7d8b9987",
49
+ "continuous_batching": true,
50
+ "ep_degree": 1,
51
+ "fused_qkv": false,
52
+ "glu_mlp": true,
53
+ "local_ranks_size": 2,
54
+ "max_batch_size": 4,
55
+ "max_context_length": 4096,
56
+ "max_topk": 256,
57
+ "n_active_tokens": 4096,
58
+ "neuronxcc_version": "2.21.33363.0+82129205",
59
+ "on_device_sampling": false,
60
+ "optimum_neuron_version": "0.4.4.dev0",
61
+ "output_logits": false,
62
+ "pp_degree": 1,
63
+ "sequence_length": 4096,
64
+ "speculation_length": 0,
65
+ "start_rank_id": 0,
66
+ "target": "trn1",
67
+ "torch_dtype": "bfloat16",
68
+ "tp_degree": 2
69
+ },
70
+ "num_attention_heads": 14,
71
+ "num_hidden_layers": 24,
72
+ "num_key_value_heads": 2,
73
+ "rms_norm_eps": 1e-06,
74
+ "rope_scaling": null,
75
+ "rope_theta": 1000000.0,
76
+ "sliding_window": null,
77
+ "tie_word_embeddings": true,
78
+ "use_cache": true,
79
+ "use_mrope": false,
80
+ "use_sliding_window": false,
81
+ "vocab_size": 151936
82
+ }
neuronxcc-2.21.33363.0+82129205/MODULE_9d35248a93d4142a3cf1+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_9d35248a93d4142a3cf1+24129607/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_9d35248a93d4142a3cf1+24129607/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1ff789deef8176afeab268439fb0770b9b965fd178c19b0aead0f8be9a02e0b6
3
+ size 739558
neuronxcc-2.21.33363.0+82129205/MODULE_9d35248a93d4142a3cf1+24129607/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f516045352a9ee7a801257fbcdc2acf8ec33bfc172b6bfd19e961216ee11254d
3
+ size 26133504
neuronxcc-2.21.33363.0+82129205/MODULE_c2b3afa51c57d431a332+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_c2b3afa51c57d431a332+a02c3a36/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_c2b3afa51c57d431a332+a02c3a36/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2c2920e5fde77a80d2d66c9ba5addf46e53d032f533c7097cba34e31243f84ed
3
+ size 588406
neuronxcc-2.21.33363.0+82129205/MODULE_c2b3afa51c57d431a332+a02c3a36/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c8017f2ca25aa30c01e839314f958eedb8512833a8e66bfd9d075b9943b0980c
3
+ size 1926144
neuronxcc-2.21.33363.0+82129205/MODULE_c2b3afa51c57d431a332+a02c3a36/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:be407fd3aecea94529ce6cc1917f9b19dd88a49aa9a67cd85524e1b93a5b54cd
3
+ size 2082478