dacorvo HF Staff commited on
Commit
7c1754b
·
verified ·
1 Parent(s): 7f7eb31

Synchronizing local compiler cache.

Browse files
.gitattributes CHANGED
@@ -13709,3 +13709,6 @@ neuronxcc-2.21.33363.0+82129205/MODULE_92667f83a1c29e318728+a02c3a36/wrapped_nef
13709
  neuronxcc-2.21.33363.0+82129205/MODULE_8a8a113f1bcb55222884+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
13710
  neuronxcc-2.21.33363.0+82129205/MODULE_fcb3f3d12a7b675ef8cd+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
13711
  neuronxcc-2.21.33363.0+82129205/MODULE_fcb3f3d12a7b675ef8cd+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
 
 
 
 
13709
  neuronxcc-2.21.33363.0+82129205/MODULE_8a8a113f1bcb55222884+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
13710
  neuronxcc-2.21.33363.0+82129205/MODULE_fcb3f3d12a7b675ef8cd+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
13711
  neuronxcc-2.21.33363.0+82129205/MODULE_fcb3f3d12a7b675ef8cd+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
13712
+ neuronxcc-2.21.33363.0+82129205/MODULE_3dd7b4668d85604269ad+a32116a7/model.neff filter=lfs diff=lfs merge=lfs -text
13713
+ neuronxcc-2.21.33363.0+82129205/MODULE_bcf6c5216d6c969dab06+ac10809c/model.neff filter=lfs diff=lfs merge=lfs -text
13714
+ neuronxcc-2.21.33363.0+82129205/MODULE_bcf6c5216d6c969dab06+ac10809c/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.3.dev2/llama/meta-llama/Llama-3.3-70B-Instruct/f23ce03b8e36586784d0.json ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "meta-llama/Llama-3.3-70B-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": 128,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 8192,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 28672,
16
+ "max_position_embeddings": 131072,
17
+ "mlp_bias": false,
18
+ "model_type": "llama",
19
+ "neuron": {
20
+ "_serialized_key": "NxDNeuronConfig",
21
+ "batch_size": 64,
22
+ "capacity_factor": null,
23
+ "checkpoint_id": "meta-llama/Llama-3.3-70B-Instruct",
24
+ "checkpoint_revision": "6f6073b423013f6a7d4d9f39144961bfbfbc386b",
25
+ "continuous_batching": true,
26
+ "ep_degree": 1,
27
+ "fused_qkv": true,
28
+ "glu_mlp": true,
29
+ "local_ranks_size": 64,
30
+ "max_batch_size": 64,
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": "0.4.3.dev2",
37
+ "output_logits": false,
38
+ "pp_degree": 1,
39
+ "sequence_length": 4096,
40
+ "speculation_length": 0,
41
+ "start_rank_id": 0,
42
+ "target": "trn2",
43
+ "torch_dtype": "bfloat16",
44
+ "tp_degree": 64
45
+ },
46
+ "num_attention_heads": 64,
47
+ "num_hidden_layers": 80,
48
+ "num_key_value_heads": 8,
49
+ "pretraining_tp": 1,
50
+ "rms_norm_eps": 1e-05,
51
+ "rope_scaling": {
52
+ "factor": 8.0,
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": false,
60
+ "use_cache": true,
61
+ "vocab_size": 128256
62
+ }
neuronxcc-2.21.33363.0+82129205/MODULE_3dd7b4668d85604269ad+a32116a7/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn2", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=2", "--logfile=/tmp/nxd_model/context_encoding/_tp0_bk0/log-neuron-cc.txt"]
neuronxcc-2.21.33363.0+82129205/MODULE_3dd7b4668d85604269ad+a32116a7/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_3dd7b4668d85604269ad+a32116a7/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f01876cb17d5f0d8318cc36da62798d16e34fe44c5f92eadc344cd0389b0552c
3
+ size 3609894
neuronxcc-2.21.33363.0+82129205/MODULE_3dd7b4668d85604269ad+a32116a7/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4cc21a2d2100bf8e0a5d4fab97279153b33ba5816edaef985abf2251f007d7a5
3
+ size 1956864
neuronxcc-2.21.33363.0+82129205/MODULE_bcf6c5216d6c969dab06+ac10809c/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ ["--target=trn2", "--auto-cast=none", "--model-type=transformer", "--tensorizer-options=--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-strided-dma ", "-O2", "--lnc=2", "--logfile=/tmp/nxd_model/token_generation/_tp0_bk0/log-neuron-cc.txt", "--enable-internal-neff-wrapper"]
neuronxcc-2.21.33363.0+82129205/MODULE_bcf6c5216d6c969dab06+ac10809c/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_bcf6c5216d6c969dab06+ac10809c/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d6a9e2b5ae6413a44f93d75bfb3830483cbd6b0f40c3dda4f4617ef2603ceb84
3
+ size 6885961
neuronxcc-2.21.33363.0+82129205/MODULE_bcf6c5216d6c969dab06+ac10809c/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e6384d1bda2315ba85461813d04da10197b6fc522e70e39c546db67ac34952ba
3
+ size 27864064
neuronxcc-2.21.33363.0+82129205/MODULE_bcf6c5216d6c969dab06+ac10809c/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:814e8d813b08583fd1bb82a06116d620047a305d77fc5b4c0eb9dd0ca0498fde
3
+ size 28233183