dacorvo HF Staff commited on
Commit
f7da935
·
verified ·
1 Parent(s): e7fe27a

Synchronizing local compiler cache.

Browse files
.gitattributes CHANGED
@@ -18269,3 +18269,6 @@ neuronxcc-2.26.6360.0+6f180f47/MODULE_8d0d6be159b7f241660b+4a27301b/wrapped_neff
18269
  neuronxcc-2.26.6360.0+6f180f47/MODULE_f4c45f346593438822e0+4a27301b/model.neff filter=lfs diff=lfs merge=lfs -text
18270
  neuronxcc-2.26.6360.0+6f180f47/MODULE_f4c45f346593438822e0+4a27301b/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
18271
  neuronxcc-2.26.6360.0+6f180f47/MODULE_f7447941defc6f6c0342+665810b0/model.neff filter=lfs diff=lfs merge=lfs -text
 
 
 
 
18269
  neuronxcc-2.26.6360.0+6f180f47/MODULE_f4c45f346593438822e0+4a27301b/model.neff filter=lfs diff=lfs merge=lfs -text
18270
  neuronxcc-2.26.6360.0+6f180f47/MODULE_f4c45f346593438822e0+4a27301b/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
18271
  neuronxcc-2.26.6360.0+6f180f47/MODULE_f7447941defc6f6c0342+665810b0/model.neff filter=lfs diff=lfs merge=lfs -text
18272
+ neuronxcc-2.26.6360.0+6f180f47/MODULE_e090011e15ec2b236cbd+665810b0/model.neff filter=lfs diff=lfs merge=lfs -text
18273
+ neuronxcc-2.26.6360.0+6f180f47/MODULE_e70e014e5ec2a1b4879d+4a27301b/model.neff filter=lfs diff=lfs merge=lfs -text
18274
+ neuronxcc-2.26.6360.0+6f180f47/MODULE_e70e014e5ec2a1b4879d+4a27301b/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.26.6360.0+6f180f47/0_REGISTRY/0.4.7.dev0/3338d63843deff76593f3d84514682d027ea1b76f25961369968d0a698b30250/5588dab006640194a286.json ADDED
@@ -0,0 +1,63 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": 1,
22
+ "capacity_factor": null,
23
+ "checkpoint_id": "meta-llama/Llama-3.3-70B-Instruct",
24
+ "checkpoint_revision": "6f6073b423013f6a7d4d9f39144961bfbfbc386b",
25
+ "continuous_batching": false,
26
+ "ep_degree": 1,
27
+ "fused_qkv": true,
28
+ "glu_mlp": true,
29
+ "local_ranks_size": 64,
30
+ "max_batch_size": 1,
31
+ "max_context_length": 16384,
32
+ "max_topk": 256,
33
+ "n_active_tokens": 16384,
34
+ "neuronxcc_version": "2.26.6360.0+6f180f47",
35
+ "on_device_sampling": true,
36
+ "optimum_neuron_version": "0.4.7.dev0",
37
+ "output_logits": false,
38
+ "pp_degree": 1,
39
+ "prefill_chunk_size": 1024,
40
+ "sequence_length": 16384,
41
+ "speculation_length": 0,
42
+ "start_rank_id": 0,
43
+ "target": "trn2",
44
+ "torch_dtype": "bfloat16",
45
+ "tp_degree": 64
46
+ },
47
+ "num_attention_heads": 64,
48
+ "num_hidden_layers": 80,
49
+ "num_key_value_heads": 8,
50
+ "pretraining_tp": 1,
51
+ "rms_norm_eps": 1e-05,
52
+ "rope_scaling": {
53
+ "factor": 8.0,
54
+ "high_freq_factor": 4.0,
55
+ "low_freq_factor": 1.0,
56
+ "original_max_position_embeddings": 8192,
57
+ "rope_type": "llama3"
58
+ },
59
+ "rope_theta": 500000.0,
60
+ "tie_word_embeddings": false,
61
+ "use_cache": true,
62
+ "vocab_size": 128256
63
+ }
neuronxcc-2.26.6360.0+6f180f47/MODULE_e090011e15ec2b236cbd+665810b0/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", "--verbose=35", "--logfile=/tmp/nxd_model/chunked_prefill/_tp0_bk0/log-neuron-cc.txt"]
neuronxcc-2.26.6360.0+6f180f47/MODULE_e090011e15ec2b236cbd+665810b0/model.done ADDED
File without changes
neuronxcc-2.26.6360.0+6f180f47/MODULE_e090011e15ec2b236cbd+665810b0/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2e4803d2ec5cf4e1d9815229618808d905572272a866e0d51827a4032cbe0ee0
3
+ size 9557152
neuronxcc-2.26.6360.0+6f180f47/MODULE_e090011e15ec2b236cbd+665810b0/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aedff946443edd1452c1a2003777a0050ebc24bdc07ca3b6cfbbf4d948f8464d
3
+ size 10469720
neuronxcc-2.26.6360.0+6f180f47/MODULE_e70e014e5ec2a1b4879d+4a27301b/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", "--verbose=35", "--logfile=/tmp/nxd_model/token_generation/_tp0_bk0/log-neuron-cc.txt", "--enable-internal-neff-wrapper"]
neuronxcc-2.26.6360.0+6f180f47/MODULE_e70e014e5ec2a1b4879d+4a27301b/model.done ADDED
File without changes
neuronxcc-2.26.6360.0+6f180f47/MODULE_e70e014e5ec2a1b4879d+4a27301b/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ee1e3639d0835f60dad536af74c8200098d562ef507b72f250c8307eda9c520a
3
+ size 7065041
neuronxcc-2.26.6360.0+6f180f47/MODULE_e70e014e5ec2a1b4879d+4a27301b/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e065dc0b49ae65298b0b921d19436e0443c7aa4e5c0c2c3cbaba896cfb4ee1c3
3
+ size 6511123
neuronxcc-2.26.6360.0+6f180f47/MODULE_e70e014e5ec2a1b4879d+4a27301b/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:67a8bd7973cb3722d0eaf8aa7d52896bab4efd07cb272fcc1ae94cc243211b63
3
+ size 6881346