dacorvo HF Staff commited on
Commit
5e64d0d
·
verified ·
1 Parent(s): f1df01a

Synchronizing local compiler cache.

Browse files
.gitattributes CHANGED
@@ -12370,3 +12370,6 @@ neuronxcc-2.21.18209.0+043b1bf7/MODULE_eb471f8a8d4b43f4ee82+a02c3a36/wrapped_nef
12370
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_36720586c4aa77bf460d+a32116a7/model.neff filter=lfs diff=lfs merge=lfs -text
12371
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_e2ae789acde0c66adff9+ac10809c/model.neff filter=lfs diff=lfs merge=lfs -text
12372
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_e2ae789acde0c66adff9+ac10809c/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
 
 
 
 
12370
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_36720586c4aa77bf460d+a32116a7/model.neff filter=lfs diff=lfs merge=lfs -text
12371
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_e2ae789acde0c66adff9+ac10809c/model.neff filter=lfs diff=lfs merge=lfs -text
12372
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_e2ae789acde0c66adff9+ac10809c/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
12373
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_0566f3c457a72aab250a+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
12374
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_0566f3c457a72aab250a+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
12375
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_1b345f92384bb3f0eada+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.1.dev0/llama/meta-llama/Meta-Llama-3-8B/ae8f9e5d930096c0a52d.json ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "meta-llama/Meta-Llama-3-8B",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "LlamaForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 4096,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 14336,
15
+ "max_position_embeddings": 8192,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "neuron": {
19
+ "_serialized_key": "NxDNeuronConfig",
20
+ "batch_size": 1,
21
+ "capacity_factor": null,
22
+ "checkpoint_id": "meta-llama/Meta-Llama-3-8B",
23
+ "checkpoint_revision": "8cde5ca8380496c9a6cc7ef3a8b46a0372a1d920",
24
+ "continuous_batching": false,
25
+ "ep_degree": 1,
26
+ "fused_qkv": true,
27
+ "glu_mlp": true,
28
+ "local_ranks_size": 2,
29
+ "max_batch_size": 1,
30
+ "max_context_length": 4096,
31
+ "max_topk": 256,
32
+ "n_active_tokens": 4096,
33
+ "neuronxcc_version": "2.21.18209.0+043b1bf7",
34
+ "on_device_sampling": true,
35
+ "optimum_neuron_version": "0.4.1.dev0",
36
+ "output_logits": false,
37
+ "pp_degree": 1,
38
+ "sequence_length": 4096,
39
+ "speculation_length": 0,
40
+ "start_rank_id": 0,
41
+ "target": "trn1",
42
+ "torch_dtype": "bfloat16",
43
+ "tp_degree": 2
44
+ },
45
+ "num_attention_heads": 32,
46
+ "num_hidden_layers": 32,
47
+ "num_key_value_heads": 8,
48
+ "pretraining_tp": 1,
49
+ "rms_norm_eps": 1e-05,
50
+ "rope_scaling": null,
51
+ "rope_theta": 500000.0,
52
+ "tie_word_embeddings": false,
53
+ "use_cache": true,
54
+ "vocab_size": 128256
55
+ }
neuronxcc-2.21.18209.0+043b1bf7/MODULE_0566f3c457a72aab250a+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-2.21.18209.0+043b1bf7/MODULE_0566f3c457a72aab250a+a02c3a36/model.done ADDED
File without changes
neuronxcc-2.21.18209.0+043b1bf7/MODULE_0566f3c457a72aab250a+a02c3a36/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:93a4c08452de4ba995d223f961e581d0c2400884fd32f61718c7e60c599f32a4
3
+ size 800406
neuronxcc-2.21.18209.0+043b1bf7/MODULE_0566f3c457a72aab250a+a02c3a36/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f405475afe093ffccb7033b16a519c4a91ebebfecf58b1b601d4d91a01051e9d
3
+ size 9217024
neuronxcc-2.21.18209.0+043b1bf7/MODULE_0566f3c457a72aab250a+a02c3a36/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:84ee4369cd29d73762ba8379e97fbbde0a32e4021265ffb4f87bbf4409d32f92
3
+ size 9365097
neuronxcc-2.21.18209.0+043b1bf7/MODULE_1b345f92384bb3f0eada+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-2.21.18209.0+043b1bf7/MODULE_1b345f92384bb3f0eada+24129607/model.done ADDED
File without changes
neuronxcc-2.21.18209.0+043b1bf7/MODULE_1b345f92384bb3f0eada+24129607/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5f6820557929dbe7cb1f481ce916122ba3ce52b25f59f34175853599d0007b05
3
+ size 909772
neuronxcc-2.21.18209.0+043b1bf7/MODULE_1b345f92384bb3f0eada+24129607/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:747f61d3b00c70dd75d60ec20a9f9db6a35931f5193c5a2e73f4d00ce99c08b5
3
+ size 6677504