dacorvo HF Staff commited on
Commit
9b935b9
·
verified ·
1 Parent(s): 468d4bc

Synchronizing local compiler cache.

Browse files
.gitattributes CHANGED
@@ -8364,3 +8364,7 @@ neuronxcc-2.17.194.0+d312836f/MODULE_1078451219dc7dbb1c51+bfe5714b/model.neff fi
8364
  neuronxcc-2.17.194.0+d312836f/MODULE_8ec406df99dd6c5b1337+431f5505/model.neff filter=lfs diff=lfs merge=lfs -text
8365
  neuronxcc-2.17.194.0+d312836f/MODULE_941d20d4622201c6330c+165e9558/model.neff filter=lfs diff=lfs merge=lfs -text
8366
  neuronxcc-2.17.194.0+d312836f/MODULE_941d20d4622201c6330c+165e9558/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
8364
  neuronxcc-2.17.194.0+d312836f/MODULE_8ec406df99dd6c5b1337+431f5505/model.neff filter=lfs diff=lfs merge=lfs -text
8365
  neuronxcc-2.17.194.0+d312836f/MODULE_941d20d4622201c6330c+165e9558/model.neff filter=lfs diff=lfs merge=lfs -text
8366
  neuronxcc-2.17.194.0+d312836f/MODULE_941d20d4622201c6330c+165e9558/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
8367
+ neuronxcc-2.17.194.0+d312836f/MODULE_1f60313545027ac57424+431f5505/model.neff filter=lfs diff=lfs merge=lfs -text
8368
+ neuronxcc-2.17.194.0+d312836f/MODULE_610f0973895e56713cd0+bfe5714b/model.neff filter=lfs diff=lfs merge=lfs -text
8369
+ neuronxcc-2.17.194.0+d312836f/MODULE_780a119f00b7adf5b848+165e9558/model.neff filter=lfs diff=lfs merge=lfs -text
8370
+ neuronxcc-2.17.194.0+d312836f/MODULE_780a119f00b7adf5b848+165e9558/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.17.194.0+d312836f/0_REGISTRY/0.2.0/llama/princeton-nlp/Sheared-LLaMA-1.3B/1d0e302d5c5a8ce527dc.json ADDED
@@ -0,0 +1,71 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "princeton-nlp/Sheared-LLaMA-1.3B",
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": 2048,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 5504,
15
+ "max_position_embeddings": 4096,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "neuron": {
19
+ "_serialized_key": "NxDNeuronConfig",
20
+ "async_mode": false,
21
+ "attn_kernel_enabled": false,
22
+ "batch_size": 1,
23
+ "capacity_factor": null,
24
+ "cc_pipeline_tiling_factor": 2,
25
+ "checkpoint_id": "princeton-nlp/Sheared-LLaMA-1.3B",
26
+ "checkpoint_revision": "a4b76938edbf571ea7d7d9904861cbdca08809b4",
27
+ "continuous_batching": false,
28
+ "enable_bucketing": false,
29
+ "ep_degree": 1,
30
+ "flash_decoding_enabled": false,
31
+ "fused_qkv": true,
32
+ "glu_mlp": true,
33
+ "is_chunked_prefill": false,
34
+ "local_ranks_size": 2,
35
+ "logical_nc_config": 1,
36
+ "max_batch_size": 1,
37
+ "max_context_length": 4096,
38
+ "max_topk": 256,
39
+ "mlp_kernel_enabled": false,
40
+ "mlp_kernel_fuse_residual_add": false,
41
+ "n_active_tokens": 4096,
42
+ "neuronxcc_version": "2.17.194.0+d312836f",
43
+ "num_cores_per_group": 1,
44
+ "on_device_sampling": true,
45
+ "optimum_neuron_version": "0.2.0",
46
+ "output_logits": false,
47
+ "padding_side": "right",
48
+ "pp_degree": 1,
49
+ "qk_layernorm": false,
50
+ "qkv_kernel_enabled": false,
51
+ "rpl_reduce_dtype": "float16",
52
+ "sequence_length": 4096,
53
+ "sequence_parallel_enabled": false,
54
+ "speculation_length": 0,
55
+ "start_rank_id": 0,
56
+ "target": null,
57
+ "torch_dtype": "float16",
58
+ "tp_degree": 2,
59
+ "vocab_parallel": false
60
+ },
61
+ "num_attention_heads": 16,
62
+ "num_hidden_layers": 24,
63
+ "num_key_value_heads": 16,
64
+ "pretraining_tp": 1,
65
+ "rms_norm_eps": 1e-05,
66
+ "rope_scaling": null,
67
+ "rope_theta": 10000.0,
68
+ "tie_word_embeddings": false,
69
+ "use_cache": true,
70
+ "vocab_size": 32000
71
+ }
neuronxcc-2.17.194.0+d312836f/MODULE_1f60313545027ac57424+431f5505/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ "--model-type=transformer -O1 --lnc=1 --internal-hlo2tensorizer-options=--experimental-unsafe-fp8e4m3fn-as-fp8e4m3 --logfile=/tmp/nxd_model/layout_opt/log-neuron-cc.txt"
neuronxcc-2.17.194.0+d312836f/MODULE_1f60313545027ac57424+431f5505/model.done ADDED
File without changes
neuronxcc-2.17.194.0+d312836f/MODULE_1f60313545027ac57424+431f5505/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:72c2a9763ca764fb5852843ee3b84c3b1bbb5dbf1ffa6e0f157d387448813b4b
3
+ size 100886
neuronxcc-2.17.194.0+d312836f/MODULE_1f60313545027ac57424+431f5505/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f14c6190a5666a11356f4fbc47ed5dc28b8c2c7a26407378d7af5efd3f66f6fe
3
+ size 1240064
neuronxcc-2.17.194.0+d312836f/MODULE_610f0973895e56713cd0+bfe5714b/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ "--auto-cast=none --model-type=transformer --tensorizer-options='--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-dge-dma --vectorize-strided-dma ' -O2 --internal-num-neuroncores-per-sengine=1 --logfile=/tmp/nxd_model/context_encoding_model/_tp0_bk0/log-neuron-cc.txt"
neuronxcc-2.17.194.0+d312836f/MODULE_610f0973895e56713cd0+bfe5714b/model.done ADDED
File without changes
neuronxcc-2.17.194.0+d312836f/MODULE_610f0973895e56713cd0+bfe5714b/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cc37dcbb75402a8ae57e89b839e16dbf862d3baff8d09fbcdea09087a9ec978c
3
+ size 614606
neuronxcc-2.17.194.0+d312836f/MODULE_610f0973895e56713cd0+bfe5714b/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1e659e0ae7b080682dd022f2d420d56f0803d79f59655764ee074d4969f24bf6
3
+ size 29697024
neuronxcc-2.17.194.0+d312836f/MODULE_6a90845afeaebacf0657+165e9558/model.hlo_module.pb CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:6111f9f26a111a1e1864e91592ff50809f794c38c7c36460f1dd7f4f6dd831f8
3
  size 756302
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c4a63f420c1792cab22b5c74765cfbbaed9645a97986a21e6e4debf8359a16a0
3
  size 756302
neuronxcc-2.17.194.0+d312836f/MODULE_780a119f00b7adf5b848+165e9558/compile_flags.json ADDED
@@ -0,0 +1 @@
 
 
1
+ "--auto-cast=none --model-type=transformer --tensorizer-options='--enable-ccop-compute-overlap --cc-pipeline-tiling-factor=2 --vectorize-dge-dma --vectorize-strided-dma ' -O2 --internal-num-neuroncores-per-sengine=1 --logfile=/tmp/nxd_model/token_generation_model/_tp0_bk0/log-neuron-cc.txt --enable-internal-neff-wrapper"
neuronxcc-2.17.194.0+d312836f/MODULE_780a119f00b7adf5b848+165e9558/model.done ADDED
File without changes
neuronxcc-2.17.194.0+d312836f/MODULE_780a119f00b7adf5b848+165e9558/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:15d5437061eb6194fb4d5d98fe794ce786baaa92263805aaf7489a94bb528ec2
3
+ size 522310
neuronxcc-2.17.194.0+d312836f/MODULE_780a119f00b7adf5b848+165e9558/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:acec861dfc116e32666d852ef282727d56e63f5138a3fc2d08e36fad629d2296
3
+ size 2079744
neuronxcc-2.17.194.0+d312836f/MODULE_780a119f00b7adf5b848+165e9558/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:79aa5cf1c464b3cf622b46365c6ee447cf9a750754216881fd47b49965eaaa21
3
+ size 2182452