dacorvo HF Staff commited on
Commit
0bfa767
·
verified ·
1 Parent(s): 5a7865f

Synchronizing local compiler cache.

Browse files
.gitattributes CHANGED
@@ -17909,3 +17909,6 @@ neuronxcc-2.26.6360.0+6f180f47/MODULE_5b87c52d288ff0e3b45b+12dd9119/model.neff f
17909
  neuronxcc-2.26.6360.0+6f180f47/MODULE_7be0c41bfa68130d2a26+84337dd9/model.neff filter=lfs diff=lfs merge=lfs -text
17910
  neuronxcc-2.26.6360.0+6f180f47/MODULE_7be0c41bfa68130d2a26+84337dd9/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
17911
  neuronxcc-2.26.6360.0+6f180f47/MODULE_df5ac55d00b940623a44+12dd9119/model.neff filter=lfs diff=lfs merge=lfs -text
 
 
 
 
17909
  neuronxcc-2.26.6360.0+6f180f47/MODULE_7be0c41bfa68130d2a26+84337dd9/model.neff filter=lfs diff=lfs merge=lfs -text
17910
  neuronxcc-2.26.6360.0+6f180f47/MODULE_7be0c41bfa68130d2a26+84337dd9/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
17911
  neuronxcc-2.26.6360.0+6f180f47/MODULE_df5ac55d00b940623a44+12dd9119/model.neff filter=lfs diff=lfs merge=lfs -text
17912
+ neuronxcc-2.26.6360.0+6f180f47/MODULE_762fb05ca99a656ad8b1+12dd9119/model.neff filter=lfs diff=lfs merge=lfs -text
17913
+ neuronxcc-2.26.6360.0+6f180f47/MODULE_aa36e4f5714f9502f617+84337dd9/model.neff filter=lfs diff=lfs merge=lfs -text
17914
+ neuronxcc-2.26.6360.0+6f180f47/MODULE_aa36e4f5714f9502f617+84337dd9/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.26.6360.0+6f180f47/0_REGISTRY/0.4.7.dev0/6e4d675089b03121ef48dedefef342d87e80260dd41c0bf51aa06adbcc2331f6/eed1fc6b9602aed1c5df.json ADDED
@@ -0,0 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "ibm-granite/granite-3.1-8b-instruct",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "GraniteForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.1,
10
+ "attention_multiplier": 0.0078125,
11
+ "dtype": "bfloat16",
12
+ "embedding_multiplier": 12.0,
13
+ "hidden_act": "silu",
14
+ "hidden_size": 4096,
15
+ "initializer_range": 0.02,
16
+ "intermediate_size": 12800,
17
+ "logits_scaling": 16.0,
18
+ "max_position_embeddings": 131072,
19
+ "mlp_bias": false,
20
+ "model_type": "granite",
21
+ "neuron": {
22
+ "_serialized_key": "NxDNeuronConfig",
23
+ "batch_size": 1,
24
+ "capacity_factor": null,
25
+ "checkpoint_id": "ibm-granite/granite-3.1-8b-instruct",
26
+ "checkpoint_revision": "4009206d5fc95d2e65a7b7633e159d6e97e25d35",
27
+ "continuous_batching": false,
28
+ "ep_degree": 1,
29
+ "fused_qkv": true,
30
+ "glu_mlp": true,
31
+ "local_ranks_size": 8,
32
+ "max_batch_size": 1,
33
+ "max_context_length": 4096,
34
+ "max_topk": 256,
35
+ "n_active_tokens": 4096,
36
+ "neuronxcc_version": "2.26.6360.0+6f180f47",
37
+ "on_device_sampling": true,
38
+ "optimum_neuron_version": "0.4.7.dev0",
39
+ "output_logits": false,
40
+ "pp_degree": 1,
41
+ "prefill_chunk_size": 1024,
42
+ "sequence_length": 4096,
43
+ "speculation_length": 0,
44
+ "start_rank_id": 0,
45
+ "target": "trn1",
46
+ "torch_dtype": "bfloat16",
47
+ "tp_degree": 8
48
+ },
49
+ "num_attention_heads": 32,
50
+ "num_hidden_layers": 40,
51
+ "num_key_value_heads": 8,
52
+ "residual_multiplier": 0.22,
53
+ "rms_norm_eps": 1e-05,
54
+ "rope_scaling": null,
55
+ "rope_theta": 10000000.0,
56
+ "tie_word_embeddings": true,
57
+ "use_cache": true,
58
+ "vocab_size": 49155
59
+ }
neuronxcc-2.26.6360.0+6f180f47/MODULE_762fb05ca99a656ad8b1+12dd9119/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", "--verbose=35", "--logfile=/tmp/nxd_model/chunked_prefill/_tp0_bk0/log-neuron-cc.txt"]
neuronxcc-2.26.6360.0+6f180f47/MODULE_762fb05ca99a656ad8b1+12dd9119/model.done ADDED
File without changes
neuronxcc-2.26.6360.0+6f180f47/MODULE_762fb05ca99a656ad8b1+12dd9119/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c5b8239f10d81684081530a5c7b9e4d233d895ae00c33eae9959e8a0d7e363f0
3
+ size 5681769
neuronxcc-2.26.6360.0+6f180f47/MODULE_762fb05ca99a656ad8b1+12dd9119/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3289fdbcbcb70b2bfdfde15bc1f23f77c44cfefea5e94cb38e7af4a502e956ab
3
+ size 5643264
neuronxcc-2.26.6360.0+6f180f47/MODULE_aa36e4f5714f9502f617+84337dd9/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", "--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_aa36e4f5714f9502f617+84337dd9/model.done ADDED
File without changes
neuronxcc-2.26.6360.0+6f180f47/MODULE_aa36e4f5714f9502f617+84337dd9/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a1c697df0490f86383c01cb092a5b40a39ec1de800f2050bfd8e3f01045128cc
3
+ size 4238344
neuronxcc-2.26.6360.0+6f180f47/MODULE_aa36e4f5714f9502f617+84337dd9/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fbaf107d3abcd9877004eb8996f319beb096aea2fabbac99bac5f0720853ee43
3
+ size 3185664
neuronxcc-2.26.6360.0+6f180f47/MODULE_aa36e4f5714f9502f617+84337dd9/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2141698a19fea1d099abef276f20eeb536e1a22633f329d15f139f2569a1e945
3
+ size 3369443