dacorvo HF Staff commited on
Commit
bcdb059
·
verified ·
1 Parent(s): c3af7a1

Synchronizing local compiler cache.

Browse files
.gitattributes CHANGED
@@ -12346,3 +12346,6 @@ neuronxcc-2.21.18209.0+043b1bf7/MODULE_e595e217d65e663a19f1+a02c3a36/wrapped_nef
12346
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_8414d37eab328896acb7+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
12347
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_8414d37eab328896acb7+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
12348
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_c0808787f3555a6627cc+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
 
 
 
 
12346
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_8414d37eab328896acb7+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
12347
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_8414d37eab328896acb7+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
12348
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_c0808787f3555a6627cc+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
12349
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_18c0230344f8d3c24650+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
12350
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_18c0230344f8d3c24650+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
12351
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_d162634b275181f3bbf3+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.1.dev0/granite/ibm-granite/granite-3.1-8b-instruct/076a9ab5cb53dd4a8905.json ADDED
@@ -0,0 +1,57 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "embedding_multiplier": 12.0,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 4096,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 12800,
16
+ "logits_scaling": 16.0,
17
+ "max_position_embeddings": 131072,
18
+ "mlp_bias": false,
19
+ "model_type": "granite",
20
+ "neuron": {
21
+ "_serialized_key": "NxDNeuronConfig",
22
+ "batch_size": 32,
23
+ "capacity_factor": null,
24
+ "checkpoint_id": "ibm-granite/granite-3.1-8b-instruct",
25
+ "checkpoint_revision": "4009206d5fc95d2e65a7b7633e159d6e97e25d35",
26
+ "continuous_batching": true,
27
+ "ep_degree": 1,
28
+ "fused_qkv": true,
29
+ "glu_mlp": true,
30
+ "local_ranks_size": 8,
31
+ "max_batch_size": 32,
32
+ "max_context_length": 4096,
33
+ "max_topk": 256,
34
+ "n_active_tokens": 4096,
35
+ "neuronxcc_version": "2.21.18209.0+043b1bf7",
36
+ "on_device_sampling": true,
37
+ "optimum_neuron_version": "0.4.1.dev0",
38
+ "output_logits": false,
39
+ "pp_degree": 1,
40
+ "sequence_length": 4096,
41
+ "speculation_length": 0,
42
+ "start_rank_id": 0,
43
+ "target": "trn1",
44
+ "torch_dtype": "bfloat16",
45
+ "tp_degree": 8
46
+ },
47
+ "num_attention_heads": 32,
48
+ "num_hidden_layers": 40,
49
+ "num_key_value_heads": 8,
50
+ "residual_multiplier": 0.22,
51
+ "rms_norm_eps": 1e-05,
52
+ "rope_scaling": null,
53
+ "rope_theta": 10000000.0,
54
+ "tie_word_embeddings": true,
55
+ "use_cache": true,
56
+ "vocab_size": 49155
57
+ }
neuronxcc-2.21.18209.0+043b1bf7/MODULE_18c0230344f8d3c24650+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_18c0230344f8d3c24650+a02c3a36/model.done ADDED
File without changes
neuronxcc-2.21.18209.0+043b1bf7/MODULE_18c0230344f8d3c24650+a02c3a36/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7c52e629c4487df7861f48a15ca691617f8f0e08a52617fac13833c27893196b
3
+ size 1021013
neuronxcc-2.21.18209.0+043b1bf7/MODULE_18c0230344f8d3c24650+a02c3a36/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:14535ddae6f339bdc658020fdebfd3003384de3b1d1bbda317bda262806746b8
3
+ size 8438784
neuronxcc-2.21.18209.0+043b1bf7/MODULE_18c0230344f8d3c24650+a02c3a36/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bfe0717080ecfda2008a96ab062d5de720776e49865ed91f272760d0c523b052
3
+ size 8622706
neuronxcc-2.21.18209.0+043b1bf7/MODULE_d162634b275181f3bbf3+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_d162634b275181f3bbf3+24129607/model.done ADDED
File without changes
neuronxcc-2.21.18209.0+043b1bf7/MODULE_d162634b275181f3bbf3+24129607/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ef6d298e87d809e14e1f1ee1bb8f84347df5c0f35d8c06903e8640243d4ed4b5
3
+ size 1001284
neuronxcc-2.21.18209.0+043b1bf7/MODULE_d162634b275181f3bbf3+24129607/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a6374a5d2ec5c37783a25b1250e8f6b5ca3244c084fe4945949f654e11ca7b05
3
+ size 2089984