dacorvo HF Staff commited on
Commit
4f8ff5b
·
verified ·
1 Parent(s): f412544

Synchronizing local compiler cache.

Browse files
.gitattributes CHANGED
@@ -16938,3 +16938,6 @@ neuronxcc-2.21.33363.0+82129205/MODULE_39bfd565dc865ea9d7a0+a02c3a36/wrapped_nef
16938
  neuronxcc-2.21.33363.0+82129205/MODULE_ad32f4c31ce53a96793e+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
16939
  neuronxcc-2.21.33363.0+82129205/MODULE_97cf303de67f78eb00f8+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
16940
  neuronxcc-2.21.33363.0+82129205/MODULE_97cf303de67f78eb00f8+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
 
 
 
 
16938
  neuronxcc-2.21.33363.0+82129205/MODULE_ad32f4c31ce53a96793e+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
16939
  neuronxcc-2.21.33363.0+82129205/MODULE_97cf303de67f78eb00f8+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
16940
  neuronxcc-2.21.33363.0+82129205/MODULE_97cf303de67f78eb00f8+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
16941
+ neuronxcc-2.21.33363.0+82129205/MODULE_12c5768296730e8339b5+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
16942
+ neuronxcc-2.21.33363.0+82129205/MODULE_12c5768296730e8339b5+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
16943
+ neuronxcc-2.21.33363.0+82129205/MODULE_c795fec87a7dac23d567+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.6.dev2/6e4d675089b03121ef48dedefef342d87e80260dd41c0bf51aa06adbcc2331f6/cc705eeb2eb13b136c6f.json ADDED
@@ -0,0 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": "2.21.33363.0+82129205",
37
+ "on_device_sampling": true,
38
+ "optimum_neuron_version": "0.4.6.dev2",
39
+ "output_logits": false,
40
+ "pp_degree": 1,
41
+ "sequence_length": 4096,
42
+ "speculation_length": 0,
43
+ "start_rank_id": 0,
44
+ "target": "trn1",
45
+ "torch_dtype": "bfloat16",
46
+ "tp_degree": 8
47
+ },
48
+ "num_attention_heads": 32,
49
+ "num_hidden_layers": 40,
50
+ "num_key_value_heads": 8,
51
+ "residual_multiplier": 0.22,
52
+ "rms_norm_eps": 1e-05,
53
+ "rope_scaling": null,
54
+ "rope_theta": 10000000.0,
55
+ "tie_word_embeddings": true,
56
+ "use_cache": true,
57
+ "vocab_size": 49155
58
+ }
neuronxcc-2.21.33363.0+82129205/MODULE_12c5768296730e8339b5+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-2.21.33363.0+82129205/MODULE_12c5768296730e8339b5+a02c3a36/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_12c5768296730e8339b5+a02c3a36/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:deb5e4855713be1db4e9bc47e40dccf3fca2886686d2ee2c0f7af66836f30f02
3
+ size 1071482
neuronxcc-2.21.33363.0+82129205/MODULE_12c5768296730e8339b5+a02c3a36/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2721900f0231625a9d8b9837553fc7273b3e40775668c7e6865d0944aceb39e8
3
+ size 3574784
neuronxcc-2.21.33363.0+82129205/MODULE_12c5768296730e8339b5+a02c3a36/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:308593fabad47d0accccd86271fdc2d0767712ea6c00b0087a195867331adcb0
3
+ size 3758843
neuronxcc-2.21.33363.0+82129205/MODULE_c795fec87a7dac23d567+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-2.21.33363.0+82129205/MODULE_c795fec87a7dac23d567+24129607/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_c795fec87a7dac23d567+24129607/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:75d49b46a600f9925e6a28ce19843eb770dc0ab599f8872a405a890487598ebb
3
+ size 912376
neuronxcc-2.21.33363.0+82129205/MODULE_c795fec87a7dac23d567+24129607/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cafc09d4a81577b7c9dc00e998d5f0ce402cc692c3a8341ee156005b1071cdcf
3
+ size 2059264