dacorvo HF Staff commited on
Commit
f107e00
·
verified ·
1 Parent(s): 624a591

Synchronizing local compiler cache.

Browse files
.gitattributes CHANGED
@@ -12316,3 +12316,6 @@ neuronxcc-2.21.18209.0+043b1bf7/MODULE_67016de67079b10e98eb+24129607/model.neff
12316
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_88b7febbfc80a0e4b2a6+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
12317
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_f1d98d5989c0fcb7f417+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
12318
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_f1d98d5989c0fcb7f417+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
 
 
 
 
12316
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_88b7febbfc80a0e4b2a6+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
12317
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_f1d98d5989c0fcb7f417+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
12318
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_f1d98d5989c0fcb7f417+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
12319
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_70c94de7da8329114b63+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
12320
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_a1a557aa2cfe582c3a40+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
12321
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_a1a557aa2cfe582c3a40+a02c3a36/wrapped_neff.hlo 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/ccd00831179ec7166449.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": 4,
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": 2,
31
+ "max_batch_size": 4,
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": 2
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_70c94de7da8329114b63+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_70c94de7da8329114b63+24129607/model.done ADDED
File without changes
neuronxcc-2.21.18209.0+043b1bf7/MODULE_70c94de7da8329114b63+24129607/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5888ba384287f48a018fb0e26636d42ccbc0f7744d11b9b713c883fd85fd3ef7
3
+ size 1055558
neuronxcc-2.21.18209.0+043b1bf7/MODULE_70c94de7da8329114b63+24129607/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b21fafb98682a2d0ef0863c094ada3e30b6d2ba83452b92db030fbd4372c190a
3
+ size 6267904
neuronxcc-2.21.18209.0+043b1bf7/MODULE_a1a557aa2cfe582c3a40+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_a1a557aa2cfe582c3a40+a02c3a36/model.done ADDED
File without changes
neuronxcc-2.21.18209.0+043b1bf7/MODULE_a1a557aa2cfe582c3a40+a02c3a36/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4303ddb62779fb637392afc9a4ae2323f186eae32c355641a83efecbd4eb7813
3
+ size 1078440
neuronxcc-2.21.18209.0+043b1bf7/MODULE_a1a557aa2cfe582c3a40+a02c3a36/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c72fac91fc3b19cf244a944c7fba942390fc53bc6b14294f92376131c8d8a37c
3
+ size 11899904
neuronxcc-2.21.18209.0+043b1bf7/MODULE_a1a557aa2cfe582c3a40+a02c3a36/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:630d56f66836bfe8eb4fabb8cbb844b3bf32d8490c698e420404f9bd9b21553a
3
+ size 12084506