dacorvo HF Staff commited on
Commit
e29eb98
·
verified ·
1 Parent(s): ca35943

Synchronizing local compiler cache.

Browse files
.gitattributes CHANGED
@@ -15312,3 +15312,6 @@ neuronxcc-2.21.33363.0+82129205/MODULE_dddf25ede440961f2803+e0c92c88/model.neff
15312
  neuronxcc-2.21.33363.0+82129205/MODULE_7910abe7593310096e14+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
15313
  neuronxcc-2.21.33363.0+82129205/MODULE_edb9cdfa9b731c820282+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
15314
  neuronxcc-2.21.33363.0+82129205/MODULE_edb9cdfa9b731c820282+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
 
 
 
 
15312
  neuronxcc-2.21.33363.0+82129205/MODULE_7910abe7593310096e14+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
15313
  neuronxcc-2.21.33363.0+82129205/MODULE_edb9cdfa9b731c820282+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
15314
  neuronxcc-2.21.33363.0+82129205/MODULE_edb9cdfa9b731c820282+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
15315
+ neuronxcc-2.21.33363.0+82129205/MODULE_324cf28a37a583b3cd91+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
15316
+ neuronxcc-2.21.33363.0+82129205/MODULE_324cf28a37a583b3cd91+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
15317
+ neuronxcc-2.21.33363.0+82129205/MODULE_de8ac3775bbbc72c1a2c+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.21.33363.0+82129205/0_REGISTRY/0.4.4.dev2/granite/ibm-granite/granite-3.1-8b-instruct/0fb21780d82eca6a23c5.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": "0.4.4.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_324cf28a37a583b3cd91+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_324cf28a37a583b3cd91+a02c3a36/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_324cf28a37a583b3cd91+a02c3a36/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cae226d97643fd4fd7d164b528eec0bd74dad00035321929d46ae8b5e3147683
3
+ size 1028298
neuronxcc-2.21.33363.0+82129205/MODULE_324cf28a37a583b3cd91+a02c3a36/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d01e38877f447b2adcfee995d6be4f0519713031281a71144a314e6e34a82c5f
3
+ size 3421184
neuronxcc-2.21.33363.0+82129205/MODULE_324cf28a37a583b3cd91+a02c3a36/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:90cadc54a3157084dff5b933305dbd21eb69b894fc2258c198f72285b852f54f
3
+ size 3605243
neuronxcc-2.21.33363.0+82129205/MODULE_de8ac3775bbbc72c1a2c+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_de8ac3775bbbc72c1a2c+24129607/model.done ADDED
File without changes
neuronxcc-2.21.33363.0+82129205/MODULE_de8ac3775bbbc72c1a2c+24129607/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d1cdbae19cdbcfd33acbe1c109e362080157393ea480d4b0158e79f344026e67
3
+ size 912376
neuronxcc-2.21.33363.0+82129205/MODULE_de8ac3775bbbc72c1a2c+24129607/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2812414a26a49f4e824f3662ba4746f4510b108904bf1600190def06dc811bcf
3
+ size 2059264