dacorvo HF Staff commited on
Commit
23ad168
·
verified ·
1 Parent(s): 75410b3

Synchronizing local compiler cache.

Browse files
.gitattributes CHANGED
@@ -13149,3 +13149,6 @@ neuronxcc-2.21.18209.0+043b1bf7/MODULE_e26c5f9afdbc6d9d4282+a32116a7/model.neff
13149
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_e57c3946dfa9c87cb07f+9936de79/model.neff filter=lfs diff=lfs merge=lfs -text
13150
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_e57c3946dfa9c87cb07f+9936de79/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
13151
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_f9cc02790e4e79aee5af+fb4cc044/model.neff filter=lfs diff=lfs merge=lfs -text
 
 
 
 
13149
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_e57c3946dfa9c87cb07f+9936de79/model.neff filter=lfs diff=lfs merge=lfs -text
13150
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_e57c3946dfa9c87cb07f+9936de79/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
13151
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_f9cc02790e4e79aee5af+fb4cc044/model.neff filter=lfs diff=lfs merge=lfs -text
13152
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_cd5f0b0df65e1b4a6bf5+24129607/model.neff filter=lfs diff=lfs merge=lfs -text
13153
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_e36f5e0ce14d1b9618e7+a02c3a36/model.neff filter=lfs diff=lfs merge=lfs -text
13154
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_e36f5e0ce14d1b9618e7+a02c3a36/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.2/granite/ibm-granite/granite-3.3-8b-instruct/63e1314219a229b693b7.json ADDED
@@ -0,0 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "ibm-granite/granite-3.3-8b-instruct",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "GraniteForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.0,
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-3.3-8b-instruct",
26
+ "checkpoint_revision": "51dd4bc2ade4059a6bd87649d68aa11e4fb2529b",
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": 16384,
34
+ "max_topk": 256,
35
+ "n_active_tokens": 16384,
36
+ "neuronxcc_version": "2.21.18209.0+043b1bf7",
37
+ "on_device_sampling": true,
38
+ "optimum_neuron_version": "0.4.2",
39
+ "output_logits": false,
40
+ "pp_degree": 1,
41
+ "sequence_length": 16384,
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": 49159
58
+ }
neuronxcc-2.21.18209.0+043b1bf7/MODULE_cd5f0b0df65e1b4a6bf5+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_cd5f0b0df65e1b4a6bf5+24129607/model.done ADDED
File without changes
neuronxcc-2.21.18209.0+043b1bf7/MODULE_cd5f0b0df65e1b4a6bf5+24129607/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9f8a711589356c8692658082894779fd7604804606263594924b5578c58dc43e
3
+ size 922776
neuronxcc-2.21.18209.0+043b1bf7/MODULE_cd5f0b0df65e1b4a6bf5+24129607/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8357cbcc340f1eaa9570839ad3fc78ef16e789e87e8ec862707ff5acbef14814
3
+ size 16661504
neuronxcc-2.21.18209.0+043b1bf7/MODULE_e36f5e0ce14d1b9618e7+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_e36f5e0ce14d1b9618e7+a02c3a36/model.done ADDED
File without changes
neuronxcc-2.21.18209.0+043b1bf7/MODULE_e36f5e0ce14d1b9618e7+a02c3a36/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:407f935149f119f2c4164c2d2938216d252069b973106eb3e92e093934a304df
3
+ size 1022685
neuronxcc-2.21.18209.0+043b1bf7/MODULE_e36f5e0ce14d1b9618e7+a02c3a36/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3d21046e265bc255cf9a5752c980c317bff3bf1d628e6ff41e4e22a2009b296e
3
+ size 3933184
neuronxcc-2.21.18209.0+043b1bf7/MODULE_e36f5e0ce14d1b9618e7+a02c3a36/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:af8f092d4ab4208c9e22144c5268b83c6e2d845fbaf3d49072a4cdc000cd4c6f
3
+ size 4117950