dacorvo HF Staff commited on
Commit
0d7aca6
·
verified ·
1 Parent(s): 52e67aa

Synchronizing local compiler cache.

Browse files
.gitattributes CHANGED
@@ -11518,3 +11518,6 @@ neuronxcc-2.21.18209.0+043b1bf7/MODULE_cebb35fcc40cba5dc00d+a9d440f5/wrapped_nef
11518
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_42f954cd17b1e3cb1cb7+a9d440f5/model.neff filter=lfs diff=lfs merge=lfs -text
11519
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_42f954cd17b1e3cb1cb7+a9d440f5/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
11520
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_a6732a46e76c97a186c7+ed72d204/model.neff filter=lfs diff=lfs merge=lfs -text
 
 
 
 
11518
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_42f954cd17b1e3cb1cb7+a9d440f5/model.neff filter=lfs diff=lfs merge=lfs -text
11519
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_42f954cd17b1e3cb1cb7+a9d440f5/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
11520
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_a6732a46e76c97a186c7+ed72d204/model.neff filter=lfs diff=lfs merge=lfs -text
11521
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_140ffe5563dbe460bca0+ed72d204/model.neff filter=lfs diff=lfs merge=lfs -text
11522
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_85fe6d3a7c65fbb1ab27+a9d440f5/model.neff filter=lfs diff=lfs merge=lfs -text
11523
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_85fe6d3a7c65fbb1ab27+a9d440f5/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.0.dev0/granite/ibm-granite/granite-3.1-2b-instruct/8dc41fb40cdb427405be.json ADDED
@@ -0,0 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_entry_class": "SingleModelCacheEntry",
3
+ "_model_id": "ibm-granite/granite-3.1-2b-instruct",
4
+ "_task": "text-generation",
5
+ "architectures": [
6
+ "GraniteForCausalLM"
7
+ ],
8
+ "attention_bias": false,
9
+ "attention_dropout": 0.1,
10
+ "attention_multiplier": 0.015625,
11
+ "embedding_multiplier": 12.0,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 2048,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 8192,
16
+ "logits_scaling": 8.0,
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-3.1-2b-instruct",
25
+ "checkpoint_revision": "bbc2aed595bd38bd770263dc3ab831db9794441d",
26
+ "continuous_batching": true,
27
+ "enable_bucketing": false,
28
+ "ep_degree": 1,
29
+ "fused_qkv": true,
30
+ "glu_mlp": true,
31
+ "local_ranks_size": 8,
32
+ "max_batch_size": 32,
33
+ "max_context_length": 4096,
34
+ "max_topk": 256,
35
+ "n_active_tokens": 4096,
36
+ "neuronxcc_version": "2.21.18209.0+043b1bf7",
37
+ "on_device_sampling": true,
38
+ "optimum_neuron_version": "0.4.0.dev0",
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": 5000000.0,
55
+ "tie_word_embeddings": true,
56
+ "use_cache": true,
57
+ "vocab_size": 49155
58
+ }
neuronxcc-2.21.18209.0+043b1bf7/MODULE_140ffe5563dbe460bca0+ed72d204/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_model/_tp0_bk0/log-neuron-cc.txt"]
neuronxcc-2.21.18209.0+043b1bf7/MODULE_140ffe5563dbe460bca0+ed72d204/model.done ADDED
File without changes
neuronxcc-2.21.18209.0+043b1bf7/MODULE_140ffe5563dbe460bca0+ed72d204/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8ba42fe8c406cd7507dbb59b08bee60d59a04b7c3db7f1e3100ef0c29c2b1cfc
3
+ size 997471
neuronxcc-2.21.18209.0+043b1bf7/MODULE_140ffe5563dbe460bca0+ed72d204/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8d8c3c5a10b8392ea736b8cb11bc495cb79ac04e4e95fc3cb89f28e9c04a3dd0
3
+ size 25201664
neuronxcc-2.21.18209.0+043b1bf7/MODULE_85fe6d3a7c65fbb1ab27+a9d440f5/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_model/_tp0_bk0/log-neuron-cc.txt", "--enable-internal-neff-wrapper"]
neuronxcc-2.21.18209.0+043b1bf7/MODULE_85fe6d3a7c65fbb1ab27+a9d440f5/model.done ADDED
File without changes
neuronxcc-2.21.18209.0+043b1bf7/MODULE_85fe6d3a7c65fbb1ab27+a9d440f5/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3c6c5805c176e02671b777aa97df3fc3fc0e885ef8b5edfa57a80c6f0e00d86a
3
+ size 1014720
neuronxcc-2.21.18209.0+043b1bf7/MODULE_85fe6d3a7c65fbb1ab27+a9d440f5/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fb0c2046f5f23b99c0976a90f9588e8586485796667cf5d5d49656ddd1b17a3b
3
+ size 7158784
neuronxcc-2.21.18209.0+043b1bf7/MODULE_85fe6d3a7c65fbb1ab27+a9d440f5/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:365674d55f98f1f4142e859684165b6a69822569d2f5802684e0ed9701ec510a
3
+ size 7343546