dacorvo HF Staff commited on
Commit
93c6afa
·
verified ·
1 Parent(s): bf0ed87

Synchronizing local compiler cache.

Browse files
.gitattributes CHANGED
@@ -11224,3 +11224,6 @@ neuronxcc-2.21.18209.0+043b1bf7/MODULE_ff2ba14ba7a42bf8f139+ed72d204/model.neff
11224
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_1035505fd438c4658518+ed72d204/model.neff filter=lfs diff=lfs merge=lfs -text
11225
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_ef6f3b1e10993380a741+a9d440f5/model.neff filter=lfs diff=lfs merge=lfs -text
11226
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_ef6f3b1e10993380a741+a9d440f5/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
 
 
 
 
11224
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_1035505fd438c4658518+ed72d204/model.neff filter=lfs diff=lfs merge=lfs -text
11225
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_ef6f3b1e10993380a741+a9d440f5/model.neff filter=lfs diff=lfs merge=lfs -text
11226
  neuronxcc-2.21.18209.0+043b1bf7/MODULE_ef6f3b1e10993380a741+a9d440f5/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
11227
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_a3d40e66d561221ab9cd+a9d440f5/model.neff filter=lfs diff=lfs merge=lfs -text
11228
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_a3d40e66d561221ab9cd+a9d440f5/wrapped_neff.hlo filter=lfs diff=lfs merge=lfs -text
11229
+ neuronxcc-2.21.18209.0+043b1bf7/MODULE_fe82938100ca85904a2b+ed72d204/model.neff filter=lfs diff=lfs merge=lfs -text
neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.3.1.dev7/granite/ibm-granite/granite-3.1-2b-instruct/fd175e949913df61edbf.json ADDED
@@ -0,0 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": 8,
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": 2,
32
+ "logical_nc_config": 1,
33
+ "max_batch_size": 8,
34
+ "max_context_length": 4096,
35
+ "max_topk": 256,
36
+ "n_active_tokens": 4096,
37
+ "neuronxcc_version": "2.21.18209.0+043b1bf7",
38
+ "on_device_sampling": true,
39
+ "optimum_neuron_version": "0.3.1.dev7",
40
+ "output_logits": false,
41
+ "pp_degree": 1,
42
+ "sequence_length": 4096,
43
+ "speculation_length": 0,
44
+ "start_rank_id": 0,
45
+ "target": null,
46
+ "torch_dtype": "bfloat16",
47
+ "tp_degree": 2
48
+ },
49
+ "num_attention_heads": 32,
50
+ "num_hidden_layers": 40,
51
+ "num_key_value_heads": 8,
52
+ "residual_multiplier": 0.22,
53
+ "rms_norm_eps": 1e-05,
54
+ "rope_scaling": null,
55
+ "rope_theta": 5000000.0,
56
+ "tie_word_embeddings": true,
57
+ "use_cache": true,
58
+ "vocab_size": 49155
59
+ }
neuronxcc-2.21.18209.0+043b1bf7/MODULE_a3d40e66d561221ab9cd+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_a3d40e66d561221ab9cd+a9d440f5/model.done ADDED
File without changes
neuronxcc-2.21.18209.0+043b1bf7/MODULE_a3d40e66d561221ab9cd+a9d440f5/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a19f29cd4351b423ff936c3807dac9b8a4924aa533484fcce3972ed58aaaae32
3
+ size 1072587
neuronxcc-2.21.18209.0+043b1bf7/MODULE_a3d40e66d561221ab9cd+a9d440f5/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e882e89d3bcd85f671f6bdde457f342de62984b4bbfd6cca42ebfd53ab622b1b
3
+ size 9165824
neuronxcc-2.21.18209.0+043b1bf7/MODULE_a3d40e66d561221ab9cd+a9d440f5/wrapped_neff.hlo ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b3ad7730578c8a5795e13cb7edf4b61ce890f9f8c45f1dfe3f48ffb54618021b
3
+ size 9350146
neuronxcc-2.21.18209.0+043b1bf7/MODULE_fe82938100ca85904a2b+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_fe82938100ca85904a2b+ed72d204/model.done ADDED
File without changes
neuronxcc-2.21.18209.0+043b1bf7/MODULE_fe82938100ca85904a2b+ed72d204/model.hlo_module.pb ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fa64925a1a3f9f09a823aa7382aa1210fa1e9a6ab14f86fd4d3299bdcb4d136e
3
+ size 1050665
neuronxcc-2.21.18209.0+043b1bf7/MODULE_fe82938100ca85904a2b+ed72d204/model.neff ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3dacf2daec485c695be0722a20b1208177be6e11d99aeaca751a7b3bc381ba68
3
+ size 3851264