optimum-internal-testing-user commited on
Commit
7c9d8c8
·
verified ·
1 Parent(s): ba4ca71

Synchronizing local compiler cache.

Browse files
neuronxcc-2.21.18209.0+043b1bf7/0_REGISTRY/0.4.2.dev0/granite/ibm-granite/granite-3.1-2b-instruct/cf4efab08fca47af8a0e.json ADDED
@@ -0,0 +1,57 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": 4,
23
+ "capacity_factor": null,
24
+ "checkpoint_id": "ibm-granite/granite-3.1-2b-instruct",
25
+ "checkpoint_revision": "bbc2aed595bd38bd770263dc3ab831db9794441d",
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": "0.4.2.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": 5000000.0,
54
+ "tie_word_embeddings": true,
55
+ "use_cache": true,
56
+ "vocab_size": 49155
57
+ }
neuronxcc-2.21.18209.0+043b1bf7/MODULE_9d35248a93d4142a3cf1+24129607/model.hlo_module.pb CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:4c3701ec3661417e95b7a0d48488bffafc155a720c62b2885bc15ad31e9a4dc1
3
  size 739558
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:89d95332373996418dd36c71e92d055f25151179af769bdcbf34f30aa3bcacb6
3
  size 739558
neuronxcc-2.21.18209.0+043b1bf7/MODULE_9d35248a93d4142a3cf1+24129607/model.neff CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e118545b87bf0dea461da4fa5ecc6540935443dde798e8227db35b0c315f5967
3
  size 26133504
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dee47034ab783eac59be02bccb1aa5ab2b746cf1bd350578fff367704fec7c2c
3
  size 26133504
neuronxcc-2.21.18209.0+043b1bf7/MODULE_c2b3afa51c57d431a332+a02c3a36/model.neff CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:92421ee451d0ae6a58b57769a43978826b242921a10cb1886b5e28c980e35afe
3
  size 1926144
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:78410670de86fb85b626bbcd51a5cb4825e7176c757a61fdf64a03dafb9915fd
3
  size 1926144
neuronxcc-2.21.18209.0+043b1bf7/MODULE_c2b3afa51c57d431a332+a02c3a36/wrapped_neff.hlo CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b3946ae0306fc1753558f9ca642ba3114c1774bbc3f5713830256f08eea19c9c
3
  size 2082478
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3146e765c0a8e68312e1844f7d08a458fe6e0e46ad72afb51d3cade89de7d9e9
3
  size 2082478