nev commited on
Commit
8db4a19
1 Parent(s): 5615738

Push model using huggingface_hub.

Browse files
Files changed (3) hide show
  1. config.json +5 -1
  2. generation_config.json +1 -1
  3. pytorch_model.bin +2 -2
config.json CHANGED
@@ -3,9 +3,12 @@
3
  "architectures": [
4
  "GPTNeoXForCausalLM"
5
  ],
 
6
  "bos_token_id": 0,
 
7
  "eos_token_id": 0,
8
  "hidden_act": "gelu",
 
9
  "hidden_size": 768,
10
  "initializer_range": 0.02,
11
  "intermediate_size": 3072,
@@ -14,11 +17,12 @@
14
  "model_type": "gpt_neox",
15
  "num_attention_heads": 12,
16
  "num_hidden_layers": 12,
 
17
  "rotary_emb_base": 10000,
18
  "rotary_pct": 0.25,
19
  "tie_word_embeddings": false,
20
  "torch_dtype": "float32",
21
- "transformers_version": "4.27.1",
22
  "use_cache": true,
23
  "use_parallel_residual": true,
24
  "vocab_size": 50304
 
3
  "architectures": [
4
  "GPTNeoXForCausalLM"
5
  ],
6
+ "attention_dropout": 0.0,
7
  "bos_token_id": 0,
8
+ "classifier_dropout": 0.1,
9
  "eos_token_id": 0,
10
  "hidden_act": "gelu",
11
+ "hidden_dropout": 0.0,
12
  "hidden_size": 768,
13
  "initializer_range": 0.02,
14
  "intermediate_size": 3072,
 
17
  "model_type": "gpt_neox",
18
  "num_attention_heads": 12,
19
  "num_hidden_layers": 12,
20
+ "rope_scaling": null,
21
  "rotary_emb_base": 10000,
22
  "rotary_pct": 0.25,
23
  "tie_word_embeddings": false,
24
  "torch_dtype": "float32",
25
+ "transformers_version": "4.33.0",
26
  "use_cache": true,
27
  "use_parallel_residual": true,
28
  "vocab_size": 50304
generation_config.json CHANGED
@@ -2,5 +2,5 @@
2
  "_from_model_config": true,
3
  "bos_token_id": 0,
4
  "eos_token_id": 0,
5
- "transformers_version": "4.27.1"
6
  }
 
2
  "_from_model_config": true,
3
  "bos_token_id": 0,
4
  "eos_token_id": 0,
5
+ "transformers_version": "4.33.0"
6
  }
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:cd57842cad583ae86cc72d114571b9fbfc09b5c379919eab6a374c6de26d9b4d
3
- size 699682733
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:087ce37f69a7a4ec487b7dd3c413c29788e19120a9fc3571ad11418bc2167906
3
+ size 649345613