baffo32 commited on
Commit
85b14e2
1 Parent(s): 890390b

add alternative config + keys is_split, subfolder

Browse files
Files changed (2) hide show
  1. config.json +51 -0
  2. model/config.json +2 -1
config.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation_function": "gelu_new",
3
+ "architectures": [
4
+ "GPTNeoForCausalLM"
5
+ ],
6
+ "attention_dropout": 0,
7
+ "attention_layers": ["global", "global", "global", "global", "global", "global", "global", "global", "global", "global", "global", "global", "global", "global", "global", "global", "global", "global", "global", "global", "global", "global", "global", "global", "global", "global", "global", "global"],
8
+ "attention_types": [
9
+ [
10
+ [
11
+ "global"
12
+ ],
13
+ 28
14
+ ]
15
+ ],
16
+ "bos_token_id": 50256,
17
+ "embed_dropout": 0,
18
+ "eos_token_id": 50256,
19
+ "gradient_checkpointing": false,
20
+ "hidden_size": 4096,
21
+ "initializer_range": 0.02,
22
+ "intermediate_size": null,
23
+ "is_split": true,
24
+ "jax": true,
25
+ "layer_norm_epsilon": 1e-05,
26
+ "max_position_embeddings": 2048,
27
+ "model_type": "gpt_neo",
28
+ "num_heads": 16,
29
+ "num_layers": 28,
30
+ "resid_dropout": 0,
31
+ "rotary": true,
32
+ "rotary_dim": 64,
33
+ "subfolder": "model",
34
+ "summary_activation": null,
35
+ "summary_first_dropout": 0.1,
36
+ "summary_proj_to_labels": true,
37
+ "summary_type": "cls_index",
38
+ "summary_use_proj": true,
39
+ "transformers_version": "4.10.0.dev0",
40
+ "use_cache": true,
41
+ "vocab_size": 50400,
42
+ "window_size": 256,
43
+ "tokenizer_class": "GPT2Tokenizer",
44
+ "task_specific_params": {
45
+ "text-generation": {
46
+ "do_sample": true,
47
+ "temperature": 1.0,
48
+ "max_length": 50
49
+ }
50
+ }
51
+ }
model/config.json CHANGED
@@ -20,6 +20,7 @@
20
  "hidden_size": 4096,
21
  "initializer_range": 0.02,
22
  "intermediate_size": null,
 
23
  "jax": true,
24
  "layer_norm_epsilon": 1e-05,
25
  "max_position_embeddings": 2048,
@@ -46,4 +47,4 @@
46
  "max_length": 50
47
  }
48
  }
49
- }
 
20
  "hidden_size": 4096,
21
  "initializer_range": 0.02,
22
  "intermediate_size": null,
23
+ "is_split": true,
24
  "jax": true,
25
  "layer_norm_epsilon": 1e-05,
26
  "max_position_embeddings": 2048,
 
47
  "max_length": 50
48
  }
49
  }
50
+ }