Crystalcareai commited on
Commit
a642d4e
1 Parent(s): ffa6a6c

Upload config.json

Browse files
Files changed (1) hide show
  1. config.json +9 -4
config.json CHANGED
@@ -1,9 +1,14 @@
1
  {
2
- "_name_or_path": "cache/qstar/1710326256/checkpoint-100",
3
  "architectures": [
4
- "MistralForCausalLM"
5
  ],
6
  "attention_dropout": 0.0,
 
 
 
 
 
7
  "bos_token_id": 1,
8
  "eos_token_id": 2,
9
  "hidden_act": "silu",
@@ -15,7 +20,7 @@
15
  "merged_lm_and_talk_heads": false,
16
  "merged_lm_and_think_heads": true,
17
  "merged_talk_heads": true,
18
- "model_type": "mistral",
19
  "num_attention_heads": 32,
20
  "num_hidden_layers": 32,
21
  "num_key_value_heads": 8,
@@ -33,4 +38,4 @@
33
  "use_shallow_think": true,
34
  "use_weighted_talk_head": true,
35
  "vocab_size": 32002
36
- }
 
1
  {
2
+ "_name_or_path": "Crystalcareai/Quiet-Star-Custom",
3
  "architectures": [
4
+ "QuietForCausalLM"
5
  ],
6
  "attention_dropout": 0.0,
7
+ "auto_map": {
8
+ "AutoConfig": "configuration_quiet.QuietConfig",
9
+ "AutoModel": "modeling_quiet.QuietModel",
10
+ "AutoModelForCausalLM": "modeling_quiet.QuietForCausalLM"
11
+ },
12
  "bos_token_id": 1,
13
  "eos_token_id": 2,
14
  "hidden_act": "silu",
 
20
  "merged_lm_and_talk_heads": false,
21
  "merged_lm_and_think_heads": true,
22
  "merged_talk_heads": true,
23
+ "model_type": "quiet",
24
  "num_attention_heads": 32,
25
  "num_hidden_layers": 32,
26
  "num_key_value_heads": 8,
 
38
  "use_shallow_think": true,
39
  "use_weighted_talk_head": true,
40
  "vocab_size": 32002
41
+ }