lanlanlan23 commited on
Commit
14e8987
·
verified ·
1 Parent(s): 83b01c7

Upload 12b update config.json

Browse files
Files changed (1) hide show
  1. config.json +7 -7
config.json CHANGED
@@ -1,5 +1,5 @@
1
  {
2
- "_name_or_path": "/workspace/eval/models/Ministral-8B-Instruct-2410",
3
  "architectures": [
4
  "MistralForCausalLM"
5
  ],
@@ -8,17 +8,17 @@
8
  "eos_token_id": 2,
9
  "head_dim": 128,
10
  "hidden_act": "silu",
11
- "hidden_size": 4096,
12
  "initializer_range": 0.02,
13
- "intermediate_size": 12288,
14
- "max_position_embeddings": 32768,
15
  "model_type": "mistral",
16
  "num_attention_heads": 32,
17
- "num_hidden_layers": 36,
18
  "num_key_value_heads": 8,
19
  "rms_norm_eps": 1e-05,
20
- "rope_theta": 100000000.0,
21
- "sliding_window": 32768,
22
  "tie_word_embeddings": false,
23
  "torch_dtype": "bfloat16",
24
  "transformers_version": "4.49.0",
 
1
  {
2
+ "_name_or_path": "/workspace/thirdpart/Mistral-12b-v0",
3
  "architectures": [
4
  "MistralForCausalLM"
5
  ],
 
8
  "eos_token_id": 2,
9
  "head_dim": 128,
10
  "hidden_act": "silu",
11
+ "hidden_size": 5120,
12
  "initializer_range": 0.02,
13
+ "intermediate_size": 14336,
14
+ "max_position_embeddings": 1024000,
15
  "model_type": "mistral",
16
  "num_attention_heads": 32,
17
+ "num_hidden_layers": 40,
18
  "num_key_value_heads": 8,
19
  "rms_norm_eps": 1e-05,
20
+ "rope_theta": 1000000.0,
21
+ "sliding_window": null,
22
  "tie_word_embeddings": false,
23
  "torch_dtype": "bfloat16",
24
  "transformers_version": "4.49.0",