perlthoughts commited on
Commit
dcb0297
1 Parent(s): a746dce

Update config.json

Browse files
Files changed (1) hide show
  1. config.json +5 -4
config.json CHANGED
@@ -1,5 +1,5 @@
1
  {
2
- "_name_or_path": "/content/new-merged",
3
  "architectures": [
4
  "MistralForCausalLM"
5
  ],
@@ -9,14 +9,15 @@
9
  "hidden_size": 4096,
10
  "initializer_range": 0.02,
11
  "intermediate_size": 14336,
12
- "max_position_embeddings": 8192,
13
  "model_type": "mistral",
14
  "num_attention_heads": 32,
15
  "num_hidden_layers": 48,
16
  "num_key_value_heads": 8,
17
  "rms_norm_eps": 1e-05,
18
- "rope_theta": 10000.0,
19
- "sliding_window": 4096,
 
20
  "tie_word_embeddings": false,
21
  "torch_dtype": "float16",
22
  "transformers_version": "4.35.2",
 
1
  {
2
+ "_name_or_path": "CallComply/Starling-LM-11B-alpha",
3
  "architectures": [
4
  "MistralForCausalLM"
5
  ],
 
9
  "hidden_size": 4096,
10
  "initializer_range": 0.02,
11
  "intermediate_size": 14336,
12
+ "max_position_embeddings": 32769,
13
  "model_type": "mistral",
14
  "num_attention_heads": 32,
15
  "num_hidden_layers": 48,
16
  "num_key_value_heads": 8,
17
  "rms_norm_eps": 1e-05,
18
+ "rope_scaling": null,
19
+ "rope_theta": 100000.0,
20
+ "sliding_window": null,
21
  "tie_word_embeddings": false,
22
  "torch_dtype": "float16",
23
  "transformers_version": "4.35.2",