WarrenWang01 commited on
Commit
aa77045
·
verified ·
1 Parent(s): 78562d6

Update config.json

Browse files
Files changed (1) hide show
  1. config.json +2 -5
config.json CHANGED
@@ -5,26 +5,23 @@
5
  "attention_bias": false,
6
  "attention_dropout": 0.0,
7
  "bos_token_id": 128000,
8
- "dtype": "float32",
9
  "eos_token_id": 128001,
10
- "head_dim": 128,
11
  "hidden_act": "silu",
12
  "hidden_size": 4096,
13
  "initializer_range": 0.02,
14
  "intermediate_size": 14336,
15
  "max_position_embeddings": 8192,
16
- "mlp_bias": false,
17
  "model_type": "llama",
18
  "num_attention_heads": 32,
19
  "num_hidden_layers": 32,
20
  "num_key_value_heads": 8,
21
- "pad_token_id": 128001,
22
  "pretraining_tp": 1,
23
  "rms_norm_eps": 1e-05,
24
  "rope_scaling": null,
25
  "rope_theta": 500000.0,
26
  "tie_word_embeddings": false,
27
- "transformers_version": "4.57.3",
 
28
  "use_cache": true,
29
  "vocab_size": 128256
30
  }
 
5
  "attention_bias": false,
6
  "attention_dropout": 0.0,
7
  "bos_token_id": 128000,
 
8
  "eos_token_id": 128001,
 
9
  "hidden_act": "silu",
10
  "hidden_size": 4096,
11
  "initializer_range": 0.02,
12
  "intermediate_size": 14336,
13
  "max_position_embeddings": 8192,
 
14
  "model_type": "llama",
15
  "num_attention_heads": 32,
16
  "num_hidden_layers": 32,
17
  "num_key_value_heads": 8,
 
18
  "pretraining_tp": 1,
19
  "rms_norm_eps": 1e-05,
20
  "rope_scaling": null,
21
  "rope_theta": 500000.0,
22
  "tie_word_embeddings": false,
23
+ "torch_dtype": "bfloat16",
24
+ "transformers_version": "4.40.0.dev0",
25
  "use_cache": true,
26
  "vocab_size": 128256
27
  }