Hjgugugjhuhjggg commited on
Commit
11129b9
1 Parent(s): 08ab8cd

Update fine-tuned/config.json

Browse files
Files changed (1) hide show
  1. fine-tuned/config.json +49 -1
fine-tuned/config.json CHANGED
@@ -1 +1,49 @@
1
- {"model_type": "llama", "torch_dtype": "float16", "eos_token_id": 128009, "bos_token_id": 128000, "pad_token_id": 128009, "vocab_size": 128256}
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "huihui-ai/Llama-3.2-3B-Instruct-abliterated",
3
+ "model_type": "llama",
4
+ "torch_dtype": "float16",
5
+ "transformers_version": "4.46.3",
6
+ "architectures": ["LlamaForCausalLM"],
7
+ "vocab_size": 128256,
8
+ "bos_token_id": 128000,
9
+ "eos_token_id": [128001, 128008, 128009],
10
+ "pad_token_id": 128009,
11
+ "hidden_size": 3072,
12
+ "intermediate_size": 8192,
13
+ "num_hidden_layers": 28,
14
+ "num_attention_heads": 24,
15
+ "num_key_value_heads": 8,
16
+ "head_dim": 128,
17
+ "max_position_embeddings": 131072,
18
+ "rope_scaling": {
19
+ "factor": 32.0,
20
+ "high_freq_factor": 4.0,
21
+ "low_freq_factor": 1.0,
22
+ "original_max_position_embeddings": 8192,
23
+ "rope_type": "llama3"
24
+ },
25
+ "rope_theta": 500000.0,
26
+ "hidden_act": "silu",
27
+ "initializer_range": 0.02,
28
+ "rms_norm_eps": 1e-05,
29
+ "layer_norm_eps": 1e-05,
30
+ "attention_bias": false,
31
+ "mlp_bias": false,
32
+ "pretraining_tp": 1,
33
+ "tie_word_embeddings": true,
34
+ "use_cache": true,
35
+ "shared_input_output_embedding": true,
36
+ "attention_dropout": 0.0,
37
+ "hidden_dropout_prob": 0.1,
38
+ "attn_dropout_prob": 0.1,
39
+ "residual_dropout_prob": 0.0,
40
+ "ffn_dropout_prob": 0.1,
41
+ "output_projection": true,
42
+ "position_embedding_type": "rotary",
43
+ "use_rope": true,
44
+ "gradient_checkpointing": false,
45
+ "multi_query_attention": false,
46
+ "output_attentions": false,
47
+ "output_hidden_states": false,
48
+ "layerdrop": 0.0
49
+ }