atsuki-yamaguchi commited on
Commit
f843f8e
1 Parent(s): 1bb362e

Upload config.json with huggingface_hub

Browse files
Files changed (1) hide show
  1. config.json +30 -30
config.json CHANGED
@@ -1,31 +1,31 @@
1
  {
2
- "_name_or_path": "/jmain02/home/J2AD003/txk102/shared/models/Meta-Llama-3-8B-te_2^15_50000_50-mean-proj",
3
- "architectures": [
4
- "LlamaForMultiCausalLMv2"
5
- ],
6
- "attention_bias": false,
7
- "attention_dropout": 0.0,
8
- "bos_token_id": 128000,
9
- "copy_lm_head": true,
10
- "eos_token_id": 128001,
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
- "num_lm_heads": 1,
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
- "torch_dtype": "float16",
28
- "transformers_version": "4.44.0.dev0",
29
- "use_cache": true,
30
- "vocab_size": 128312
31
- }
 
1
  {
2
+ "_name_or_path": "meta-llama/Meta-Llama-3-8B",
3
+ "architectures": [
4
+ "LlamaForMultiCausalLMv2"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 128000,
9
+ "copy_lm_head": true,
10
+ "eos_token_id": 128001,
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
+ "num_lm_heads": 1,
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
+ "torch_dtype": "float16",
28
+ "transformers_version": "4.44.0.dev0",
29
+ "use_cache": true,
30
+ "vocab_size": 128312
31
+ }