config.json
826 B · 35 lines · json Raw
1 {
2 "architectures": [
3 "LlamaForCausalLM"
4 ],
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": 131072,
14 "mlp_bias": false,
15 "model_type": "llama",
16 "num_attention_heads": 32,
17 "num_hidden_layers": 32,
18 "num_key_value_heads": 8,
19 "pretraining_tp": 1,
20 "rms_norm_eps": 1e-05,
21 "rope_scaling": {
22 "factor": 8.0,
23 "low_freq_factor": 1.0,
24 "high_freq_factor": 4.0,
25 "original_max_position_embeddings": 8192,
26 "rope_type": "llama3"
27 },
28 "rope_theta": 500000.0,
29 "tie_word_embeddings": false,
30 "torch_dtype": "bfloat16",
31 "transformers_version": "4.43.0.dev0",
32 "use_cache": true,
33 "vocab_size": 128256
34 }
35