LLaMA-65B-HF / config.json
Neko-Institute-of-Science's picture
LLaMA: convert_llama_weights_to_hf.py as of c0f99b4d2ec73090595914dde4c16da207e21d73
dff3a11
raw
history blame
507 Bytes
{
"architectures": [
"LlamaForCausalLM"
],
"bos_token_id": 1,
"eos_token_id": 2,
"hidden_act": "silu",
"hidden_size": 8192,
"initializer_range": 0.02,
"intermediate_size": 22016,
"max_position_embeddings": 2048,
"model_type": "llama",
"num_attention_heads": 64,
"num_hidden_layers": 80,
"pad_token_id": 0,
"rms_norm_eps": 1e-05,
"tie_word_embeddings": false,
"torch_dtype": "float16",
"transformers_version": "4.28.0.dev0",
"use_cache": true,
"vocab_size": 32000
}