chinese-chat-81m / config.json
BravePaul's picture
full epoch
1697829
raw
history blame contribute delete
644 Bytes
{
"_name_or_path": "/data/pythia-retrain/data_model/chinese-chat-81m",
"architectures": [
"GPTNeoXForCausalLM"
],
"bos_token_id": 0,
"eos_token_id": 2,
"hidden_act": "gelu",
"hidden_size": 768,
"initializer_range": 0.02,
"intermediate_size": 2048,
"layer_norm_eps": 1e-05,
"max_position_embeddings": 768,
"model_type": "gpt_neox",
"num_attention_heads": 12,
"num_hidden_layers": 12,
"rotary_emb_base": 10000,
"rotary_pct": 0.25,
"tie_word_embeddings": false,
"torch_dtype": "float32",
"transformers_version": "4.28.0.dev0",
"use_cache": true,
"use_parallel_residual": true,
"vocab_size": 12831
}