plamo-13b / config.json
dhigurashi
Fix context length
b26c18c
raw history blame
No virus
783 Bytes
{
"architectures": [
"PlamoForCausalLM"
],
"auto_map": {
"AutoConfig": "modeling_plamo.PlamoConfig",
"AutoModelForCausalLM": "modeling_plamo.PlamoForCausalLM"
},
"bos_token_id": 1,
"eos_token_id": 2,
"hidden_act": "silu",
"hidden_size": 5120,
"initializer_range": 0.02,
"intermediate_size": 16640,
"max_position_embeddings": 4096,
"model_type": "plamo",
"n_shared_head": 8,
"num_attention_heads": 40,
"num_hidden_layers": 40,
"num_key_value_heads": 40,
"pad_token_id": 0,
"rms_norm_eps": 1e-06,
"tie_word_embeddings": false,
"tokenizer_class": "PlamoTokenizer",
"torch_dtype": "bfloat16",
"transformers_version": "4.32.0",
"use_cache": true,
"vocab_size": 50432
}