CLEX-LLaMA-2-7B-64K / config.json
Guanzheng's picture
Update config.json
fe79b60 verified
raw
history blame contribute delete
999 Bytes
{
"architectures": [
"LlamaForCausalLM"
],
"auto_map": {
"AutoConfig": "configuration_llama_clex.CLEXLlamaConfig",
"AutoModelForCausalLM": "modeling_llama_clex.LlamaForCausalLM"
},
"attention_bias": false,
"attention_dropout": 0.0,
"bos_token_id": 1,
"data_length": 4096,
"eos_token_id": 2,
"hidden_act": "silu",
"hidden_size": 4096,
"initializer_range": 0.02,
"intermediate_size": 11008,
"log_scale": false,
"max_position_embeddings": 4096,
"model_type": "llama",
"num_attention_heads": 32,
"num_hidden_layers": 32,
"num_key_value_heads": 32,
"pad_token_id": 0,
"pretraining_tp": 1,
"rms_norm_eps": 1e-05,
"rope_scaling": {
"factor": 1,
"max_factor": 64,
"param_factor": 1,
"time_dt": 0.01,
"type": "clex",
"act": "tanh"
},
"rope_theta": 10000,
"tie_word_embeddings": false,
"torch_dtype": "bfloat16",
"transformers_version": "4.36.2",
"use_cache": true,
"use_flashattn": true,
"vocab_size": 32000
}