32 lines
822 B
JSON
32 lines
822 B
JSON
|
{
|
||
|
|
"_name_or_path": "/data03/moe_chatlaw/models/moe_chatlaw_pt_4/checkpoint-26000",
|
||
|
|
"architectures": [
|
||
|
|
"MixtralForCausalLM"
|
||
|
|
],
|
||
|
|
"attention_bias": true,
|
||
|
|
"attention_dropout": 0.0,
|
||
|
|
"bos_token_id": 1,
|
||
|
|
"eos_token_id": 2,
|
||
|
|
"hidden_act": "silu",
|
||
|
|
"hidden_size": 2048,
|
||
|
|
"initializer_range": 0.02,
|
||
|
|
"intermediate_size": 5504,
|
||
|
|
"max_position_embeddings": 8192,
|
||
|
|
"model_type": "mixtral",
|
||
|
|
"num_attention_heads": 16,
|
||
|
|
"num_experts_per_tok": 2,
|
||
|
|
"num_hidden_layers": 24,
|
||
|
|
"num_key_value_heads": 16,
|
||
|
|
"num_local_experts": 4,
|
||
|
|
"output_router_logits": true,
|
||
|
|
"rms_norm_eps": 1e-06,
|
||
|
|
"rope_theta": 10000.0,
|
||
|
|
"router_aux_loss_coef": 0.001,
|
||
|
|
"sliding_window": 4096,
|
||
|
|
"tie_word_embeddings": false,
|
||
|
|
"torch_dtype": "bfloat16",
|
||
|
|
"transformers_version": "4.36.2",
|
||
|
|
"use_cache": true,
|
||
|
|
"vocab_size": 151936
|
||
|
|
}
|