mixtral-ep64-320m / config.json
ylf1017's picture
Upload config
9578d61 verified
raw
history blame contribute delete
694 Bytes
{
"architectures": [
"MixtralForCausalLM"
],
"attention_dropout": 0.0,
"bos_token_id": 1,
"eos_token_id": 2,
"hidden_act": "silu",
"hidden_size": 1024,
"initializer_range": 0.02,
"intermediate_size": 2560,
"max_position_embeddings": 131072,
"model_type": "mixtral",
"num_attention_heads": 16,
"num_experts_per_tok": 2,
"num_hidden_layers": 24,
"num_key_value_heads": 16,
"num_local_experts": 64,
"output_router_logits": false,
"rms_norm_eps": 1e-05,
"rope_theta": 1000000.0,
"router_aux_loss_coef": 0.001,
"sliding_window": 4096,
"tie_word_embeddings": false,
"transformers_version": "4.36.0.dev0",
"use_cache": true,
"vocab_size": 32000
}