[BUGFIX]修复w4a8低延迟报错

vllm/model_executor/layers/fused_moe/config.py

@@ -599,6 +599,7 @@ def int8_w4a8_moe_quant_config(
         per_act_token_quant=per_act_token_quant,
         per_out_ch_quant=False,
         block_shape=block_shape,
+        weight_dtype="int4"
     )
 
 

vllm/model_executor/layers/quantization/slimquant_w4a8_marlin.py

@@ -3,6 +3,7 @@ import os
 import torch
 from torch.nn.parameter import Parameter
 
+from vllm.logger import init_logger
 import vllm.envs as envs
 from vllm import _custom_ops as ops
 from vllm.config import get_current_vllm_config