fix format error

Signed-off-by: Jinzhen Lin <[email protected]>

Author: jinzhen-lin

vllm/model_executor/layers/fused_moe/fused_moe.py

@@ -557,8 +557,6 @@ def invoke_fused_moe_kernel(A: torch.Tensor,
         )
 
     else:
-        ndim = lambda x: 0 if x is None else x.ndim
-
         fused_moe_kernel[grid](
             A,
             B,

vllm/model_executor/layers/quantization/moe_quant_int.py

@@ -104,11 +104,11 @@ def get_quant_method(self, layer: torch.nn.Module,
             return UnquantizedLinearMethod()
         elif isinstance(layer, LinearBase):
             if self.linear_quant_method == "gptq":
-                config = GPTQMarlinConfig.from_config(self.full_config)
-                return GPTQMarlinLinearMethod(config)
+                gptq_config = GPTQMarlinConfig.from_config(self.full_config)
+                return GPTQMarlinLinearMethod(gptq_config)
             elif self.linear_quant_method == "awq":
-                config = AWQMarlinConfig.from_config(self.full_config)
-                return AWQMarlinLinearMethod(config)
+                awq_config = AWQMarlinConfig.from_config(self.full_config)
+                return AWQMarlinLinearMethod(awq_config)
             else:
                 raise ValueError("moe_quant_int only support gptq and awq.")
         elif isinstance(layer, FusedMoE):