[Bugfix] fix adding bias twice in ipex GPTQ quantization (vllm-project#18363)

Signed-off-by: rand-fly <[email protected]>
Signed-off-by: Chenheli Hua <[email protected]>

Author: rand-fly

vllm/model_executor/layers/quantization/ipex_quant.py

@@ -181,8 +181,6 @@ def apply(self,
               bias: Optional[torch.Tensor] = None) -> torch.Tensor:
         reshaped_x = x.reshape(-1, x.shape[-1])
         out = layer.ipex_qlinear(reshaped_x)
-        if bias is not None:
-            out.add_(bias)
         return out.reshape(x.shape[:-1] + (layer.ipex_output_size, ))