back to forward_native

vllm/model_executor/layers/activation.py

@@ -85,10 +85,10 @@ class SiluAndMul(CustomOp):
     @staticmethod
     def forward_native(x: torch.Tensor) -> torch.Tensor:
         """PyTorch-native implementation equivalent to forward()."""
-        if not torch.compiler.is_compiling() and envs.VLLM_USE_OPT_OP:
-            return self.forward_cuda(x)
-        else:
-            d = x.shape[-1] // 2
+        # if not torch.compiler.is_compiling() and envs.VLLM_USE_OPT_OP:
+        #     return self.forward_cuda(x)
+        # else:
+        d = x.shape[-1] // 2
         return F.silu(x[..., :d]) * x[..., d:]
 
     def forward_cuda(self, x: torch.Tensor) -> torch.Tensor: