[AutoParallel] use inplace multiply in grad_clip (PaddlePaddle#71421)

* use inplace multiply in grad_clip

* use inplace multiply in grad_clip

Author: Waynezee

python/paddle/nn/clip.py

@@ -840,8 +840,12 @@ def async_add_n(var_list):
                         clip_input = paddle.distributed.reshard(
                             clip_input, g.process_mesh, clip_input.placements
                         )
-                new_grad = paddle.multiply(g, clip_input)
-                params_and_grads.append((p, new_grad))
+                if g.is_dist() or g.is_dense():
+                    g.multiply_(clip_input)
+                    params_and_grads.append((p, g))
+                else:
+                    new_grad = paddle.multiply(g, clip_input)
+                    params_and_grads.append((p, new_grad))
             else:
                 params_and_grads.append((p, g))