update for npu. (#8210)

Author: wuhuachaocoding

paddlenlp/peft/lora/mc2_lora_npu.py

@@ -40,12 +40,12 @@ def backward(ctx, dy):
  input_, weight = ctx.saved_tensor()
  out_grad = dy
  sub_grad = out_grad.reshape([-1, out_grad.shape[-1]])
- input_grad = paddle.matmul(sub_grad, weight.t())
+ input_grad = paddle.matmul(sub_grad, weight, transpose_y=True)
  if weight.stop_gradient:
  return input_grad.reshape(input_.shape)
  else:
  input_reshape = input_.reshape([-1, input_.shape[-1]])
- weight_grad = input_reshape.t().matmul(sub_grad)
+ weight_grad = paddle.matmul(input_reshape, sub_grad, transpose_x=True)
  return input_grad.reshape(input_.shape), weight_grad
 
 
@@ -65,7 +65,11 @@ def backward(ctx, dy):
  rank = paddle.distributed.get_rank()
  hcom_name = ctx.group.process_group.get_comm_name(rank)
 
- d_weight = input_.reshape([-1, input_.shape[-1]]).t().matmul(sub_grad) if not weight.stop_gradient else None
+ d_weight = (
+ paddle.matmul(input_.reshape([-1, input_.shape[-1]]), sub_grad, transpose_x=True)
+ if not weight.stop_gradient
+ else None
+ )
  d_input = paddle_custom_device.npu.fused_mm_allreduce(
  sub_grad, weight.t(), bias=None, hcom=hcom_name, reduce_op="sum", comm_turn=0
  )