// Copyright (c) 2023 Huawei Technologies Co., Ltd
// All rights reserved.
//
// Licensed under the BSD 3-Clause License (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// https://opensource.org/licenses/BSD-3-Clause
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
#include "op_plugin/AclOpsInterface.h"
#include "op_plugin/utils/OpAdapter.h"
namespace acl_op {
using npu_preparation = at_npu::native::OpPreparation;
namespace {
at::Tensor threshold_backward_out_npu(
at::Tensor& result,
const at::Tensor& grad_output,
const at::Tensor& self,
at::Scalar threshold)
{
at_npu::native::OpCommand cmd;
// The performance of the ReluGrad operator is better than that of ThresholdGradV2D.
// However, ReluGrad does not support the scenario where threshold is not 0.
if (op_plugin::utils::get_scalar_float_value(threshold) != static_cast<float>(0)) {
cmd.Name("ThresholdGradV2D")
.Input(grad_output)
.Input(self)
.Output(result)
.Attr("threshold", threshold)
.Run();
} else {
cmd.Name("ReluGrad")
.Input(grad_output)
.Input(self)
.Output(result)
.Run();
}
return result;
}
} // namespace
at::Tensor threshold_backward(
const at::Tensor& grad_output,
const at::Tensor& self,
const at::Scalar& threshold)
{
auto output_size = op_infer::broadcast_ops_npu_output_size(grad_output, self);
at::Tensor result = npu_preparation::apply_tensor(self, output_size);
threshold_backward_out_npu(result, grad_output, self, threshold);
return result;
}
} // namespace acl_op