ae206e73创建于 2025年3月18日历史提交
// 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;
using npu_utils = at_npu::native::NpuUtils;

namespace {
at::Tensor& soft_margin_loss_out_nocheck(
    at::Tensor& result,
    const at::Tensor& self,
    const at::Tensor& target,
    int64_t reduction)
{
    at::Tensor target_broadcast = target;
    if (target.sizes() != self.sizes()) {
        target_broadcast = acl_op::npu_broadcast(target, self.sizes());
    }
    string reduction_str(op_plugin::utils::get_reduction_str(reduction));
    at_npu::native::OpCommand cmd;
    cmd.Name("SoftMarginLoss")
        .Input(self)
        .Input(target_broadcast)
        .Output(result)
        .Attr("reduction", reduction_str)
        .Run();
    return result;
}
} // namespace

at::Tensor& soft_margin_loss_out(
    const at::Tensor& self,
    const at::Tensor& target,
    int64_t reduction,
    at::Tensor& out)
{
    auto output_size = op_infer::soft_margin_loss_npu_output_size(self, reduction);
    npu_preparation::CheckOut(
        {self, target},
        out,
        self,
        output_size);

    if (!npu_utils::check_match(&out)) {
        at::Tensor contiguous_out = npu_utils::format_contiguous(out);
        soft_margin_loss_out_nocheck(contiguous_out, self, target, reduction);
        npu_utils::format_fresh_view(out, contiguous_out);
    } else {
        soft_margin_loss_out_nocheck(out, self, target, reduction);
    }
    return out;
}

at::Tensor soft_margin_loss(const at::Tensor& self, const at::Tensor& target, int64_t reduction)
{
    auto output_size = op_infer::soft_margin_loss_npu_output_size(self, reduction);
    at::Tensor result = npu_preparation::apply_tensor(self, output_size);

    soft_margin_loss_out_nocheck(result, self, target, reduction);

    if (reduction == at::Reduction::None) {
        return result;
    } else {
        return result.reshape({});
    }
}
} // namespace acl_op