* Copyright (c) 2025 Huawei Technologies Co., Ltd.
* This program is free software, you can redistribute it and/or modify it under the terms and conditions of
* CANN Open Software License Agreement Version 2.0 (the "License").
* Please refer to the License for details. You may not use this file except in compliance with the License.
* THIS SOFTWARE IS PROVIDED ON AN "AS IS" BASIS, WITHOUT WARRANTIES OF ANY KIND, EITHER EXPRESS OR IMPLIED,
* INCLUDING BUT NOT LIMITED TO NON-INFRINGEMENT, MERCHANTABILITY, OR FITNESS FOR A PARTICULAR PURPOSE.
* See LICENSE in the root of the software repository for the full text of the License.
*/
#include "onnx_common.h"
#include "math/reduce_max/op_graph/reduce_max_proto.h"
using namespace ge;
namespace domi {
using NodeProto = ge::onnx::NodeProto;
static Status ParseParamsReduceMax(const Message* op_src, ge::Operator& op_dest)
{
const NodeProto* node = dynamic_cast<const NodeProto*>(op_src);
if (node == nullptr) {
OP_LOGE(GetOpName(op_dest).c_str(), "Dynamic cast op_src to NodeProto failed.");
return FAILED;
}
std::vector<int> axes = {};
bool keep_dims = true;
for (const auto& attr : node->attribute()) {
if (attr.name() == "axes" && attr.type() == ge::onnx::AttributeProto::INTS) {
for (int i = 0; i < attr.ints_size(); i++) {
axes.push_back(attr.ints(i));
}
} else if (attr.name() == "keepdims" && attr.type() == ge::onnx::AttributeProto::INT) {
keep_dims = (attr.i() == 1);
}
if (attr.name() == "noop_with_empty_axes" && attr.type() == ge::onnx::AttributeProto::INT && attr.i() == 1) {
OP_LOGW(GetOpName(op_dest).c_str(), "Only support noop_with_empty_axes=0, but 1 is obtained now");
}
}
int num = axes.size();
std::vector<int64_t> dims = {};
if (num != 0) {
dims.push_back(num);
} else {
dims.push_back(0);
}
ge::Tensor tensor = Vec2Tensor(axes, dims, ge::DT_INT32, ge::FORMAT_NCHW);
op_dest.SetAttr("name", node->name());
op_dest.SetAttr("axes", tensor);
op_dest.SetAttr("keep_dims", keep_dims);
const int input = 2;
const int output = 1;
op_dest.DynamicInputRegister("x", input);
op_dest.DynamicOutputRegister("y", output);
op_dest.SetAttr("original_type", "ai.onnx::11::ReduceMax");
return SUCCESS;
}
static Status ParseOpToGraphReduceMax(const ge::Operator& op, Graph& graph)
{
std::string ori_name;
if (op.GetAttr("name", ori_name) != SUCCESS) {
OP_LOGE(GetOpName(op).c_str(), "get name from op failed.");
return FAILED;
}
auto data0 = op::Data((ori_name + "_data0").c_str()).set_attr_index(0);
ge::Tensor axes;
if (op.GetAttr("axes", axes) != SUCCESS) {
OP_LOGE(GetOpName(op).c_str(), "get axes from op failed");
return FAILED;
}
auto data1 = op::Const((ori_name + "_data1").c_str()).set_attr_value(axes);
auto reducemax = op::ReduceMax((ori_name + "_ReduceMax").c_str()).set_input_x(data0).set_input_axes(data1);
bool keep_dims = false;
if (op.GetAttr("keep_dims", keep_dims) != SUCCESS) {
OP_LOGE(GetOpName(op).c_str(), "get keep_dims from op failed");
return FAILED;
}
reducemax.set_attr_keep_dims(keep_dims);
std::vector<ge::Operator> inputs{data0};
std::vector<std::pair<ge::Operator, std::vector<size_t> > > outputs;
outputs.emplace_back(reducemax, std::vector<std::size_t>{0});
graph.SetInputs(inputs).SetOutputs(outputs);
return SUCCESS;
}
REGISTER_CUSTOM_OP("PartitionedCall")
.FrameworkType(ONNX)
.OriginOpType({ge::AscendString("ai.onnx::8::ReduceMax"),
ge::AscendString("ai.onnx::9::ReduceMax"),
ge::AscendString("ai.onnx::10::ReduceMax"),
ge::AscendString("ai.onnx::11::ReduceMax"),
ge::AscendString("ai.onnx::12::ReduceMax"),
ge::AscendString("ai.onnx::13::ReduceMax"),
ge::AscendString("ai.onnx::14::ReduceMax"),
ge::AscendString("ai.onnx::15::ReduceMax"),
ge::AscendString("ai.onnx::16::ReduceMax"),
ge::AscendString("ai.onnx::17::ReduceMax"),
ge::AscendString("ai.onnx::18::ReduceMax")})
.ParseParamsFn(ParseParamsReduceMax)
.ParseOpToGraphFn(ParseOpToGraphReduceMax)
.ImplyType(ImplyType::TVM);
}
