You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
Paddle/paddle/fluid/inference/tensorrt/plugin/slice_op_plugin.h

90 lines
3.1 KiB

// Copyright (c) 2018 PaddlePaddle Authors. All Rights Reserved.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// 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.
#pragma once
#include <algorithm>
#include <string>
#include <vector>
#include "paddle/fluid/inference/tensorrt/engine.h"
#include "paddle/fluid/inference/tensorrt/plugin/trt_plugin.h"
namespace paddle {
namespace inference {
namespace tensorrt {
namespace plugin {
#if IS_TRT_VERSION_GE(6000)
class SlicePluginDynamic : public DynamicPluginTensorRT {
public:
explicit SlicePluginDynamic(std::vector<int> starts, std::vector<int> ends,
std::vector<int> axes, bool ban_fp16)
: starts_(starts), ends_(ends), axes_(axes), ban_fp16_(ban_fp16) {}
SlicePluginDynamic(void const* serialData, size_t serialLength) {}
nvinfer1::IPluginV2DynamicExt* clone() const override {
return new SlicePluginDynamic(starts_, ends_, axes_, ban_fp16_);
}
const char* getPluginType() const override { return "slice_plugin"; }
int getNbOutputs() const override { return 1; }
int initialize() override;
size_t getSerializationSize() const override;
void serialize(void* buffer) const override;
nvinfer1::DimsExprs getOutputDimensions(
int output_index, const nvinfer1::DimsExprs* inputs, int nb_inputs,
nvinfer1::IExprBuilder& expr_builder) override;
bool supportsFormatCombination(int pos,
const nvinfer1::PluginTensorDesc* inOut,
int nbInputs, int nbOutputs) override;
void configurePlugin(const nvinfer1::DynamicPluginTensorDesc* in,
int nbInputs,
const nvinfer1::DynamicPluginTensorDesc* out,
int nbOutputs) override {}
size_t getWorkspaceSize(const nvinfer1::PluginTensorDesc* inputs,
int nbInputs,
const nvinfer1::PluginTensorDesc* outputs,
int nbOutputs) const override {
return 0;
}
int enqueue(const nvinfer1::PluginTensorDesc* inputDesc,
const nvinfer1::PluginTensorDesc* outputDesc,
const void* const* inputs, void* const* outputs, void* workspace,
cudaStream_t stream) override;
nvinfer1::DataType getOutputDataType(int index,
const nvinfer1::DataType* inputTypes,
int nbInputs) const override;
void destroy() override { delete this; }
private:
std::vector<int> starts_;
std::vector<int> ends_;
std::vector<int> axes_;
bool ban_fp16_{false};
};
#endif
} // namespace plugin
} // namespace tensorrt
} // namespace inference
} // namespace paddle