mirror of
https://github.com/NVIDIA/TensorRT-LLM.git
synced 2026-01-14 06:27:45 +08:00
b777bd6475
* Update TensorRT-LLM --------- Co-authored-by: RunningLeon <mnsheng@yeah.net> Co-authored-by: Tlntin <TlntinDeng01@Gmail.com> Co-authored-by: ZHENG, Zhen <zhengzhen.z@qq.com> Co-authored-by: Pham Van Ngoan <ngoanpham1196@gmail.com> Co-authored-by: Nathan Price <nathan@abridge.com> Co-authored-by: Tushar Goel <tushar.goel.ml@gmail.com> Co-authored-by: Mati <132419219+matichon-vultureprime@users.noreply.github.com>
103 lines
4.0 KiB
C++
103 lines
4.0 KiB
C++
/*
|
|
* SPDX-FileCopyrightText: Copyright (c) 1993-2024 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
|
|
* SPDX-License-Identifier: Apache-2.0
|
|
*
|
|
* 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.
|
|
*/
|
|
|
|
#ifndef TRT_CUMSUM_LAST_DIM_PLUGIN_H
|
|
#define TRT_CUMSUM_LAST_DIM_PLUGIN_H
|
|
|
|
#include "tensorrt_llm/kernels/cumsumLastDim.h"
|
|
#include "tensorrt_llm/plugins/common/plugin.h"
|
|
#include <cassert>
|
|
|
|
namespace tensorrt_llm::plugins
|
|
{
|
|
class CumsumLastDimPlugin : public BasePlugin
|
|
{
|
|
public:
|
|
using SizeType32 = tensorrt_llm::kernels::SizeType32;
|
|
|
|
CumsumLastDimPlugin(SizeType32 inputLength, nvinfer1::DataType type, size_t tempStorageBytes = 0);
|
|
CumsumLastDimPlugin(void const* data, size_t length);
|
|
~CumsumLastDimPlugin() override = default;
|
|
// IPluginV2DynamicExt Methods
|
|
nvinfer1::IPluginV2DynamicExt* clone() const noexcept override;
|
|
nvinfer1::DimsExprs getOutputDimensions(int outputIndex, nvinfer1::DimsExprs const* inputs, int nbInputs,
|
|
nvinfer1::IExprBuilder& exprBuilder) noexcept override;
|
|
bool supportsFormatCombination(
|
|
int pos, nvinfer1::PluginTensorDesc const* inOut, int nbInputs, int nbOutputs) noexcept override;
|
|
void configurePlugin(nvinfer1::DynamicPluginTensorDesc const* in, int nbInputs,
|
|
nvinfer1::DynamicPluginTensorDesc const* out, int nbOutputs) noexcept override;
|
|
size_t getWorkspaceSize(nvinfer1::PluginTensorDesc const* inputs, int nbInputs,
|
|
nvinfer1::PluginTensorDesc const* outputs, int nbOutputs) const noexcept override;
|
|
int enqueue(nvinfer1::PluginTensorDesc const* inputDesc, nvinfer1::PluginTensorDesc const* outputDesc,
|
|
void const* const* inputs, void* const* outputs, void* workspace, cudaStream_t stream) noexcept override;
|
|
template <typename T>
|
|
int enqueueImpl(nvinfer1::PluginTensorDesc const* inputDesc, nvinfer1::PluginTensorDesc const* outputDesc,
|
|
void const* const* inputs, void* const* outputs, void* workspace, cudaStream_t stream);
|
|
size_t getWorkspaceSizeNeeded(SizeType32 inputLength, nvinfer1::DataType type);
|
|
|
|
// IPluginV2Ext Methods
|
|
nvinfer1::DataType getOutputDataType(
|
|
int index, nvinfer1::DataType const* inputTypes, int nbInputs) const noexcept override;
|
|
|
|
// IPluginV2 Methods
|
|
char const* getPluginType() const noexcept override;
|
|
char const* getPluginVersion() const noexcept override;
|
|
int getNbOutputs() const noexcept override;
|
|
int initialize() noexcept override;
|
|
void terminate() noexcept override;
|
|
size_t getSerializationSize() const noexcept override;
|
|
void serialize(void* buffer) const noexcept override;
|
|
void destroy() noexcept override;
|
|
|
|
private:
|
|
using IndexType = std::int32_t;
|
|
|
|
IndexType getInputTensorIdx() const
|
|
{
|
|
return 0;
|
|
};
|
|
|
|
private:
|
|
SizeType32 mInputLength;
|
|
size_t mTempStorageBytes;
|
|
nvinfer1::DataType mType;
|
|
};
|
|
|
|
class CumsumLastDimPluginCreator : public BaseCreator
|
|
{
|
|
public:
|
|
CumsumLastDimPluginCreator();
|
|
char const* getPluginName() const noexcept override;
|
|
|
|
char const* getPluginVersion() const noexcept override;
|
|
|
|
nvinfer1::PluginFieldCollection const* getFieldNames() noexcept override;
|
|
|
|
nvinfer1::IPluginV2* createPlugin(char const* name, nvinfer1::PluginFieldCollection const* fc) noexcept override;
|
|
|
|
nvinfer1::IPluginV2* deserializePlugin(
|
|
char const* name, void const* serialData, size_t serialLength) noexcept override;
|
|
|
|
private:
|
|
static nvinfer1::PluginFieldCollection mFC;
|
|
static std::vector<nvinfer1::PluginField> mPluginAttributes;
|
|
};
|
|
|
|
} // namespace tensorrt_llm::plugins
|
|
|
|
#endif
|