TensorRT-LLMs/cpp/tensorrt_llm/layers/baseSamplingLayer.cpp
Kaiyu Xie 0f041b7b57
Update TensorRT-LLM (#1098)
* Update TensorRT-LLM

* update submodule

* Remove unused binaries
2024-02-18 15:48:08 +08:00

50 lines
1.6 KiB
C++

/*
* Copyright (c) 2019-2024, NVIDIA CORPORATION. All rights reserved.
* Copyright (c) 2021, NAVER Corp. Authored by CLOVA.
*
* 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.
*/
#include "tensorrt_llm/layers/baseSamplingLayer.h"
#include "tensorrt_llm/common/cudaUtils.h"
#include "tensorrt_llm/common/memoryUtils.h"
#include "tensorrt_llm/kernels/decodingCommon.h"
#include "tensorrt_llm/kernels/penaltyKernels.h"
#include "tensorrt_llm/kernels/samplingTopKKernels.h"
#include "tensorrt_llm/layers/fillBuffers.h"
#include <algorithm>
using namespace tensorrt_llm::common;
using namespace tensorrt_llm::kernels;
namespace tensorrt_llm
{
namespace layers
{
template <typename T>
BaseSamplingLayer<T>::BaseSamplingLayer(size_t maxBatchSize, size_t vocabSize, size_t vocabSizePadded,
cudaStream_t stream, std::shared_ptr<IAllocator> allocator, cudaDeviceProp* prop)
: BaseLayer(stream, std::move(allocator), prop)
, mMaxBatchSize(maxBatchSize)
, mVocabSize(vocabSize)
, mVocabSizePadded(vocabSizePadded)
{
}
template class BaseSamplingLayer<float>;
template class BaseSamplingLayer<half>;
} // namespace layers
} // namespace tensorrt_llm