mirror of
https://github.com/NVIDIA/TensorRT-LLM.git
synced 2026-01-14 06:27:45 +08:00
c89653021e
* Update TensorRT-LLM --------- Co-authored-by: Eddie-Wang1120 <81598289+Eddie-Wang1120@users.noreply.github.com> Co-authored-by: Shixiaowei02 <39303645+Shixiaowei02@users.noreply.github.com>
172 lines
10 KiB
C++
172 lines
10 KiB
C++
/*
|
|
* Copyright (c) 2020-2023, NVIDIA CORPORATION. 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 "tensorrt_llm/kernels/decodingCommon.h"
|
|
#include <curand_kernel.h>
|
|
|
|
namespace tensorrt_llm
|
|
{
|
|
namespace kernels
|
|
{
|
|
|
|
//! \brief Initialize buffers for topP inference
|
|
//!
|
|
//! \param topPIdValBuf output buffer [batchSize, vocabSize]. Value at {bi,vi} position contains vi token id.
|
|
//! \param topPOffsetBuf output buffer [batchSize+1].
|
|
//! \param beginTopPOffsetBuf output buffer [batchSize+1].
|
|
//! \param batchSize number of requests in batch
|
|
//! \param vocabSize size of the inner dimension
|
|
//! \param stream stream
|
|
void invokeTopPInitialize(int* topPIdValBuf, int* topPOffsetBuf, int* beginTopPOffsetBuf, size_t const batchSize,
|
|
int const vocabSize, cudaStream_t stream);
|
|
|
|
//! \brief Given logProbs, performs top P sampling. Fills sampled tokens to outputIds.
|
|
//! Computes sequenceLength, finished state, cumLogProbs inplace.
|
|
//! Sampling per request can be controlled using skipDecode and topPs parameters.
|
|
//! Function sets workspaceSize and cubTempStorageSize and exits early if workspace is nullptr.
|
|
//!
|
|
//! \param workspace pointer to the workspace. Has to be pre-allocated by caller. Function does not take ownership of
|
|
//! the buffer.
|
|
//! \param workspaceSize size of the workspace in bytes.
|
|
//! \param cubTempStorageSize workspace size for cub radix sort.
|
|
//! \param outputIds output buffer [batchSize][maxSeqLen]. Contains pointers to rows with output tokens per request.
|
|
//! \param sequenceLength input/output buffer [batchSize]. Current sequence length of the request up to, but excluding
|
|
//! endId token.
|
|
//! \param finishedInput input buffer [batchSize]. Exit early if true.
|
|
//! \param finishedOutput output buffer [batchSize]. Set flag if sequence has finished (if finished || outputId ==
|
|
//! endId).
|
|
//! \param cumLogProbs input/output buffer [batchSize]. Cumulative log probability of selected tokens. Ignored if
|
|
//! nullptr.
|
|
//! \param outputLogProbs output buffer [batchSize]. Log probs is the probability
|
|
//! induced by the top-k sampling. We normalize the probability 'expLogit' of the selected token by the probability
|
|
//! 's_sum' of a set of top-k tokens, meaning the logProb is the probability of the selected token, conditioned on the
|
|
//! event that it is selected, i.e., log_prob = log P(i | i is in top-k) = log(expLogit / s_sum). Ignored if nullptr.
|
|
//! \param logProbs input buffer [batchSize x vocabSizePadded]. Log probabilities of each token in the vocab.
|
|
//! If cumLogProbs or outputLogProbs are specified, logProbs must contain **just** probabilities instead of log
|
|
//! probabilities.
|
|
//! \param idVals input buffer [batchSize, vocabSize]. Value at {bi,vi} position contains vi token id.
|
|
//! Initialized using invokeTopPInitialize.
|
|
//! \param offsetBuf input buffer [batchSize+1]. Array of offsets initialized using invokeTopPInitialize.
|
|
//! \param beginOffsetBuf input buffer [batchSize+1]. Array of offsets initialized using invokeTopPInitialize.
|
|
//! \param curandstate input buffer [batchSize]. Curand states properly initialized using
|
|
//! invokeCurandInitialize per request.
|
|
//! \param batchSize batch size
|
|
//! \param vocabSizePadded size of padded vocab
|
|
//! \param endIds input buffer [batchSize]. EOS token ids per request
|
|
//! \param maxTopP maximum among all topPs P for topP sampling
|
|
//! \param topPs input buffer [batchSize]. P for topP sampling per request. Supported P is in range (0.0; 1.0].
|
|
//! If nullptr maxTopP is used for all requests.
|
|
//! \param stream cuda stream
|
|
//! \param cudaDeviceProp
|
|
//! \param skipDecode input buffer [batchSize]. Flags whether to skip decoding per request
|
|
|
|
template <typename T>
|
|
void invokeBatchTopPSampling(void* workspace, size_t& workspaceSize, size_t& cubTempStorageSize, int** outputIds,
|
|
int* sequenceLength, FinishedState const* finishedInput, FinishedState* finishedOutput, float* cumLogProbs,
|
|
float* outputLogProbs, T const* logProbs, int const* idVals, int* offsetBuf, int* beginOffsetBuf,
|
|
curandState_t* curandstate, int const batchSize, size_t const vocabSizePadded, int const* endIds,
|
|
float const maxTopP, float const* topPs, cudaStream_t stream, bool const* skipDecode);
|
|
|
|
//! \brief Specialization of invokeBatchTopPSampling with topPs=nullptr
|
|
template <typename T>
|
|
void invokeTopPSampling(void* workspace, size_t& workspaceSize, size_t& cubTempStorageSize, int** outputIds,
|
|
int* sequenceLength, FinishedState const* finishedInput, FinishedState* finishedOutput, float* cumLogProbs,
|
|
float* outputLogProbs, T const* logProbs, int const* idVals, int* offsetBuf, int* beginOffsetBuf,
|
|
curandState_t* curandstate, int const batchSize, size_t const vocabSizePadded, int const* endIds, float const topPp,
|
|
cudaStream_t stream, bool const* skipDecode);
|
|
|
|
//! \brief Given logProbs, performs top P sampling.
|
|
//! Note different from invokeTopPSampling() and invokeBatchTopPSampling() there two functions invokeAirTopPSampling
|
|
//! and invokeBatchAirTopPSampling is non-deterministic.
|
|
//! Fills sampled tokens to outputIds. Computes sequenceLength, finished state, cumLogProbs inplace.
|
|
//! Sampling per request can be controlled using skipDecode and topPs parameters.
|
|
//! Function sets workspaceSize and exits early if workspace is nullptr.
|
|
//!
|
|
//! \param workspace pointer to the workspace. Has to be pre-allocated by caller. Function does not take ownership of
|
|
//! the buffer.
|
|
//! \param workspaceSize size of the workspace in bytes.
|
|
//! \param outputIds output buffer [batchSize][maxSeqLen]. Contains pointers to rows with output tokens per request.
|
|
//! \param sequenceLength input/output buffer [batchSize]. Current sequence length of the request up to, but excluding
|
|
//! endId token.
|
|
//! \param finishedInput input buffer[batchSize].Exit early if true.
|
|
//! \param finishedOutput output buffer [batchSize]. Set flag if sequence has finished (if finished || outputId ==
|
|
//! endId).
|
|
//! \param cumLogProbs input/output buffer [batchSize]. Cumulative log probability of selected tokens. Ignored
|
|
//! if nullptr.
|
|
//! \param outputLogProbs output buffer [batchSize]. Log probs is the probability induced by the top-k
|
|
//! sampling. We normalize the probability 'expLogit' of the selected token by the probability 's_sum' of a set of top-k
|
|
//! tokens, meaning the logProb is the probability of the selected token, conditioned on the event that it is selected,
|
|
//! i.e., log_prob = log P(i | i is in top-k) = log(expLogit / s_sum). Ignored if nullptr.
|
|
//! \param logProbs input buffer [batchSize x vocabSizePadded]. Log probabilities of each token in the vocab.
|
|
//! If cumLogProbs or outputLogProbs are specified, logProbs must contain **just** probabilities instead of log
|
|
//! probabilities.
|
|
//! \param curandstate input buffer [batchSize]. Curand states properly initialized using invokeCurandInitialize per
|
|
//! request.
|
|
//! \param batchSize batch size
|
|
//! \param vocabSizePadded size of padded vocab
|
|
//! \param endIds input buffer [batchSize]. EOS token ids per request
|
|
//! \param maxTopP maximum among all topPs P for topP sampling
|
|
//! \param topPs input buffer [batchSize]. P for topP sampling per request. Supported P is in range (0.0; 1.0].
|
|
//! If nullptr maxTopP is used for all requests.
|
|
//! \param stream cuda stream
|
|
//! \param blockNum The appropriate block configuration calculated based on the number of multiprocessors, occupancy,
|
|
//! batchSize and vocabSizePadded
|
|
//! \param skipDecode input buffer [batchSize]. Flags whether to skip decoding per request
|
|
template <typename T>
|
|
void invokeBatchAirTopPSampling(void* workspace, size_t& workspaceSize, int** outputIds, int* sequenceLength,
|
|
FinishedState const* finishedInput, FinishedState* finishedOutput, float* cumLogProbs, float* outputLogProbs,
|
|
T const* logProbs, curandState_t* curandstate, int const batchSize, size_t const vocabSizePadded, int const* endIds,
|
|
float const maxTopP, float const* topPs, cudaStream_t stream, int blockNum, bool const* skipDecode);
|
|
|
|
//! \brief Specialization of invokeBatchAirTopPSampling with topPs=nullptr
|
|
template <typename T>
|
|
void invokeAirTopPSampling(void* workspace, size_t& workspaceSize, int** outputIds, int* sequenceLength,
|
|
FinishedState const* finishedInput, FinishedState* finishedOutput, float* cumLogProbs, float* outputLogProbs,
|
|
T const* logProbs, curandState_t* curandstate, int const batchSize, size_t const vocabSizePadded, int const* endIds,
|
|
float const topP, cudaStream_t stream, int blockNum, bool const* skipDecode);
|
|
|
|
//! \brief Calculate the number of blocks based on the number of multiprocessors, batchSize and vocabSize.
|
|
//! \tparam T the data type of value
|
|
//! \tparam IdxT the data type of index
|
|
//! \tparam AccT the data type of variables related to accumulation
|
|
//! \tparam BitsPerPass the number of bits for each pass. Can be 8 or 11. Use 11 for default.
|
|
//! \tparam BlockSize the block size
|
|
//! \param batchSize
|
|
//! \param len the number of candidates for each case
|
|
//! \param smCnt number of multiprocessors on device
|
|
template <typename T, typename IdxT, typename AccT, int BitsPerPass = 11, int BlockSize = 512>
|
|
unsigned calcAirTopPBlockNum(int batchSize, IdxT len, int smCnt);
|
|
|
|
//! \brief Compute the topp decay by https://arxiv.org/pdf/2206.04624.pdf
|
|
//! In short, the formula is
|
|
//! runtimeTopP = max(runtimeTopP * topPDecay, topPMin)
|
|
//! If generating the topPResetIds, then reset the runtimeTopP.
|
|
//!
|
|
//! \param runtimeTopP
|
|
//! \param runtimeInitialTopP
|
|
//! \param outputIds
|
|
//! \param topPDecay
|
|
//! \param topPMin
|
|
//! \param topPResetIds
|
|
//! \param localBatchSize
|
|
void invokeComputeToppDecay(float* runtimeTopP, float const* runtimeInitialTopP, int const** outputIds,
|
|
float const* topPDecay, float const* topPMin, int32_t const* topPResetIds, int const* sequenceLengths,
|
|
int const localBatchSize, cudaStream_t stream);
|
|
|
|
} // namespace kernels
|
|
} // namespace tensorrt_llm
|