kanshan/TensorRT-LLMs
1
0
Fork 0
mirror of https://github.com/NVIDIA/TensorRT-LLM.git synced 2026-01-14 06:27:45 +08:00
Code Issues Actions 1 Packages Projects Releases Wiki Activity
78f5c2936b
TensorRT-LLMs/cpp/tensorrt_llm/runtime/gptDecoder.cpp
Kaiyu Xie 78f5c2936b
Update TensorRT-LLM (#2184)
2024-09-03 12:14:23 +02:00

794 lines
34 KiB
C++
Raw Blame History

/*
* Copyright (c) 2022-2024, 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.
*/
#include "tensorrt_llm/runtime/gptDecoder.h"
#include "tensorrt_llm/kernels/decodingKernels.h"
#include "tensorrt_llm/kernels/speculativeDecoding/externalDraftTokensKernels.h"
#include "tensorrt_llm/layers/decodingParams.h"
#include "tensorrt_llm/layers/dynamicDecodeLayer.h"
#include "tensorrt_llm/runtime/decodingLayerWorkspace.h"
#include <memory>
#include <NvInferRuntime.h>
namespace tle = tensorrt_llm::executor;
namespace tl = tensorrt_llm::layers;
namespace tksd = tensorrt_llm::kernels::speculative_decoding;
using namespace tensorrt_llm::runtime;
using BufferConstPtr = IBuffer::SharedConstPtr;
using BufferPtr = IBuffer::SharedPtr;
using TensorConstPtr = ITensor::SharedConstPtr;
using TensorPtr = ITensor::SharedPtr;
template <typename T>
GptDecoder<T>::GptDecoder(executor::DecodingMode const& mode, size_t maxBatchSize, size_t maxBeamWidth,
size_t vocabSize, size_t vocabSizePadded, size_t maxSequenceLength, CudaStreamPtr const& stream,
std::shared_ptr<SpeculativeDecodingModule const> speculativeDecodingModule)
: mManager{std::make_shared<BufferManager>(stream)}
, mMaxBatchSize(maxBatchSize)
, mDecodingMode{mode}
{
TLLM_LOG_TRACE("%s start", __PRETTY_FUNCTION__);
auto const decodingDomain = tensorrt_llm::layers::DecoderDomain(
maxBatchSize, maxBeamWidth, vocabSize, vocabSizePadded, speculativeDecodingModule);
mDynamicDecodeLayer = std::make_shared<tensorrt_llm::layers::DynamicDecodeLayer<T>>(mode, decodingDomain, mManager);
auto constexpr nvFloatType = TRTDataType<float>::value;
mLogProbsTiled = mManager->gpu(ITensor::makeShape({static_cast<SizeType32>(maxSequenceLength),
static_cast<SizeType32>(maxBatchSize), static_cast<SizeType32>(maxBeamWidth)}),
nvFloatType);
mManager->setZero(*mLogProbsTiled);
mDecodingLayerWorkspace = std::make_unique<tensorrt_llm::runtime::DecodingLayerWorkspace>(
mManager, decodingDomain, TRTDataType<T>::value, mDynamicDecodeLayer->getWorkspaceSize());
TLLM_LOG_TRACE("%s stop", __PRETTY_FUNCTION__);
}
template <typename T>
void GptDecoder<T>::setup(SamplingConfig const& samplingConfig, size_t batchSize, TensorConstPtr const& batchSlots,
std::optional<DecodingOutput> const& output,
std::optional<std::vector<decoder_batch::Request> const> const& requestsOpt)
{
TLLM_LOG_TRACE("%s start", __PRETTY_FUNCTION__);
mSamplingConfig = samplingConfig;
auto setupParams = std::make_shared<layers::DynamicDecodeSetupParams>();
TLLM_CHECK_WITH_INFO(mSamplingConfig.validate(), "Sampling config is invalid");
TLLM_CHECK_WITH_INFO(batchSlots != nullptr, "Batch slots are mandatory to set up the decoder.");
auto penaltyParams = std::make_shared<tl::PenaltySetupParams>();
penaltyParams->repetitionPenalty = mSamplingConfig.repetitionPenalty;
penaltyParams->presencePenalty = mSamplingConfig.presencePenalty;
penaltyParams->frequencyPenalty = mSamplingConfig.frequencyPenalty;
penaltyParams->temperature = mSamplingConfig.temperature;
penaltyParams->minLength = mSamplingConfig.minLength;
setupParams->penaltyParams = std::move(penaltyParams);
auto banWordsParams = std::make_shared<tl::BanWordsSetupParams>();
banWordsParams->noRepeatNgramSize = mSamplingConfig.noRepeatNgramSize;
setupParams->banWordsParams = std::move(banWordsParams);
if (mDecodingMode.isTopKorTopP())
{
auto samplingParams = std::make_shared<tl::SamplingSetupParams>();
samplingParams->normalizeLogProbs = mSamplingConfig.normalizeLogProbs;
// signed to unsigned
if (mSamplingConfig.topK)
{
auto const& topK = mSamplingConfig.topK.value();
samplingParams->runtimeTopK = std::vector<SizeType32>(std::begin(topK), std::end(topK));
}
samplingParams->runtimeTopP = mSamplingConfig.topP;
samplingParams->topPDecay = mSamplingConfig.topPDecay;
samplingParams->topPMin = mSamplingConfig.topPMin;
samplingParams->topPResetIds = mSamplingConfig.topPResetIds;
samplingParams->outputLogProbs = mSamplingConfig.outputLogProbs;
samplingParams->cumLogProbs = mSamplingConfig.cumLogProbs;
setupParams->decodingParams = std::move(samplingParams);
}
else if (mDecodingMode.isBeamSearch())
{
auto beamSearchParams = std::make_shared<tl::BeamSearchSetupParams>();
beamSearchParams->beamSearchDiversityRate = mSamplingConfig.beamSearchDiversityRate;
beamSearchParams->lengthPenalty = mSamplingConfig.lengthPenalty;
beamSearchParams->earlyStopping = mSamplingConfig.earlyStopping;
setupParams->decodingParams = std::move(beamSearchParams);
}
else if (mDecodingMode.isMedusa())
{
auto medusaParams = std::make_shared<tl::MedusaSetupParams>();
// signed to unsigned
if (mSamplingConfig.topK)
{
auto const& topK = mSamplingConfig.topK.value();
medusaParams->runtimeTopK = std::vector<SizeType32>(std::begin(topK), std::end(topK));
}
medusaParams->runtimeHeadsTopK = mSamplingConfig.topKMedusaHeads;
setupParams->decodingParams = std::move(medusaParams);
}
else if (mDecodingMode.isExplicitDraftTokens())
{
TLLM_CHECK_WITH_INFO(output.has_value(), "Output tensors must be provided for ExplicitDraftTokens");
auto explicitDraftTokensParams = std::make_shared<tl::ExplicitDraftTokensSetupParams>();
explicitDraftTokensParams->temperature = mSamplingConfig.temperature;
explicitDraftTokensParams->randomDataSample = output->explicitDraftTokensBuffers->randomDataSample;
explicitDraftTokensParams->temperatures = output->explicitDraftTokensBuffers->temperatures;
TLLM_CHECK(requestsOpt);
// Ignore the dtype from all other requests assuming that it is the same for all.
explicitDraftTokensParams->dtype = requestsOpt.value()[0].dtype;
setupParams->decodingParams = explicitDraftTokensParams;
}
else if (mDecodingMode.isLookahead())
{
TLLM_CHECK_WITH_INFO(output.has_value(), "Output tensors must be provided for Lookahead decoding");
TLLM_LOG_DEBUG("gptDecoder setup lookahead, batchSize=%d", batchSize);
auto lookaheadParams = std::make_shared<tl::LookaheadSetupParams>();
TLLM_CHECK(requestsOpt);
auto& requests = requestsOpt.value();
lookaheadParams->prompt.resize(0);
lookaheadParams->prompt.reserve(batchSize);
lookaheadParams->algoConfigs.resize(0);
lookaheadParams->algoConfigs.reserve(batchSize);
for (size_t bi = 0; bi < batchSize; bi++)
{
lookaheadParams->prompt.emplace_back(ITensor::slice(requests[bi].ids, 0, requests[bi].inputLen));
TLLM_CHECK(requests[bi].lookaheadRuntimeConfig);
lookaheadParams->algoConfigs.emplace_back(requests[bi].lookaheadRuntimeConfig.value());
}
lookaheadParams->generationLengths = output->lookaheadOutputs->generationLengths;
lookaheadParams->positionOffsets = output->lookaheadOutputs->positionOffsets;
lookaheadParams->attentionPackedMasks = output->lookaheadOutputs->packedMasks;
setupParams->decodingParams = std::move(lookaheadParams);
}
setupParams->decodingParams->randomSeed = mSamplingConfig.randomSeed;
mDecodingLayerWorkspace->setDeviceBatchSlots(batchSlots);
mDynamicDecodeLayer->setup(batchSize, mSamplingConfig.beamWidth, batchSlots, setupParams, mDecodingLayerWorkspace);
TLLM_LOG_TRACE("%s stop", __PRETTY_FUNCTION__);
}
namespace
{
std::shared_ptr<tl::BanWordsDecodingInputs> prepareBanWordsInputs(DecodingInput const& input)
{
auto banWordsParams = std::make_shared<tl::BanWordsDecodingInputs>(input.batchSize);
if (input.badWordsPtrs)
{
TLLM_CHECK_WITH_INFO(input.badWordsPtrs, "Bad word lengths must be provided when badWordsPtrs is given");
banWordsParams->badWordsPtr = input.badWordsPtrs;
banWordsParams->badWordsLengths = input.badWordsLens;
banWordsParams->maxBadWordsLen = input.maxBadWordsLen;
}
return banWordsParams;
}
std::shared_ptr<tl::StopCriteriaDecodingInputs> prepareStopCriteriaInputs(DecodingInput const& input)
{
auto stopCriteriaParams = std::make_shared<tl::StopCriteriaDecodingInputs>(input.batchSize);
if (input.stopWordsPtrs)
{
TLLM_CHECK_WITH_INFO(input.stopWordsLens, "Stop word lengths must be provided when stopWordsPtrs is given");
stopCriteriaParams->stopWordsPtr = input.stopWordsPtrs;
stopCriteriaParams->stopWordsLengths = input.stopWordsLens;
stopCriteriaParams->maxStopWordsLen = input.maxStopWordsLen;
}
if (input.sequenceLimitLength)
{
stopCriteriaParams->sequenceLimitLength = input.sequenceLimitLength;
}
return stopCriteriaParams;
}
void prepareMedusaInputs(
DecodingInput const& inputs, size_t maxBatchSize, std::shared_ptr<tl::DecodingInputs>& baseInputs)
{
TLLM_LOG_TRACE("%s start", __PRETTY_FUNCTION__);
auto inputParams = std::dynamic_pointer_cast<tl::MedusaDecodingInputs>(baseInputs);
auto const& medusaInputs = inputs.medusaInputs.value();
inputParams->curTokensPerStep = medusaInputs.medusaCurTokensPerStep;
inputParams->targetTokensPerStep = medusaInputs.medusaTargetTokensPerStep;
inputParams->paths = medusaInputs.medusaPaths;
inputParams->treeIds = medusaInputs.medusaTreeIds;
auto const batchSlots = bufferCast<SizeType32>(*inputs.batchSlots);
if (medusaInputs.medusaLogits.size())
{
std::vector<std::vector<TensorPtr>> medusaLogits;
auto const batchSize = medusaInputs.medusaLogits.size();
medusaLogits.resize(maxBatchSize);
for (size_t bi = 0; bi < batchSize; ++bi)
{
auto const slot = batchSlots[bi];
auto const& logitsHeads = medusaInputs.medusaLogits.at(slot);
auto const medusaHeads = logitsHeads.size();
medusaLogits[slot].resize(medusaHeads);
for (size_t hi = 0; hi < medusaHeads; ++hi)
{
if (logitsHeads[hi])
{
medusaLogits[slot][hi] = logitsHeads[hi];
}
}
}
inputParams->medusaLogits = medusaLogits;
}
TLLM_LOG_TRACE("%s stop", __PRETTY_FUNCTION__);
}
void prepareExplicitDraftTokensInput(DecodingInput const& inputs, std::shared_ptr<tl::DecodingInputs>& baseInputs)
{
TLLM_LOG_TRACE("%s start", __PRETTY_FUNCTION__);
auto inputParams = std::dynamic_pointer_cast<tl::ExplicitDraftTokensInputs>(baseInputs);
auto& explicitDraftTokensInputs = inputs.explicitDraftTokensInputs;
TLLM_CHECK_WITH_INFO(explicitDraftTokensInputs.has_value(), "ExplicitDraftTokensInputs are not set");
inputParams->nextDraftTokens = explicitDraftTokensInputs->nextDraftTokens;
inputParams->nextFlatTokens = explicitDraftTokensInputs->nextFlatTokens;
inputParams->nextDraftIndices = explicitDraftTokensInputs->nextDraftIndices;
inputParams->nextDraftProbs = explicitDraftTokensInputs->nextDraftProbs;
inputParams->lastDraftTokens = explicitDraftTokensInputs->lastDraftTokens;
inputParams->lastDraftIndices = explicitDraftTokensInputs->lastDraftIndices;
inputParams->masks = explicitDraftTokensInputs->masks;
inputParams->packedPosIds = explicitDraftTokensInputs->packedPositionIds;
inputParams->bestPathLengths = explicitDraftTokensInputs->bestPathLengths;
inputParams->bestPathIndices = explicitDraftTokensInputs->bestPathIndices;
inputParams->generationLengths = explicitDraftTokensInputs->nextGenerationLengths;
inputParams->positionIdsBase = explicitDraftTokensInputs->lastPositionIdsBase;
inputParams->lastGenerationLengths = explicitDraftTokensInputs->lastGenerationLengths;
inputParams->maxGenLengthDevice = explicitDraftTokensInputs->maxGenLengthDevice;
inputParams->seqSlots = explicitDraftTokensInputs->seqSlots;
TLLM_LOG_TRACE("%s stop", __PRETTY_FUNCTION__);
}
void prepareLookaheadInputs(
DecodingInput const& inputs, size_t maxBatchSize, std::shared_ptr<tl::DecodingInputs>& baseInputs)
{
TLLM_LOG_TRACE("%s start", __PRETTY_FUNCTION__);
auto inputParams = std::dynamic_pointer_cast<tl::LookaheadDecodingInputs>(baseInputs);
auto const& lookaheadInputs = inputs.lookaheadInputs.value();
inputParams->curTokensPerStep = lookaheadInputs.tokensPerStep;
TLLM_LOG_TRACE("%s stop", __PRETTY_FUNCTION__);
}
template <typename T>
std::shared_ptr<tl::BaseDecodingInputs> prepareInputs(
DecodingInput const& input, size_t maxBatchSize, tle::DecodingMode const& decodingMode)
{
auto constexpr ite = 0;
TLLM_CHECK_WITH_INFO(input.batchSlots != nullptr, "Batch slots are mandatory to call the decoder.");
std::shared_ptr<tl::DecodingInputs> forwardParams;
if (decodingMode.isTopKorTopP())
{
forwardParams
= std::make_shared<tl::SamplingInputs>(input.endIds, input.batchSlots, input.step, ite, input.batchSize);
}
else if (decodingMode.isBeamSearch())
{
forwardParams = std::make_shared<tl::DecodingInputs>(input.endIds, input.batchSlots, input.step, ite,
input.batchSize, input.maxAttentionWindow, input.sinkTokenLength);
}
else if (decodingMode.isMedusa())
{
forwardParams = std::make_shared<tl::MedusaDecodingInputs>(input.endIds, input.batchSlots, input.batchSize);
}
else if (decodingMode.isLookahead())
{
forwardParams = std::make_shared<tl::LookaheadDecodingInputs>(input.endIds, input.batchSlots);
}
else if (decodingMode.isExplicitDraftTokens())
{
forwardParams
= std::make_shared<tl::ExplicitDraftTokensInputs>(input.endIds, input.batchSlots, input.batchSize);
}
// No logits for explicit draft tokens
if (!decodingMode.isExplicitDraftTokens())
{
if (input.logitsVec)
{
std::vector<TensorConstPtr> logitsVec;
for (auto const& logits : input.logitsVec.value())
{
TLLM_CHECK(logits->getDataType() == TRTDataType<T>::value);
logitsVec.push_back(logits);
}
forwardParams->logitsVec = logitsVec;
}
else if (input.logits)
{
TLLM_CHECK(input.logits->getDataType() == TRTDataType<T>::value);
forwardParams->logits = input.logits;
}
}
if (input.cacheIndirection)
{
forwardParams->srcCacheIndirection = input.cacheIndirection;
}
if (input.embeddingBias)
{
forwardParams->embeddingBias = input.embeddingBias;
}
if (input.lengths)
{
forwardParams->inputLengths = input.lengths;
}
forwardParams->banWordsInputs = prepareBanWordsInputs(input);
forwardParams->stopCriteriaInputs = prepareStopCriteriaInputs(input);
if (input.finishReasons)
{
forwardParams->finished = input.finishReasons;
}
// Medusa
if (decodingMode.isMedusa())
{
prepareMedusaInputs(input, maxBatchSize, forwardParams);
}
// Explicit draft tokens
if (decodingMode.isExplicitDraftTokens())
{
prepareExplicitDraftTokensInput(input, forwardParams);
}
if (input.lookaheadInputs)
{
prepareLookaheadInputs(input, maxBatchSize, forwardParams);
forwardParams->localBatchSize = input.batchSize;
}
TLLM_LOG_TRACE("%s stop", __PRETTY_FUNCTION__);
return forwardParams;
}
void prepareBeamSearchOutputs(DecodingOutput& output, std::shared_ptr<tl::BaseDecodingOutputs>& baseOutputs)
{
TLLM_LOG_TRACE("%s start", __PRETTY_FUNCTION__);
auto outputParams = std::dynamic_pointer_cast<tl::BeamSearchOutputs>(baseOutputs);
outputParams->beamHypotheses = std::make_unique<tensorrt_llm::kernels::BeamHypotheses>();
if (output.beamHypotheses.outputIdsCBA)
{
outputParams->beamHypotheses->outputIdsCBA = bufferCast<int>(*output.beamHypotheses.outputIdsCBA);
}
if (output.beamHypotheses.logProbsCBA)
{
outputParams->beamHypotheses->logProbsCBA = bufferCast<float>(*output.beamHypotheses.logProbsCBA);
}
if (output.beamHypotheses.sequenceLengthsCBA)
{
outputParams->beamHypotheses->sequenceLengthsCBA = bufferCast<int>(*output.beamHypotheses.sequenceLengthsCBA);
}
if (output.beamHypotheses.cumLogProbsCBA)
{
outputParams->beamHypotheses->cumLogProbsCBA = bufferCast<float>(*output.beamHypotheses.cumLogProbsCBA);
}
if (output.beamHypotheses.normedScoresCBA)
{
outputParams->beamHypotheses->normedScoresCBA = bufferCast<float>(*output.beamHypotheses.normedScoresCBA);
}
if (output.beamHypotheses.numBeamsCBA)
{
outputParams->beamHypotheses->numBeamsCBA = bufferCast<int>(*output.beamHypotheses.numBeamsCBA);
}
if (output.beamHypotheses.minNormedScoresCBA)
{
outputParams->beamHypotheses->minNormedScoresCBA = bufferCast<float>(*output.beamHypotheses.minNormedScoresCBA);
}
if (output.beamHypotheses.batchDones)
{
outputParams->beamHypotheses->batchDones = bufferCast<bool>(*output.beamHypotheses.batchDones);
}
if (output.cacheIndirection)
{
outputParams->tgtCacheIndirection = output.cacheIndirection;
}
TLLM_LOG_TRACE("%s stop", __PRETTY_FUNCTION__);
}
void prepareSpeculativeDecodingOutputs(DecodingOutput& output, std::shared_ptr<tl::BaseDecodingOutputs>& baseOutputs,
tle::DecodingMode const& decodingMode)
{
TLLM_LOG_TRACE("%s start", __PRETTY_FUNCTION__);
auto outputParams = std::dynamic_pointer_cast<tl::SpeculativeDecodingOutputs>(baseOutputs);
auto const& speculativeDecodingOutputs = output.speculativeDecodingOutputs;
TLLM_CHECK_WITH_INFO(speculativeDecodingOutputs.has_value(), "speculativeDecodingOutputs is not set");
outputParams->nextDraftTokens = speculativeDecodingOutputs->nextDraftTokens;
outputParams->numNewTokens = speculativeDecodingOutputs->acceptedTokensLen;
outputParams->numNewTokensCumSum = speculativeDecodingOutputs->acceptedLengthsCumSum;
outputParams->pathsOffsets = speculativeDecodingOutputs->pathsOffsets;
if (speculativeDecodingOutputs->nextDraftTokensLen)
{
outputParams->nextDraftLengths = speculativeDecodingOutputs->nextDraftTokensLen;
}
if (speculativeDecodingOutputs->prevDraftTokensLen)
{
outputParams->prevDraftLengths = speculativeDecodingOutputs->prevDraftTokensLen;
}
if (decodingMode.isExplicitDraftTokens())
{
auto outputParams = std::dynamic_pointer_cast<tl::ExplicitDraftTokensOutputs>(baseOutputs);
auto const& explicitDraftTokensBuffers = output.explicitDraftTokensBuffers;
TLLM_CHECK_WITH_INFO(explicitDraftTokensBuffers.has_value(), "explicitDraftTokensBuffers is not set");
outputParams->packedMasks = explicitDraftTokensBuffers->packedMasks;
outputParams->nextDraftPosIds = explicitDraftTokensBuffers->positionIds;
outputParams->unpackedNextDraftTokens = explicitDraftTokensBuffers->draftTokens;
outputParams->unpackedNextDraftIndices = explicitDraftTokensBuffers->draftIndices;
outputParams->nextDraftProbs = explicitDraftTokensBuffers->draftProbs;
outputParams->positionIdsBase = explicitDraftTokensBuffers->positionIdsBase;
outputParams->randomDataSample = explicitDraftTokensBuffers->randomDataSample;
outputParams->randomDataValidation = explicitDraftTokensBuffers->randomDataValidation;
outputParams->temperatures = explicitDraftTokensBuffers->temperatures;
outputParams->generationLengths = explicitDraftTokensBuffers->generationLengths;
outputParams->generationLengthsHost = explicitDraftTokensBuffers->generationLengthsHost;
outputParams->maxGenLengthHost = explicitDraftTokensBuffers->maxGenLengthHost;
}
if (decodingMode.isLookahead())
{
TLLM_CHECK(output.lookaheadOutputs);
auto outputParams = std::dynamic_pointer_cast<tl::LookaheadDecodingOutputs>(baseOutputs);
outputParams->packedMasks = output.lookaheadOutputs->packedMasks;
outputParams->positionIds = output.lookaheadOutputs->positionIds;
outputParams->positionOffsets = output.lookaheadOutputs->positionOffsets;
outputParams->generationLengths = output.lookaheadOutputs->generationLengths;
}
TLLM_LOG_TRACE("%s stop", __PRETTY_FUNCTION__);
}
std::shared_ptr<tl::BaseDecodingOutputs> prepareOutputs(
DecodingOutput& output, DecodingOutput::TensorPtr& logProbsTiled, tle::DecodingMode const& decodingMode)
{
TLLM_LOG_TRACE("%s start", __PRETTY_FUNCTION__);
std::shared_ptr<tl::BaseDecodingOutputs> outputParams;
if (decodingMode.isBeamSearch())
{
outputParams = std::make_shared<tl::BeamSearchOutputs>(output.ids);
}
else if (decodingMode.isMedusa())
{
outputParams = std::make_shared<tl::SpeculativeDecodingOutputs>(output.ids);
}
else if (decodingMode.isLookahead())
{
outputParams = std::make_shared<tl::LookaheadDecodingOutputs>(output.ids);
}
else if (decodingMode.isExplicitDraftTokens())
{
outputParams = std::make_shared<tl::ExplicitDraftTokensOutputs>(output.ids);
}
else
{
outputParams = std::make_shared<tl::BaseDecodingOutputs>(output.ids);
}
// Common outputs
outputParams->newTokens = output.newTokens;
if (output.cumLogProbs)
{
outputParams->cumLogProbs = output.cumLogProbs;
}
if (output.parentIds)
{
outputParams->parentIds = output.parentIds;
}
if (output.finishReasons)
{
outputParams->finished = output.finishReasons;
}
if (output.finishedSum)
{
outputParams->finishedSum = output.finishedSum;
}
if (output.lengths)
{
outputParams->sequenceLength = output.lengths;
}
if (output.logProbs)
{
outputParams->outputLogProbs = output.logProbs;
outputParams->outputLogProbsTiled = logProbsTiled;
}
// Beam search outputs
if (decodingMode.isBeamSearch())
{
prepareBeamSearchOutputs(output, outputParams);
}
// Speculative decoding outputs
if (decodingMode.isMedusa() || decodingMode.isLookahead() || decodingMode.isExplicitDraftTokens())
{
prepareSpeculativeDecodingOutputs(output, outputParams, decodingMode);
}
TLLM_LOG_TRACE("%s stop", __PRETTY_FUNCTION__);
return outputParams;
}
} // namespace
template <typename T>
void GptDecoder<T>::forwardAsync(DecodingOutput& output, DecodingInput const& input)
{
TLLM_LOG_TRACE("%s start", __PRETTY_FUNCTION__);
auto forwardParams = prepareInputs<T>(input, mMaxBatchSize, mDecodingMode);
auto outputParams = prepareOutputs(output, mLogProbsTiled, mDecodingMode);
mDynamicDecodeLayer->forwardAsync(outputParams, forwardParams, mDecodingLayerWorkspace);
TLLM_LOG_TRACE("%s stop", __PRETTY_FUNCTION__);
}
template <typename T>
void GptDecoder<T>::forwardSync(DecodingOutput& output, DecodingInput const& input)
{
TLLM_LOG_TRACE("%s start", __PRETTY_FUNCTION__);
auto forwardParams = prepareInputs<T>(input, mMaxBatchSize, mDecodingMode);
auto outputParams = prepareOutputs(output, mLogProbsTiled, mDecodingMode);
mDynamicDecodeLayer->forwardSync(outputParams, forwardParams, mDecodingLayerWorkspace);
TLLM_LOG_TRACE("%s stop", __PRETTY_FUNCTION__);
}
// Must be similar to [cpp/tensorrt_llm/thop/gatherTreeOp.cpp] gatherTree
template <typename T>
void GptDecoder<T>::gatherTree(DecodingOutput const& decodingOutput, DecodingInput const& decodingInput,
BufferManager const& manager, std::optional<std::reference_wrapper<SamplingConfig const>> samplingConfig)
{
TLLM_LOG_TRACE("%s start", __PRETTY_FUNCTION__);
auto& finalOutputIds = *decodingOutput.gatheredIds;
auto const& finalOutputIdsShape = finalOutputIds.getShape();
auto const& decodingOutputIdsShape = decodingOutput.ids->getShape();
auto const batchSize = finalOutputIdsShape.d[0];
auto const beamWidth = finalOutputIdsShape.d[1];
auto const maxSeqLength = finalOutputIdsShape.d[2];
TLLM_CHECK_WITH_INFO(beamWidth > 1, "gatherTree is only needed for beam search.");
TLLM_CHECK_WITH_INFO(decodingOutputIdsShape.d[0] == batchSize,
common::fmtstr("Decoder batch size (" FMT_DIM ") does not match final batch size (" FMT_DIM ")",
decodingOutputIdsShape.d[0], batchSize));
TLLM_CHECK_WITH_INFO(decodingOutputIdsShape.d[1] == beamWidth,
common::fmtstr("Decoder beam width (" FMT_DIM ") does not match final beam width (" FMT_DIM ")",
decodingOutputIdsShape.d[1], beamWidth));
TLLM_CHECK_WITH_INFO(decodingOutputIdsShape.d[2] <= maxSeqLength,
common::fmtstr("Decoder seq length size (" FMT_DIM ") is too large for final seq length (" FMT_DIM ")",
decodingOutputIdsShape.d[2], maxSeqLength));
auto const& stream = manager.getStream().get();
// prefill finalOutputIds with the EOS tokens from decodingInput.endIds
tensorrt_llm::kernels::invokeInitializeOutput(bufferCast<TokenIdType>(finalOutputIds),
bufferCast<TokenIdType>(*decodingInput.endIds), batchSize * beamWidth, maxSeqLength, stream);
sync_check_cuda_error();
// Prepare length penalty, use the value from samplingConfig or 1.0f by default
SamplingConfig const& samplingConf = samplingConfig ? (*samplingConfig).get() : mSamplingConfig;
std::vector<float> lengthPenaltyVec;
TensorPtr lengthPenaltyPtr
= std::shared_ptr(manager.gpu(ITensor::makeShape({batchSize}), TRTDataType<float>::value));
if (!samplingConf.lengthPenalty.has_value() || samplingConf.lengthPenalty.value().size() == 0)
{
lengthPenaltyVec = std::vector<float>(batchSize, 1.0f);
}
else if (long int const size = samplingConf.lengthPenalty.value().size(); size == 1)
{
lengthPenaltyVec = std::vector<float>(batchSize, samplingConf.lengthPenalty.value()[0]);
}
else
{
TLLM_CHECK_WITH_INFO(size == batchSize,
common::fmtstr("Size of lengthPenalty in SamplingConfig (" FMT_DIM ") is different from batchSize (" FMT_DIM
")",
size, batchSize));
lengthPenaltyVec = samplingConf.lengthPenalty.value();
}
lengthPenaltyPtr = manager.copyFrom(lengthPenaltyVec, ITensor::makeShape({batchSize}), runtime::MemoryType::kGPU);
tensorrt_llm::kernels::BeamHypotheses bh;
bh.nMaxBatchSize = batchSize;
bh.nBatchSize = batchSize;
bh.nBeamWidth = beamWidth;
bh.nMaxSeqLen = maxSeqLength;
bh.lengthPenalties = bufferCast<float>(*lengthPenaltyPtr);
bh.inputLengths = bufferCast<SizeType32>(*decodingInput.lengths);
bh.outputIds = bufferCast<TokenIdType>(finalOutputIds);
bh.logProbs = bufferCastOrNull<float>(decodingOutput.logProbs);
bh.logProbsTiled = bufferCast<float>(*mLogProbsTiled);
bh.sequenceLengths = bufferCast<SizeType32>(*decodingOutput.lengths);
bh.cumLogProbs = bufferCast<float>(*decodingOutput.cumLogProbs);
bh.outputIdsCBA = bufferCast<TokenIdType>(*decodingOutput.beamHypotheses.outputIdsCBA);
bh.logProbsCBA = bufferCast<float>(*decodingOutput.beamHypotheses.logProbsCBA);
bh.sequenceLengthsCBA = bufferCast<SizeType32>(*decodingOutput.beamHypotheses.sequenceLengthsCBA);
bh.cumLogProbsCBA = bufferCast<float>(*decodingOutput.beamHypotheses.cumLogProbsCBA);
bh.normedScoresCBA = bufferCast<float>(*decodingOutput.beamHypotheses.normedScoresCBA);
bh.numBeamsCBA = bufferCast<SizeType32>(*decodingOutput.beamHypotheses.numBeamsCBA);
bh.minNormedScoresCBA = bufferCast<float>(*decodingOutput.beamHypotheses.minNormedScoresCBA);
bh.batchDones = bufferCast<bool>(*decodingOutput.beamHypotheses.batchDones);
bh.finished = bufferCast<tensorrt_llm::kernels::FinishedState>(*decodingOutput.finishReasons);
bh.outputIdsUnfinish = bufferCast<TokenIdType>(*decodingOutput.ids);
bh.parentIdsUnfinish = bufferCast<TokenIdType>(*decodingOutput.parentIds);
// This is where transpose is done
tensorrt_llm::kernels::invokeInsertUnfinishedPath(bh, stream);
sync_check_cuda_error();
tensorrt_llm::kernels::invokeFinalize(bh, stream);
sync_check_cuda_error();
TLLM_LOG_TRACE("%s stop", __PRETTY_FUNCTION__);
}
namespace tensorrt_llm::runtime
{
template class GptDecoder<float>;
template class GptDecoder<half>;
} // namespace tensorrt_llm::runtime
void IGptDecoder::acceptDraftTokensByIds(ITensor const& targetTokenIds, ITensor const& draftTokenIds,
ITensor const& contextLengths, ITensor const& numDraftTokens, ITensor& sequenceLengths, ITensor const& finishedVec,
ITensor& finishedFinal, ITensor& finishedSum, ITensor const& batchSlots, BufferManager::CudaStreamPtr const& stream)
{
TLLM_LOG_TRACE("%s start", __PRETTY_FUNCTION__);
auto const finishedVecShape = finishedVec.getShape();
auto const maxBatchSize = finishedVecShape.d[1];
auto const batchSlotsShape = batchSlots.getShape();
auto const batchSize = batchSlotsShape.d[0];
auto const targetTokenIdsShape = targetTokenIds.getShape();
auto const beamWidth = targetTokenIdsShape.d[1];
auto const maxSeqLength = targetTokenIdsShape.d[2];
auto const maxDraftTokens = draftTokenIds.getDimension<1>();
TLLM_CHECK_WITH_INFO(beamWidth == 1,
common::fmtstr("Beam width (" FMT_DIM ") > 1 is not supported for the speculative decoding", beamWidth));
TLLM_CHECK_WITH_INFO(batchSize <= maxBatchSize,
common::fmtstr("Batch size (" FMT_DIM ") is not smaller or equal to max batch size (" FMT_DIM ")", batchSize,
maxBatchSize));
TLLM_CHECK_WITH_INFO(draftTokenIds.getDimension<0>() == maxBatchSize,
common::fmtstr("Draft tokens batch size (" FMT_DIM ") is not equal to target batch size (" FMT_DIM ")",
draftTokenIds.getDimension<0>(), maxBatchSize));
TLLM_CHECK_WITH_INFO(contextLengths.getDimension<0>() == maxBatchSize,
common::fmtstr("Context length batch size (" FMT_DIM ") is not equal to batch size (" FMT_DIM ")",
contextLengths.getDimension<0>(), maxBatchSize));
TLLM_CHECK_WITH_INFO(numDraftTokens.getDimension<0>() == maxBatchSize,
common::fmtstr("Num draft tokens batch size (" FMT_DIM ") is not equal to batch size (" FMT_DIM ")",
numDraftTokens.getDimension<0>(), maxBatchSize));
TLLM_CHECK_WITH_INFO(sequenceLengths.getDimension<0>() == maxBatchSize,
common::fmtstr("Sequence length batch size (" FMT_DIM ") is not equal to batch size (" FMT_DIM ")",
sequenceLengths.getDimension<0>(), maxBatchSize));
tksd::invokeAcceptDraftTokensByIds(bufferCast<TokenIdType>(draftTokenIds), bufferCast<TokenIdType>(targetTokenIds),
bufferCast<SizeType32>(contextLengths), bufferCast<SizeType32>(numDraftTokens),
bufferCast<SizeType32>(sequenceLengths),
reinterpret_cast<tensorrt_llm::kernels::FinishedState const*>(
bufferCast<tensorrt_llm::kernels::FinishedState::UnderlyingType>(finishedVec)),
reinterpret_cast<tensorrt_llm::kernels::FinishedState*>(
bufferCast<tensorrt_llm::kernels::FinishedState::UnderlyingType>(finishedFinal)),
bufferCast<int>(finishedSum), bufferCast<SizeType32>(batchSlots), batchSize, maxBatchSize, beamWidth,
maxSeqLength, maxDraftTokens, stream->get());
sync_check_cuda_error();
TLLM_LOG_TRACE("%s stop", __PRETTY_FUNCTION__);
}
void IGptDecoder::acceptDraftTokensByLogits(ITensor& draftLogits, ITensor const& targetLogits, ITensor& draftProbs,
ITensor& targetProbs, ITensor const& numDraftTokens, ITensor& finished, ITensor const& batchSlots,
SizeType32 vocabSize, SizeType32 vocabSizePadded, bool useRandomAcceptThreshold, float randomAcceptThreshold,
curandState_t* curandState, BufferManager::CudaStreamPtr const& stream)
{
TLLM_LOG_TRACE("%s start", __PRETTY_FUNCTION__);
auto const draftLogitsShape = draftLogits.getShape();
auto const maxBatchSize = draftLogitsShape.d[0];
auto const maxTokensPerStep = draftLogitsShape.d[1];
auto const batchSlotsShape = batchSlots.getShape();
auto const batchSize = batchSlotsShape.d[0];
auto constexpr beamWidth = 1;
TLLM_CHECK_WITH_INFO(
beamWidth == 1, common::fmtstr("Beam width (%d) > 1 is not supported for the speculative decoding", beamWidth));
TLLM_CHECK(draftLogitsShape.d[2] == vocabSize);
if (draftLogits.getDataType() == nvinfer1::DataType::kFLOAT)
{
tksd::acceptDraftTokensByLogits(bufferCast<float>(draftLogits),
const_cast<float**>(reinterpret_cast<float const* const*>(bufferCast<int64_t>(targetLogits))),
bufferCast<float>(draftProbs), bufferCast<float>(targetProbs), bufferCast<SizeType32>(numDraftTokens),
reinterpret_cast<tensorrt_llm::kernels::FinishedState*>(
bufferCast<tensorrt_llm::kernels::FinishedState::UnderlyingType>(finished)),
curandState, bufferCast<SizeType32>(batchSlots), batchSize, maxBatchSize, beamWidth, vocabSize,
vocabSizePadded, maxTokensPerStep, useRandomAcceptThreshold, randomAcceptThreshold, stream->get());
}
else if (draftLogits.getDataType() == nvinfer1::DataType::kHALF)
{
tksd::acceptDraftTokensByLogits(bufferCast<half>(draftLogits),
const_cast<half**>(reinterpret_cast<half const* const*>(bufferCast<int64_t>(targetLogits))),
bufferCast<half>(draftProbs), bufferCast<half>(targetProbs), bufferCast<SizeType32>(numDraftTokens),
reinterpret_cast<tensorrt_llm::kernels::FinishedState*>(
bufferCast<tensorrt_llm::kernels::FinishedState::UnderlyingType>(finished)),
curandState, bufferCast<SizeType32>(batchSlots), batchSize, maxBatchSize, beamWidth, vocabSize,
vocabSizePadded, maxTokensPerStep, useRandomAcceptThreshold, randomAcceptThreshold, stream->get());
}
else
{
TLLM_THROW("Incorrect logits dtype. Only float32 and float16 are supported");
}
sync_check_cuda_error();
TLLM_LOG_TRACE("%s stop", __PRETTY_FUNCTION__);
}
Reference in New Issue View Git Blame Copy Permalink