164 static constexpr int32_t kVALUE = 14;
202 mImpl->setName(name);
214 return mImpl->getName();
233 mImpl->setDimensions(dimensions);
247 return mImpl->getDimensions();
283 mImpl->setType(type);
298 return mImpl->getType();
315 return mImpl->setDynamicRange(min, max);
323 return mImpl->isNetworkInput();
331 return mImpl->isNetworkOutput();
348 mImpl->setBroadcastAcrossBatch(broadcastAcrossBatch);
362 return mImpl->getBroadcastAcrossBatch();
374 return mImpl->getLocation();
393 mImpl->setLocation(location);
405 return mImpl->dynamicRangeIsSet();
413 mImpl->resetDynamicRange();
423 return mImpl->getDynamicRangeMin();
433 return mImpl->getDynamicRangeMax();
455 mImpl->setAllowedFormats(formats);
468 return mImpl->getAllowedFormats();
499 return mImpl->isShapeTensor();
520 return mImpl->isExecutionTensor();
546 mImpl->setDimensionName(index, name);
561 return mImpl->getDimensionName(index);
586 return mLayer->getType();
600 mLayer->setName(name);
610 return mLayer->getName();
618 return mLayer->getNbInputs();
631 return mLayer->getInput(index);
639 return mLayer->getNbOutputs();
649 return mLayer->getOutput(index);
666 return mLayer->setInput(index, tensor);
699 mLayer->setPrecision(dataType);
711 return mLayer->getPrecision();
725 return mLayer->precisionIsSet();
737 mLayer->resetPrecision();
787 mLayer->setOutputType(index, dataType);
802 return mLayer->getOutputType(index);
818 return mLayer->outputTypeIsSet(index);
832 return mLayer->resetOutputType(index);
850 mLayer->setMetadata(metadata);
863 return mLayer->getMetadata();
868 apiv::VLayer* mLayer;
1045 static constexpr int32_t kVALUE = 4;
1073 mImpl->setNbOutputMaps(nbOutputMaps);
1083 return mImpl->getNbOutputMaps();
1103 mImpl->setNbGroups(nbGroups);
1113 return mImpl->getNbGroups();
1127 mImpl->setKernelWeights(weights);
1137 return mImpl->getKernelWeights();
1152 mImpl->setBiasWeights(weights);
1162 return mImpl->getBiasWeights();
1179 mImpl->setPrePadding(padding);
1189 return mImpl->getPrePadding();
1206 mImpl->setPostPadding(padding);
1216 return mImpl->getPostPadding();
1230 mImpl->setPaddingMode(paddingMode);
1242 return mImpl->getPaddingMode();
1255 mImpl->setKernelSizeNd(kernelSize);
1265 return mImpl->getKernelSizeNd();
1280 mImpl->setStrideNd(stride);
1290 return mImpl->getStrideNd();
1308 mImpl->setPaddingNd(padding);
1320 return mImpl->getPaddingNd();
1334 mImpl->setDilationNd(dilation);
1344 return mImpl->getDilationNd();
1393 mImpl->setActivationType(type);
1403 return mImpl->getActivationType();
1418 mImpl->setAlpha(alpha);
1432 mImpl->setBeta(beta);
1441 return mImpl->getAlpha();
1450 return mImpl->getBeta();
1480 static constexpr int32_t kVALUE = 3;
1507 mImpl->setPoolingType(type);
1517 return mImpl->getPoolingType();
1532 mImpl->setBlendFactor(blendFactor);
1545 return mImpl->getBlendFactor();
1559 mImpl->setAverageCountExcludesPadding(exclusive);
1570 return mImpl->getAverageCountExcludesPadding();
1588 mImpl->setPrePadding(padding);
1598 return mImpl->getPrePadding();
1616 mImpl->setPostPadding(padding);
1626 return mImpl->getPostPadding();
1639 mImpl->setPaddingMode(paddingMode);
1650 return mImpl->getPaddingMode();
1663 mImpl->setWindowSizeNd(windowSize);
1673 return mImpl->getWindowSizeNd();
1688 mImpl->setStrideNd(stride);
1698 return mImpl->getStrideNd();
1717 mImpl->setPaddingNd(padding);
1729 return mImpl->getPaddingNd();
1760 mImpl->setWindowSize(windowSize);
1770 return mImpl->getWindowSize();
1782 mImpl->setAlpha(alpha);
1792 return mImpl->getAlpha();
1804 mImpl->setBeta(beta);
1814 return mImpl->getBeta();
1836 return mImpl->getK();
1902 mImpl->setMode(mode);
1912 return mImpl->getMode();
1922 mImpl->setShift(shift);
1932 return mImpl->getShift();
1942 mImpl->setScale(scale);
1952 return mImpl->getScale();
1962 mImpl->setPower(power);
1972 return mImpl->getPower();
1987 return mImpl->getChannelAxis();
2008 mImpl->setChannelAxis(channelAxis);
2061 mImpl->setAxes(axes);
2071 return mImpl->getAxes();
2107 mImpl->setAxis(axis);
2117 return mImpl->getAxis();
2144 mImpl->setNbOutputMaps(nbOutputMaps);
2154 return mImpl->getNbOutputMaps();
2174 mImpl->setNbGroups(nbGroups);
2184 return mImpl->getNbGroups();
2198 mImpl->setKernelWeights(weights);
2208 return mImpl->getKernelWeights();
2223 mImpl->setBiasWeights(weights);
2233 return mImpl->getBiasWeights();
2250 mImpl->setPrePadding(padding);
2260 return mImpl->getPrePadding();
2277 mImpl->setPostPadding(padding);
2287 return mImpl->getPostPadding();
2301 mImpl->setPaddingMode(paddingMode);
2313 return mImpl->getPaddingMode();
2328 mImpl->setKernelSizeNd(kernelSize);
2338 return mImpl->getKernelSizeNd();
2355 mImpl->setStrideNd(stride);
2365 return mImpl->getStrideNd();
2383 mImpl->setPaddingNd(padding);
2395 return mImpl->getPaddingNd();
2421 mImpl->setDilationNd(dilation);
2431 return mImpl->getDilationNd();
2479 static constexpr int32_t kVALUE = 14;
2516 return mImpl->setOperation(op);
2528 return mImpl->getOperation();
2649 mImpl->setGatherAxis(axis);
2661 return mImpl->getGatherAxis();
2684 mImpl->setNbElementWiseDims(elementWiseDims);
2694 return mImpl->getNbElementWiseDims();
2704 mImpl->setMode(mode);
2714 return mImpl->getMode();
2743 return mImpl->getPlugin();
2770 return mImpl->getPlugin();
2853 mImpl->setOperation(op);
2863 return mImpl->getOperation();
2926 mImpl->setOperation(op);
2936 return mImpl->getOperation();
2946 mImpl->setReduceAxes(reduceAxes);
2956 return mImpl->getReduceAxes();
2966 mImpl->setKeepDimensions(keepDimensions);
2976 return mImpl->getKeepDimensions();
3010 mImpl->setPrePaddingNd(padding);
3022 return mImpl->getPrePaddingNd();
3036 mImpl->setPostPaddingNd(padding);
3048 return mImpl->getPostPaddingNd();
3098 mImpl->setFirstTranspose(permutation);
3110 return mImpl->getFirstTranspose();
3138 mImpl->setReshapeDimensions(dimensions);
3151 return mImpl->getReshapeDimensions();
3198 mImpl->setSecondTranspose(permutation);
3210 return mImpl->getSecondTranspose();
3226 return mImpl->setZeroIsPlaceholder(zeroIsPlaceholder);
3239 return mImpl->getZeroIsPlaceholder();
3350 mImpl->setStart(start);
3365 return mImpl->getStart();
3379 return mImpl->setSize(size);
3394 return mImpl->getSize();
3408 mImpl->setStride(stride);
3423 return mImpl->getStride();
3433 mImpl->setMode(mode);
3443 return mImpl->getMode();
3486 mImpl->setAxes(axes);
3501 return mImpl->getAxes();
3571 mImpl->setOperation(op);
3581 return mImpl->getOperation();
3609 return mImpl->getK();
3619 mImpl->setReduceAxes(reduceAxes);
3629 return mImpl->getReduceAxes();
3731 mImpl->setOperation(index, op);
3743 return mImpl->getOperation(index);
3868 mImpl->setToType(toType);
3879 return mImpl->getToType();
3908 mImpl->setWeights(weights);
3918 return mImpl->getWeights();
3930 mImpl->setDimensions(dimensions);
3942 return mImpl->getDimensions();
3988 static constexpr int32_t kVALUE = 3;
4042 static constexpr int32_t kVALUE = 3;
4072 static constexpr int32_t kVALUE = 2;
4108 static constexpr int32_t kVALUE = 4;
4171 return mImpl->setOutputDimensions(dimensions);
4181 return mImpl->getOutputDimensions();
4209 void setScales(
float const* scales, int32_t nbScales)
noexcept
4211 mImpl->setScales(scales, nbScales);
4228 int32_t
getScales(int32_t size,
float* scales)
const noexcept
4230 return mImpl->getScales(size, scales);
4242 mImpl->setResizeMode(interpolationMode);
4252 return mImpl->getResizeMode();
4287 mImpl->setCoordinateTransformation(coordTransform);
4297 return mImpl->getCoordinateTransformation();
4312 mImpl->setSelectorForSinglePixel(selector);
4322 return mImpl->getSelectorForSinglePixel();
4336 mImpl->setNearestRounding(value);
4346 return mImpl->getNearestRounding();
4368 mImpl->setCubicCoeff(A);
4378 return mImpl->getCubicCoeff();
4391 mImpl->setExcludeOutside(excludeFlag);
4401 return mImpl->getExcludeOutside();
4483 return mBoundary->getLoop();
4488 apiv::VLoopBoundaryLayer* mBoundary;
4506 return mBoundary->getConditional();
4511 apiv::VConditionalBoundaryLayer* mBoundary;
4595 return mImpl->setCondition(condition);
4613 return mImpl->addOutput(trueSubgraphOutput, falseSubgraphOutput);
4625 return mImpl->addInput(input);
4640 mImpl->setName(name);
4650 return mImpl->getName();
4720 return mImpl->getLoopOutput();
4737 mImpl->setAxis(axis);
4745 return mImpl->getAxis();
4794 return mImpl->getTripLimit();
4820 mImpl->setAxis(axis);
4828 return mImpl->getAxis();
4842 mImpl->setReverse(reverse);
4852 return mImpl->getReverse();
4881 return mImpl->addRecurrence(initialValue);
4902 return mImpl->addTripLimit(tensor, limit);
4915 return mImpl->addIterator(tensor, axis, reverse);
4928 return mImpl->addLoopOutput(tensor, outputKind, axis);
4943 mImpl->setName(name);
4953 return mImpl->getName();
5008 mImpl->setMessage(message);
5018 return mImpl->getMessage();
5120 mImpl->setDimensions(dimensions);
5135 return mImpl->getDimensions();
5145 mImpl->setOperation(op);
5155 return mImpl->getOperation();
5174 mImpl->setAlpha(alpha);
5189 return mImpl->getAlpha();
5208 mImpl->setBeta(beta);
5223 return mImpl->getBeta();
5284 mImpl->setAlphaInt64(alpha);
5299 return mImpl->getAlphaInt64();
5318 mImpl->setBetaInt64(beta);
5333 return mImpl->getBetaInt64();
5341 return mImpl->isAlphaBetaInt64();
5358 mImpl->setToType(toType);
5370 return mImpl->getToType();
5465 return mImpl->getAxis();
5476 mImpl->setAxis(axis);
5492 mImpl->setToType(toType);
5504 return mImpl->getToType();
5596 return mImpl->getAxis();
5607 mImpl->setAxis(axis);
5623 mImpl->setToType(toType);
5635 return mImpl->getToType();
5690 mImpl->setToType(toType);
5703 return mImpl->getToType();
5716 mImpl->setScaleType(scaleType);
5729 return mImpl->getScaleType();
5742 mImpl->setAxis(axis);
5752 return mImpl->getAxis();
5765 mImpl->setBlockSize(size);
5775 return mImpl->getBlockSize();
5831 return mImpl->setEquation(equation);
5841 return mImpl->getEquation();
5940 mImpl->setMode(mode);
5950 return mImpl->getMode();
5960 mImpl->setAxis(axis);
5968 return mImpl->getAxis();
6012 mImpl->setAxis(axis);
6020 return mImpl->getAxis();
6049 mImpl->setInterpolationMode(mode);
6061 return mImpl->getInterpolationMode();
6071 mImpl->setAlignCorners(alignCorners);
6083 return mImpl->getAlignCorners();
6095 return mImpl->setSampleMode(mode);
6107 return mImpl->getSampleMode();
6201 mImpl->setBoundingBoxFormat(fmt);
6213 return mImpl->getBoundingBoxFormat();
6227 mImpl->setTopKBoxLimit(limit);
6237 return mImpl->getTopKBoxLimit();
6290 mImpl->setBatchAxis(batchAxis);
6300 return mImpl->getBatchAxis();
6313 mImpl->setSequenceAxis(sequenceAxis);
6323 return mImpl->getSequenceAxis();
6361 return mImpl->setEpsilon(eps);
6371 return mImpl->getEpsilon();
6381 return mImpl->setAxes(axesMask);
6391 return mImpl->getAxes();
6412 return mImpl->setNbGroups(nbGroups);
6422 return mImpl->getNbGroups();
6448 return mImpl->setComputePrecision(type);
6458 return mImpl->getComputePrecision();
6553 static constexpr int32_t kVALUE = 1;
6599 return mImpl->setOperation(op);
6611 return mImpl->getOperation();
6623 mImpl->setExclusive(exclusive);
6635 return mImpl->getExclusive();
6647 mImpl->setReverse(reverse);
6659 return mImpl->getReverse();
6726 return mImpl->addInput(name, type, dimensions);
6740 mImpl->markOutput(tensor);
6758 return mImpl->markDebug(tensor);
6774 return mImpl->unmarkDebug(tensor);
6784 return mImpl->isDebugTensor(tensor);
6806 return mImpl->markUnfusedTensorsAsDebugTensors();
6820 return mImpl->unmarkUnfusedTensorsAsDebugTensors();
6840 return mImpl->addActivation(input, type);
6859 return mImpl->addLRN(input, window, alpha, beta, k);
6885 return mImpl->addScale(input, mode, shift, scale, power);
6898 return mImpl->addSoftMax(input);
6915 return mImpl->addConcatenation(inputs, nbInputs);
6942 return mImpl->addElementWise(input1, input2, op);
6964 return mImpl->addUnary(input, operation);
6978 return mImpl->addShuffle(input);
6995 return mImpl->addOneHot(indices, values, depth, axis);
7007 return mImpl->getNbLayers();
7021 return mImpl->getLayer(index);
7033 return mImpl->getNbInputs();
7049 return mImpl->getInput(index);
7063 return mImpl->getNbOutputs();
7079 return mImpl->getOutput(index);
7106 return mImpl->addReduce(input, operation, reduceAxes, keepDimensions);
7138 return mImpl->addTopK(input, op, k, reduceAxes);
7154 return mImpl->addGather(data, indices, axis);
7170 return mImpl->addGatherV2(data, indices, mode);
7189 return mImpl->addRaggedSoftMax(input, bounds);
7211 return mImpl->addMatrixMultiply(input0, op0, input1, op1);
7225 return mImpl->addNonZero(input);
7249 return mImpl->addConstant(dimensions, weights);
7263 return mImpl->addIdentity(input);
7278 return mImpl->addCast(input, toType);
7293 mImpl->removeTensor(tensor);
7305 mImpl->unmarkOutput(tensor);
7326 return mImpl->addPluginV2(inputs, nbInputs, plugin);
7343 int32_t nbShapeInputs,
IPluginV3& plugin)
noexcept
7345 return mImpl->addPluginV3(inputs, nbInputs, shapeInputs, nbShapeInputs, plugin);
7364 return mImpl->addSlice(input, start, size, stride);
7388 mImpl->setName(name);
7402 return mImpl->getName();
7418 return mImpl->addShape(input);
7432 return mImpl->hasImplicitBatchDimension();
7442 return mImpl->getFlags();
7454 return mImpl->getFlag(networkDefinitionCreationFlag);
7471 return mImpl->markOutputForShapes(tensor);
7483 return mImpl->unmarkOutputForShapes(tensor);
7501 return mImpl->addParametricReLU(input, slope);
7524 return mImpl->addConvolutionNd(input, nbOutputMaps, kernelSize, kernelWeights, biasWeights);
7543 return mImpl->addPoolingNd(input, type, windowSize);
7566 return mImpl->addDeconvolutionNd(input, nbOutputMaps, kernelSize, kernelWeights, biasWeights);
7603 return mImpl->addScaleNd(input, mode, shift, scale, power, channelAxis);
7619 return mImpl->addResize(input);
7633 return mImpl->addLoop();
7648 return mImpl->addIfConditional();
7687 return mImpl->addSelect(condition, thenInput, elseInput);
7704 return mImpl->addAssertion(condition, message);
7729 return mImpl->addFill(dimensions, op);
7755 return mImpl->addFillV2(dimensions, op, outputType);
7771 return mImpl->addPaddingNd(input, prePadding, postPadding);
7795 return mImpl->setWeightsName(weights, name);
7814 mImpl->setErrorRecorder(recorder);
7829 return mImpl->getErrorRecorder();
7850 return mImpl->addDequantize(input, scale);
7872 return mImpl->addDequantizeV2(input, scale, outputType);
7892 return mImpl->addScatter(data, indices, updates, mode);
7913 return mImpl->addQuantize(input, scale);
7935 return mImpl->addQuantizeV2(input, scale, outputType);
7963 return mImpl->addDynamicQuantize(input, axis, blockSize, outputType, scaleType);
7978 return mImpl->addEinsum(inputs, nbInputs, equation);
7996 return mImpl->addGridSample(input, grid);
8014 return mImpl->addNMS(boxes, scores, maxOutputBoxesPerClass);
8031 return mImpl->addReverseSequence(input, sequenceLens);
8057 return mImpl->addNormalization(input, scale, bias, axesMask);
8079 return mImpl->addCumulative(input, axis, operation, exclusive, reverse);
8090 return mImpl->getBuilder();
8103 return mImpl->markWeightsRefittable(name);
8115 return mImpl->unmarkWeightsRefittable(name);
8128 return mImpl->areWeightsMarkedRefittable(name);
8147 return mImpl->addSqueeze(input, axes);
8168 return mImpl->addUnsqueeze(input, axes);
8240 virtual
bool getBatch(
void* bindings[],
char const* names[], int32_t nbBindings) noexcept = 0;
8256 virtual
void const* readCalibrationCache(std::
size_t& length) noexcept = 0;
8266 virtual
void writeCalibrationCache(
void const* ptr, std::
size_t length) noexcept = 0;
8432 virtual
double getRegressionCutoff() const noexcept = 0;
8446 virtual
void const* readHistogramCache(std::
size_t& length) noexcept = 0;
8456 virtual
void writeHistogramCache(
void const* ptr, std::
size_t length) noexcept = 0;
8499 return mImpl->getDataType();
8510 return mImpl->getStrides();
8520 return mImpl->getVectorizedDim();
8531 return mImpl->getComponentsPerElement();
8560 return mImpl->getImplementation();
8568 return mImpl->getTactic();
8596 return mImpl->getName();
8608 return mImpl->getDimensions(index, select);
8616 return mImpl->getNbInputs();
8624 return mImpl->getNbOutputs();
8653 return mImpl->getAlgorithmVariant();
8661 return mImpl->getTimingMSec();
8669 return mImpl->getWorkspaceSize();
8683 return mImpl->getAlgorithmIOInfoByIndex(index);
8718 int32_t nbChoices, int32_t* selection)
noexcept = 0;
8731 int32_t nbAlgorithms)
noexcept = 0;
8825 static constexpr int32_t kVALUE = 2;
9065 static constexpr uint64_t kINVALID_TACTIC_HASH = UINT64_MAX;
9098 return mImpl->serialize();
9122 return mImpl->combine(inputCache, ignoreMismatch);
9132 return mImpl->reset();
9149 int64_t
queryKeys(TimingCacheKey* keyBuffer, int64_t capacity)
const noexcept
9151 return mImpl->queryKeys(keyBuffer, capacity);
9166 TimingCacheValue
query(TimingCacheKey
const& key)
const noexcept
9168 return mImpl->query(key);
9188 bool update(TimingCacheKey
const& key, TimingCacheValue
const& value)
noexcept
9190 return mImpl->update(key, value);
9311 static constexpr int32_t kVALUE = 3;
9363 static constexpr int32_t kVALUE = 3;
9403 static constexpr int32_t kVALUE = 4;
9442 virtual void phaseStart(
char const* phaseName,
char const* parentPhase, int32_t nbSteps)
noexcept = 0;
9456 virtual bool stepComplete(
char const* phaseName, int32_t step)
noexcept = 0;
9515 virtual
void setAvgTimingIterations(int32_t avgTiming) noexcept
9517 mImpl->setAvgTimingIterations(avgTiming);
9529 return mImpl->getAvgTimingIterations();
9542 mImpl->setEngineCapability(capability);
9554 return mImpl->getEngineCapability();
9566 mImpl->setInt8Calibrator(calibrator);
9576 return mImpl->getInt8Calibrator();
9593 mImpl->setFlags(builderFlags);
9605 return mImpl->getFlags();
9617 mImpl->clearFlag(builderFlag);
9629 mImpl->setFlag(builderFlag);
9641 return mImpl->getFlag(builderFlag);
9658 mImpl->setDeviceType(layer, deviceType);
9668 return mImpl->getDeviceType(layer);
9680 return mImpl->isDeviceTypeSet(layer);
9690 mImpl->resetDeviceType(layer);
9700 return mImpl->canRunOnDLA(layer);
9716 mImpl->setDLACore(dlaCore);
9726 return mImpl->getDLACore();
9737 mImpl->setDefaultDeviceType(deviceType);
9747 return mImpl->getDefaultDeviceType();
9769 return mImpl->setProfileStream(stream);
9781 return mImpl->getProfileStream();
9798 return mImpl->addOptimizationProfile(profile);
9811 return mImpl->getNbOptimizationProfiles();
9823 mImpl->setProfilingVerbosity(verbosity);
9836 return mImpl->getProfilingVerbosity();
9848 mImpl->setAlgorithmSelector(selector);
9858 return mImpl->getAlgorithmSelector();
9876 return mImpl->setCalibrationProfile(profile);
9888 return mImpl->getCalibrationProfile();
9907 mImpl->setQuantizationFlags(flags);
9921 return mImpl->getQuantizationFlags();
9935 mImpl->clearQuantizationFlag(flag);
9949 mImpl->setQuantizationFlag(flag);
9963 return mImpl->getQuantizationFlag(flag);
9985 return mImpl->setTacticSources(tacticSources);
10000 return mImpl->getTacticSources();
10020 return mImpl->createTimingCache(blob, size);
10043 return mImpl->setTimingCache(cache, ignoreMismatch);
10053 return mImpl->getTimingCache();
10085 mImpl->setMemoryPoolLimit(pool, poolSize);
10104 return mImpl->getMemoryPoolLimit(pool);
10122 mImpl->setPreviewFeature(feature, enable);
10136 return mImpl->getPreviewFeature(feature);
10169 mImpl->setBuilderOptimizationLevel(level);
10181 return mImpl->getBuilderOptimizationLevel();
10198 mImpl->setHardwareCompatibilityLevel(hardwareCompatibilityLevel);
10211 return mImpl->getHardwareCompatibilityLevel();
10224 mImpl->setPluginsToSerialize(paths, nbPaths);
10237 return mImpl->getPluginToSerialize(index);
10247 return mImpl->getNbPluginsToSerialize();
10276 mImpl->setMaxAuxStreams(nbStreams);
10286 return mImpl->getMaxAuxStreams();
10302 return mImpl->setProgressMonitor(monitor);
10312 return mImpl->getProgressMonitor();
10328 mImpl->setRuntimePlatform(runtimePlatform);
10340 return mImpl->getRuntimePlatform();
10352 mImpl->setMaxNbTactics(maxNbTactics);
10364 return mImpl->getMaxNbTactics();
10380 return mImpl->setTilingOptimizationLevel(level);
10392 return mImpl->getTilingOptimizationLevel();
10408 return mImpl->setL2LimitForTiling(size);
10420 return mImpl->getL2LimitForTiling();
10498 return mImpl->platformHasFastFp16();
10508 return mImpl->platformHasFastInt8();
10520 return mImpl->getMaxDLABatchSize();
10528 return mImpl->getNbDLACores();
10546 mImpl->setGpuAllocator(allocator);
10560 return mImpl->createBuilderConfig();
10586 return mImpl->createNetworkV2(flags);
10601 return mImpl->createOptimizationProfile();
10620 mImpl->setErrorRecorder(recorder);
10635 return mImpl->getErrorRecorder();
10653 return mImpl->platformHasTf32();
10672 return mImpl->buildSerializedNetwork(network, config);
10692 return mImpl->buildEngineWithConfig(network, config);
10714 return mImpl->isNetworkSupported(network, config);
10724 return mImpl->getLogger();
10740 return mImpl->setMaxThreads(maxThreads);
10754 return mImpl->getMaxThreads();
10764 return mImpl->getPluginRegistry();
10777extern "C" TENSORRTAPI void* createInferBuilder_INTERNAL(
void* logger, int32_t version)
noexcept;
#define TENSORRTAPI
Definition: NvInferRuntimeBase.h:69
#define NV_TENSORRT_VERSION
Definition: NvInferRuntimeBase.h:101
#define TRT_DEPRECATED
Definition: NvInferRuntimeBase.h:42
#define TRT_DEPRECATED_ENUM
Definition: NvInferRuntimeBase.h:43
Definition: NvInferRuntimeBase.h:216
static constexpr int32_t MAX_DIMS
The maximum rank (number of dimensions) supported for a tensor.
Definition: NvInferRuntimeBase.h:219
An Activation layer in a network definition.
Definition: NvInfer.h:1382
void setBeta(float beta) noexcept
Set the beta parameter (must be finite).
Definition: NvInfer.h:1430
void setActivationType(ActivationType type) noexcept
Set the type of activation to be performed.
Definition: NvInfer.h:1391
ActivationType getActivationType() const noexcept
Get the type of activation to be performed.
Definition: NvInfer.h:1401
float getAlpha() const noexcept
Get the alpha parameter.
Definition: NvInfer.h:1439
virtual ~IActivationLayer() noexcept=default
float getBeta() const noexcept
Get the beta parameter.
Definition: NvInfer.h:1448
void setAlpha(float alpha) noexcept
Set the alpha parameter (must be finite).
Definition: NvInfer.h:1416
Describes the context and requirements, that could be fulfilled by one or more instances of IAlgorith...
Definition: NvInfer.h:8587
int32_t getNbOutputs() const noexcept
Return number of outputs of the algorithm.
Definition: NvInfer.h:8622
int32_t getNbInputs() const noexcept
Return number of inputs of the algorithm.
Definition: NvInfer.h:8614
char const * getName() const noexcept
Return name of the algorithm node.
Definition: NvInfer.h:8594
virtual ~IAlgorithmContext() noexcept=default
Dims getDimensions(int32_t index, OptProfileSelector select) const noexcept
Get the minimum / optimum / maximum dimensions for input or output tensor.
Definition: NvInfer.h:8606
Describes a variation of execution of a layer. An algorithm is represented by IAlgorithmVariant and t...
Definition: NvInfer.h:8646
std::size_t getWorkspaceSize() const noexcept
The size of the GPU temporary memory in bytes which the algorithm uses at execution time.
Definition: NvInfer.h:8667
float getTimingMSec() const noexcept
The time in milliseconds to execute the algorithm.
Definition: NvInfer.h:8659
IAlgorithmIOInfo const * getAlgorithmIOInfoByIndex(int32_t index) const noexcept
Returns the format of an Algorithm input or output. Algorithm inputs are incrementally numbered first...
Definition: NvInfer.h:8681
virtual ~IAlgorithm() noexcept=default
IAlgorithmVariant const & getAlgorithmVariant() const noexcept
Returns the algorithm variant.
Definition: NvInfer.h:8651
Carries information about input or output of the algorithm. IAlgorithmIOInfo for all the input and ou...
Definition: NvInfer.h:8490
virtual ~IAlgorithmIOInfo() noexcept=default
int64_t getVectorizedDim() const noexcept
Return the index of the vectorized dimension or -1 for non-vectorized formats.
Definition: NvInfer.h:8518
Dims getStrides() const noexcept
Return strides of the input/output tensor of algorithm. For vectorized formats, strides are given in ...
Definition: NvInfer.h:8508
DataType getDataType() const noexcept
Return DataType of the input/output of algorithm.
Definition: NvInfer.h:8497
int64_t getComponentsPerElement() const noexcept
Return the number of components per element. This is always 1 for non-vectorized formats.
Definition: NvInfer.h:8529
provides a unique 128-bit identifier, which along with the input and output information denotes the v...
Definition: NvInfer.h:8553
virtual ~IAlgorithmVariant() noexcept=default
int64_t getTactic() const noexcept
Return tactic of the algorithm.
Definition: NvInfer.h:8566
int64_t getImplementation() const noexcept
Return implementation of the algorithm.
Definition: NvInfer.h:8558
An assertion layer in a network.
Definition: NvInfer.h:4996
void setMessage(char const *message) noexcept
Set the message to print if the assertion fails.
Definition: NvInfer.h:5006
char const * getMessage() const noexcept
Return the assertion message.
Definition: NvInfer.h:5016
virtual ~IAssertionLayer() noexcept=default
Holds properties for configuring a builder to produce an engine.
Definition: NvInfer.h:9503
void setMemoryPoolLimit(MemoryPoolType pool, std::size_t poolSize) noexcept
Set the memory size for the memory pool.
Definition: NvInfer.h:10083
TRT_DEPRECATED void setQuantizationFlags(QuantizationFlags flags) noexcept
Set the quantization flags.
Definition: NvInfer.h:9905
nvinfer1::ITimingCache * createTimingCache(void const *blob, std::size_t size) const noexcept
Create timing cache.
Definition: NvInfer.h:10018
void setPreviewFeature(PreviewFeature feature, bool enable) noexcept
Enable or disable a specific preview feature.
Definition: NvInfer.h:10120
TRT_DEPRECATED void setAlgorithmSelector(IAlgorithmSelector *selector) noexcept
Set Algorithm Selector.
Definition: NvInfer.h:9846
TRT_DEPRECATED void setInt8Calibrator(IInt8Calibrator *calibrator) noexcept
Set Int8 Calibration interface.
Definition: NvInfer.h:9564
bool getPreviewFeature(PreviewFeature feature) const noexcept
Get status of preview feature.
Definition: NvInfer.h:10134
int32_t getBuilderOptimizationLevel() noexcept
Get builder optimization level.
Definition: NvInfer.h:10179
bool setTacticSources(TacticSources tacticSources) noexcept
Set tactic sources.
Definition: NvInfer.h:9983
void setPluginsToSerialize(char const *const *paths, int32_t nbPaths) noexcept
Set the plugin libraries to be serialized with version-compatible engines.
Definition: NvInfer.h:10222
bool setTilingOptimizationLevel(TilingOptimizationLevel level) noexcept
Set the Tiling optimization level.
Definition: NvInfer.h:10378
bool setL2LimitForTiling(int64_t size) noexcept
Set the L2 cache usage limit for Tiling optimization.
Definition: NvInfer.h:10406
TRT_DEPRECATED IInt8Calibrator * getInt8Calibrator() const noexcept
Get Int8 Calibration interface.
Definition: NvInfer.h:9574
std::size_t getMemoryPoolLimit(MemoryPoolType pool) const noexcept
Get the memory size limit of the memory pool.
Definition: NvInfer.h:10102
int32_t getDLACore() const noexcept
Get the DLA core that the engine executes on.
Definition: NvInfer.h:9724
int32_t getNbPluginsToSerialize() const noexcept
Get the number of plugin library paths to be serialized with version-compatible engines.
Definition: NvInfer.h:10245
void setDeviceType(ILayer const *layer, DeviceType deviceType) noexcept
Set the device that this layer must execute on.
Definition: NvInfer.h:9656
void setEngineCapability(EngineCapability capability) noexcept
Configure the builder to target specified EngineCapability flow.
Definition: NvInfer.h:9540
int32_t getMaxAuxStreams() const noexcept
Get the maximum number of auxiliary streams that TRT is allowed to use.
Definition: NvInfer.h:10284
bool getFlag(BuilderFlag builderFlag) const noexcept
Returns true if the build mode flag is set.
Definition: NvInfer.h:9639
void setMaxNbTactics(int32_t maxNbTactics) noexcept
Set the maximum number of tactics to time when there is a choice of tactics.
Definition: NvInfer.h:10350
TRT_DEPRECATED void clearQuantizationFlag(QuantizationFlag flag) noexcept
clear a quantization flag.
Definition: NvInfer.h:9933
int64_t getL2LimitForTiling() const noexcept
Get the L2 cache usage limit for tiling optimization.
Definition: NvInfer.h:10418
void setProgressMonitor(IProgressMonitor *monitor) noexcept
Sets the progress monitor for building a network.
Definition: NvInfer.h:10300
void setProfilingVerbosity(ProfilingVerbosity verbosity) noexcept
Set verbosity level of layer information exposed in NVTX annotations and IEngineInspector.
Definition: NvInfer.h:9821
int32_t getNbOptimizationProfiles() const noexcept
Get number of optimization profiles.
Definition: NvInfer.h:9809
nvinfer1::ITimingCache const * getTimingCache() const noexcept
Get the pointer to the timing cache from current IBuilderConfig.
Definition: NvInfer.h:10051
void reset() noexcept
Resets the builder configuration to defaults.
Definition: NvInfer.h:9755
bool setTimingCache(ITimingCache const &cache, bool ignoreMismatch) noexcept
Attach a timing cache to IBuilderConfig.
Definition: NvInfer.h:10041
char const * getPluginToSerialize(int32_t index) const noexcept
Get the plugin library path to be serialized with version-compatible engines.
Definition: NvInfer.h:10235
EngineCapability getEngineCapability() const noexcept
Query EngineCapability flow configured for the builder.
Definition: NvInfer.h:9552
RuntimePlatform getRuntimePlatform() const noexcept
Get the target platform for runtime execution.
Definition: NvInfer.h:10338
DeviceType getDefaultDeviceType() const noexcept
Get the default DeviceType which was set by setDefaultDeviceType.
Definition: NvInfer.h:9745
void setRuntimePlatform(RuntimePlatform runtimePlatform) noexcept
Set the target platform for runtime execution.
Definition: NvInfer.h:10326
int32_t getMaxNbTactics() const noexcept
Query the maximum number of tactics timed when there is a choice.
Definition: NvInfer.h:10362
BuilderFlags getFlags() const noexcept
Get the build mode flags for this builder config. Defaults to 0.
Definition: NvInfer.h:9603
void setFlags(BuilderFlags builderFlags) noexcept
Set the build mode flags to turn on builder options for this network.
Definition: NvInfer.h:9591
TacticSources getTacticSources() const noexcept
Get tactic sources.
Definition: NvInfer.h:9998
void resetDeviceType(ILayer const *layer) noexcept
reset the DeviceType for this layer
Definition: NvInfer.h:9688
void setDLACore(int32_t dlaCore) noexcept
Sets the DLA core used by the network. Defaults to -1.
Definition: NvInfer.h:9714
HardwareCompatibilityLevel getHardwareCompatibilityLevel() const noexcept
Get the hardware compatibility level.
Definition: NvInfer.h:10209
TRT_DEPRECATED QuantizationFlags getQuantizationFlags() const noexcept
Get the quantization flags.
Definition: NvInfer.h:9919
void clearFlag(BuilderFlag builderFlag) noexcept
clear a single build mode flag.
Definition: NvInfer.h:9615
int32_t addOptimizationProfile(IOptimizationProfile const *profile) noexcept
Add an optimization profile.
Definition: NvInfer.h:9796
IProgressMonitor * getProgressMonitor() const noexcept
Definition: NvInfer.h:10310
apiv::VBuilderConfig * mImpl
Definition: NvInfer.h:10425
TRT_DEPRECATED IOptimizationProfile const * getCalibrationProfile() noexcept
Get the current calibration profile.
Definition: NvInfer.h:9886
int32_t getAvgTimingIterations() const noexcept
Query the number of averaging iterations.
Definition: NvInfer.h:9527
void setDefaultDeviceType(DeviceType deviceType) noexcept
Sets the default DeviceType to be used by the builder. It ensures that all the layers that can run on...
Definition: NvInfer.h:9735
void setFlag(BuilderFlag builderFlag) noexcept
Set a single build mode flag.
Definition: NvInfer.h:9627
TRT_DEPRECATED bool setCalibrationProfile(IOptimizationProfile const *profile) noexcept
Add a calibration profile.
Definition: NvInfer.h:9874
virtual ~IBuilderConfig() noexcept=default
DeviceType getDeviceType(ILayer const *layer) const noexcept
Get the device that this layer executes on.
Definition: NvInfer.h:9666
bool canRunOnDLA(ILayer const *layer) const noexcept
Checks if a layer can run on DLA.
Definition: NvInfer.h:9698
TRT_DEPRECATED bool getQuantizationFlag(QuantizationFlag flag) const noexcept
Returns true if the quantization flag is set.
Definition: NvInfer.h:9961
cudaStream_t getProfileStream() const noexcept
Get the CUDA stream that is used to profile this network.
Definition: NvInfer.h:9779
void setHardwareCompatibilityLevel(HardwareCompatibilityLevel hardwareCompatibilityLevel) noexcept
Set the hardware compatibility level.
Definition: NvInfer.h:10196
TilingOptimizationLevel getTilingOptimizationLevel() const noexcept
Get the Tiling optimization level.
Definition: NvInfer.h:10390
TRT_DEPRECATED void setQuantizationFlag(QuantizationFlag flag) noexcept
Set a single quantization flag.
Definition: NvInfer.h:9947
void setMaxAuxStreams(int32_t nbStreams) noexcept
Set the maximum number of auxiliary streams that TRT is allowed to use.
Definition: NvInfer.h:10274
ProfilingVerbosity getProfilingVerbosity() const noexcept
Get verbosity level of layer information exposed in NVTX annotations and IEngineInspector.
Definition: NvInfer.h:9834
bool isDeviceTypeSet(ILayer const *layer) const noexcept
whether the DeviceType has been explicitly set for this layer
Definition: NvInfer.h:9678
void setBuilderOptimizationLevel(int32_t level) noexcept
Set builder optimization level.
Definition: NvInfer.h:10167
void setProfileStream(const cudaStream_t stream) noexcept
Set the CUDA stream that is used to profile this network.
Definition: NvInfer.h:9767
TRT_DEPRECATED IAlgorithmSelector * getAlgorithmSelector() const noexcept
Get Algorithm Selector.
Definition: NvInfer.h:9856
Builds an engine from a network definition.
Definition: NvInfer.h:10487
int32_t getMaxDLABatchSize() const noexcept
Get the maximum batch size DLA can support. For any tensor the total volume of index dimensions combi...
Definition: NvInfer.h:10518
int32_t getNbDLACores() const noexcept
Return the number of DLA engines available to this builder.
Definition: NvInfer.h:10526
IErrorRecorder * getErrorRecorder() const noexcept
get the ErrorRecorder assigned to this interface.
Definition: NvInfer.h:10633
apiv::VBuilder * mImpl
Definition: NvInfer.h:10768
ILogger * getLogger() const noexcept
get the logger with which the builder was created
Definition: NvInfer.h:10722
bool isNetworkSupported(INetworkDefinition const &network, IBuilderConfig const &config) const noexcept
Checks that a network is within the scope of the IBuilderConfig settings.
Definition: NvInfer.h:10712
int32_t getMaxThreads() const noexcept
get the maximum number of threads that can be used by the builder.
Definition: NvInfer.h:10752
IPluginRegistry & getPluginRegistry() noexcept
get the local plugin registry that can be used by the builder.
Definition: NvInfer.h:10762
TRT_DEPRECATED bool platformHasFastInt8() const noexcept
Determine whether the platform has fast native int8.
Definition: NvInfer.h:10506
nvinfer1::IOptimizationProfile * createOptimizationProfile() noexcept
Create a new optimization profile.
Definition: NvInfer.h:10599
void setGpuAllocator(IGpuAllocator *allocator) noexcept
Set the GPU allocator.
Definition: NvInfer.h:10544
nvinfer1::INetworkDefinition * createNetworkV2(NetworkDefinitionCreationFlags flags) noexcept
Create a network definition object.
Definition: NvInfer.h:10584
nvinfer1::IBuilderConfig * createBuilderConfig() noexcept
Create a builder configuration object.
Definition: NvInfer.h:10558
void reset() noexcept
Resets the builder state to default values.
Definition: NvInfer.h:10641
bool setMaxThreads(int32_t maxThreads) noexcept
Set the maximum number of threads.
Definition: NvInfer.h:10738
void setErrorRecorder(IErrorRecorder *recorder) noexcept
Set the ErrorRecorder for this interface.
Definition: NvInfer.h:10618
nvinfer1::IHostMemory * buildSerializedNetwork(INetworkDefinition &network, IBuilderConfig &config) noexcept
Builds and serializes a network for the given INetworkDefinition and IBuilderConfig.
Definition: NvInfer.h:10670
virtual ~IBuilder() noexcept=default
TRT_DEPRECATED bool platformHasTf32() const noexcept
Determine whether the platform has TF32 support.
Definition: NvInfer.h:10651
nvinfer1::ICudaEngine * buildEngineWithConfig(INetworkDefinition &network, IBuilderConfig &config) noexcept
Builds a network for the given INetworkDefinition and IBuilderConfig.
Definition: NvInfer.h:10690
A cast layer in a network.
Definition: NvInfer.h:3857
virtual ~ICastLayer() noexcept=default
apiv::VCastLayer * mImpl
Definition: NvInfer.h:3883
DataType getToType() const noexcept
Return cast layer output type.
Definition: NvInfer.h:3877
void setToType(DataType toType) noexcept
Set cast layer output type.
Definition: NvInfer.h:3866
A concatenation layer in a network definition.
Definition: NvInfer.h:2092
void setAxis(int32_t axis) noexcept
Set the axis along which concatenation occurs.
Definition: NvInfer.h:2105
int32_t getAxis() const noexcept
Get the axis along which concatenation occurs.
Definition: NvInfer.h:2115
virtual ~IConcatenationLayer() noexcept=default
This layer represents a condition input to an IIfConditional.
Definition: NvInfer.h:4520
virtual ~IConditionLayer() noexcept=default
Layer that represents a constant value.
Definition: NvInfer.h:3896
void setWeights(Weights weights) noexcept
Set the weights for the layer.
Definition: NvInfer.h:3906
Weights getWeights() const noexcept
Get the weights for the layer.
Definition: NvInfer.h:3916
void setDimensions(Dims const &dimensions) noexcept
Set the dimensions for the layer.
Definition: NvInfer.h:3928
apiv::VConstantLayer * mImpl
Definition: NvInfer.h:3946
virtual ~IConstantLayer() noexcept=default
Dims getDimensions() const noexcept
Get the dimensions for the layer.
Definition: NvInfer.h:3940
A convolution layer in a network definition.
Definition: NvInfer.h:1062
Dims getPrePadding() const noexcept
Get the pre-padding.
Definition: NvInfer.h:1187
Weights getBiasWeights() const noexcept
Get the bias weights for the convolution.
Definition: NvInfer.h:1160
void setPaddingMode(PaddingMode paddingMode) noexcept
Set the padding mode.
Definition: NvInfer.h:1228
void setDilationNd(Dims const &dilation) noexcept
Set the multi-dimension dilation of the convolution.
Definition: NvInfer.h:1332
Dims getPaddingNd() const noexcept
Get the multi-dimension padding of the convolution.
Definition: NvInfer.h:1318
Dims getStrideNd() const noexcept
Get the multi-dimension stride of the convolution.
Definition: NvInfer.h:1288
Weights getKernelWeights() const noexcept
Get the kernel weights of the convolution.
Definition: NvInfer.h:1135
void setStrideNd(Dims const &stride) noexcept
Set the multi-dimension stride of the convolution.
Definition: NvInfer.h:1278
Dims getDilationNd() const noexcept
Get the multi-dimension dilation of the convolution.
Definition: NvInfer.h:1342
int64_t getNbOutputMaps() const noexcept
Get the number of output maps for the convolution.
Definition: NvInfer.h:1081
void setKernelWeights(Weights weights) noexcept
Set the kernel weights for the convolution.
Definition: NvInfer.h:1125
Dims getPostPadding() const noexcept
Get the post-padding.
Definition: NvInfer.h:1214
int64_t getNbGroups() const noexcept
Get the number of groups of the convolution.
Definition: NvInfer.h:1111
PaddingMode getPaddingMode() const noexcept
Get the padding mode.
Definition: NvInfer.h:1240
virtual ~IConvolutionLayer() noexcept=default
void setNbGroups(int64_t nbGroups) noexcept
Set the number of groups for a convolution.
Definition: NvInfer.h:1101
void setNbOutputMaps(int64_t nbOutputMaps) noexcept
Set the number of output maps for the convolution.
Definition: NvInfer.h:1071
void setBiasWeights(Weights weights) noexcept
Set the bias weights for the convolution.
Definition: NvInfer.h:1150
Dims getKernelSizeNd() const noexcept
Get the multi-dimension kernel size of the convolution.
Definition: NvInfer.h:1263
void setPaddingNd(Dims const &padding) noexcept
Set the multi-dimension padding of the convolution.
Definition: NvInfer.h:1306
void setPrePadding(Dims const &padding) noexcept
Set the multi-dimension pre-padding of the convolution.
Definition: NvInfer.h:1177
void setPostPadding(Dims const &padding) noexcept
Set the multi-dimension post-padding of the convolution.
Definition: NvInfer.h:1204
void setKernelSizeNd(Dims const &kernelSize) noexcept
Set the multi-dimension kernel size of the convolution.
Definition: NvInfer.h:1253
An engine for executing inference on a built network, with functionally unsafe features.
Definition: NvInferRuntime.h:3120
Layer that represents a cumulative operation across a tensor.
Definition: NvInfer.h:6586
bool setOperation(CumulativeOperation op) noexcept
Set the cumulative operation for the layer.
Definition: NvInfer.h:6597
void setReverse(bool reverse) noexcept
Specify whether the cumulative operation should be applied backward.
Definition: NvInfer.h:6645
apiv::VCumulativeLayer * mImpl
Definition: NvInfer.h:6663
bool getExclusive() const noexcept
Get whether it is exclusive accumulation or inclusive accumulation.
Definition: NvInfer.h:6633
virtual ~ICumulativeLayer() noexcept=default
bool getReverse() const noexcept
Get the boolean that specifies whether the cumulative operation should be applied backward.
Definition: NvInfer.h:6657
void setExclusive(bool exclusive) noexcept
Set whether it is an exclusive accumulation or inclusive accumulation.
Definition: NvInfer.h:6621
CumulativeOperation getOperation() const noexcept
Get the cumulative operation for the layer.
Definition: NvInfer.h:6609
A deconvolution layer in a network definition.
Definition: NvInfer.h:2133
void setBiasWeights(Weights weights) noexcept
Set the bias weights for the deconvolution.
Definition: NvInfer.h:2221
int64_t getNbGroups() const noexcept
Get the number of groups for a deconvolution.
Definition: NvInfer.h:2182
Weights getKernelWeights() const noexcept
Get the kernel weights for the deconvolution.
Definition: NvInfer.h:2206
void setPrePadding(Dims const &padding) noexcept
Set the multi-dimension pre-padding of the deconvolution.
Definition: NvInfer.h:2248
Dims getStrideNd() const noexcept
Get the multi-dimension stride of the deconvolution.
Definition: NvInfer.h:2363
Dims getDilationNd() const noexcept
Get the multi-dimension dilation of the deconvolution.
Definition: NvInfer.h:2429
Weights getBiasWeights() const noexcept
Get the bias weights for the deconvolution.
Definition: NvInfer.h:2231
void setKernelWeights(Weights weights) noexcept
Set the kernel weights for the deconvolution.
Definition: NvInfer.h:2196
int64_t getNbOutputMaps() const noexcept
Get the number of output feature maps for the deconvolution.
Definition: NvInfer.h:2152
void setStrideNd(Dims const &stride) noexcept
Set the multi-dimension stride of the deconvolution.
Definition: NvInfer.h:2353
Dims getPostPadding() const noexcept
Get the padding.
Definition: NvInfer.h:2285
Dims getKernelSizeNd() const noexcept
Get the multi-dimension kernel size of the deconvolution.
Definition: NvInfer.h:2336
void setPostPadding(Dims const &padding) noexcept
Set the multi-dimension post-padding of the deconvolution.
Definition: NvInfer.h:2275
void setKernelSizeNd(Dims const &kernelSize) noexcept
Set the multi-dimension kernel size of the deconvolution.
Definition: NvInfer.h:2326
virtual ~IDeconvolutionLayer() noexcept=default
void setPaddingNd(Dims const &padding) noexcept
Set the multi-dimension padding of the deconvolution.
Definition: NvInfer.h:2381
void setNbOutputMaps(int64_t nbOutputMaps) noexcept
Set the number of output feature maps for the deconvolution.
Definition: NvInfer.h:2142
Dims getPaddingNd() const noexcept
Get the multi-dimension padding of the deconvolution.
Definition: NvInfer.h:2393
void setDilationNd(Dims const &dilation) noexcept
Set the multi-dimension dilation of the deconvolution.
Definition: NvInfer.h:2419
void setPaddingMode(PaddingMode paddingMode) noexcept
Set the padding mode.
Definition: NvInfer.h:2299
void setNbGroups(int64_t nbGroups) noexcept
Set the number of groups for a deconvolution.
Definition: NvInfer.h:2172
Dims getPrePadding() const noexcept
Get the pre-padding.
Definition: NvInfer.h:2258
PaddingMode getPaddingMode() const noexcept
Get the padding mode.
Definition: NvInfer.h:2311
A Dequantize layer in a network definition.
Definition: NvInfer.h:5584
void setToType(DataType toType) noexcept
Set the Dequantize layer output type.
Definition: NvInfer.h:5621
virtual ~IDequantizeLayer() noexcept=default
int32_t getAxis() const noexcept
Get the quantization axis.
Definition: NvInfer.h:5594
DataType getToType() const noexcept
Return the Dequantize layer output type.
Definition: NvInfer.h:5633
void setAxis(int32_t axis) noexcept
Set the quantization axis.
Definition: NvInfer.h:5605
A network layer to perform dynamic quantization.
Definition: NvInfer.h:5661
int32_t getAxis() const noexcept
Get the axis along which blocking occurs.
Definition: NvInfer.h:5750
int32_t getBlockSize() const noexcept
Get the size of the quantization block.
Definition: NvInfer.h:5773
DataType getScaleType() const noexcept
Return the scale factors data type.
Definition: NvInfer.h:5727
void setScaleType(DataType scaleType) noexcept
Set the data type of the scale factors used to quantize the data.
Definition: NvInfer.h:5714
DataType getToType() const noexcept
Return DynamicQuantizeLayer's quantized output type.
Definition: NvInfer.h:5701
virtual ~IDynamicQuantizeLayer() noexcept=default
void setToType(DataType toType) noexcept
Set DynamicQuantizeLayer's quantized output type.
Definition: NvInfer.h:5688
void setAxis(int32_t axis) noexcept
Set the axis along which block quantization occurs.
Definition: NvInfer.h:5740
void setBlockSize(int32_t size) noexcept
Set the size of the quantization block.
Definition: NvInfer.h:5763
An Einsum layer in a network.
Definition: NvInfer.h:5818
bool setEquation(char const *equation) noexcept
Set the equation. The equation is a comma-separated list of subscript labels, where each label refers...
Definition: NvInfer.h:5829
virtual ~IEinsumLayer() noexcept=default
char const * getEquation() const noexcept
Return the equation.
Definition: NvInfer.h:5839
A elementwise layer in a network definition.
Definition: NvInfer.h:2503
virtual ~IElementWiseLayer() noexcept=default
apiv::VElementWiseLayer * mImpl
Definition: NvInfer.h:2532
ElementWiseOperation getOperation() const noexcept
Get the binary operation for the layer.
Definition: NvInfer.h:2526
void setOperation(ElementWiseOperation op) noexcept
Set the binary operation for the layer.
Definition: NvInfer.h:2514
Generate a tensor according to a specified mode.
Definition: NvInfer.h:5107
bool isAlphaBetaInt64() const noexcept
Return true if alpha/beta have type int64, false if they have type double.
Definition: NvInfer.h:5339
FillOperation getOperation() const noexcept
Get the fill operation for the layer.
Definition: NvInfer.h:5153
void setOperation(FillOperation op) noexcept
Set the fill operation for the layer.
Definition: NvInfer.h:5143
DataType getToType() const noexcept
Get the fill layer output type.
Definition: NvInfer.h:5368
void setAlphaInt64(int64_t alpha) noexcept
Set the alpha parameter with int64 datatype.
Definition: NvInfer.h:5282
void setBetaInt64(int64_t beta) noexcept
Set the beta parameter with int64 datatype.
Definition: NvInfer.h:5316
void setBeta(double beta) noexcept
Set the beta parameter.
Definition: NvInfer.h:5206
int64_t getAlphaInt64() const noexcept
Get the value of alpha parameter with int64 datatype.
Definition: NvInfer.h:5297
int64_t getBetaInt64() const noexcept
Get the value of beta parameter with int64 datatype.
Definition: NvInfer.h:5331
double getAlpha() const noexcept
Get the value of alpha parameter.
Definition: NvInfer.h:5187
void setDimensions(Dims const &dimensions) noexcept
Set the output tensor's dimensions.
Definition: NvInfer.h:5118
void setAlpha(double alpha) noexcept
Set the alpha parameter.
Definition: NvInfer.h:5172
void setToType(DataType toType) noexcept
Set the fill layer output type.
Definition: NvInfer.h:5356
Dims getDimensions() const noexcept
Get the output tensor's dimensions.
Definition: NvInfer.h:5133
double getBeta() const noexcept
Get the value of beta parameter.
Definition: NvInfer.h:5221
virtual ~IFillLayer() noexcept=default
A Gather layer in a network definition. Supports several kinds of gathering.
Definition: NvInfer.h:2636
void setGatherAxis(int32_t axis) noexcept
Set the axis used by GatherMode::kELEMENTS and GatherMode::kDEFAULT The axis must be less than the nu...
Definition: NvInfer.h:2647
void setNbElementWiseDims(int32_t elementWiseDims) noexcept
Set the number of leading dimensions of indices tensor to be handled elementwise.
Definition: NvInfer.h:2682
apiv::VGatherLayer * mImpl
Definition: NvInfer.h:2718
int32_t getNbElementWiseDims() const noexcept
Get the number of leading dimensions of indices tensor to be handled elementwise.
Definition: NvInfer.h:2692
void setMode(GatherMode mode) noexcept
Set the gather mode.
Definition: NvInfer.h:2702
int32_t getGatherAxis() const noexcept
Get the axis to gather on.
Definition: NvInfer.h:2659
GatherMode getMode() const noexcept
Get the gather mode.
Definition: NvInfer.h:2712
virtual ~IGatherLayer() noexcept=default
A GridSample layer in a network definition.
Definition: NvInfer.h:6040
void setInterpolationMode(InterpolationMode mode) noexcept
Set the grid sample interpolation mode.
Definition: NvInfer.h:6047
bool setSampleMode(SampleMode mode) noexcept
Set the sample mode.
Definition: NvInfer.h:6093
void setAlignCorners(bool alignCorners) noexcept
Set the align corners mode.
Definition: NvInfer.h:6069
apiv::VGridSampleLayer * mImpl
Definition: NvInfer.h:6111
SampleMode getSampleMode() const noexcept
Get the sample mode.
Definition: NvInfer.h:6105
InterpolationMode getInterpolationMode() const noexcept
Get the grid sample interpolation mode.
Definition: NvInfer.h:6059
bool getAlignCorners() const noexcept
Get the align corners mode.
Definition: NvInfer.h:6081
virtual ~IGridSampleLayer() noexcept=default
Class to handle library allocated memory that is accessible to the user.
Definition: NvInferRuntime.h:142
A layer that represents the identity function.
Definition: NvInfer.h:3844
apiv::VIdentityLayer * mImpl
Definition: NvInfer.h:3846
virtual ~IIdentityLayer() noexcept=default
This is a base class for Conditional boundary layers.
Definition: NvInfer.h:4499
IIfConditional * getConditional() const noexcept
Get a pointer to the IIfConditional associated with this boundary layer.
Definition: NvInfer.h:4504
virtual ~IIfConditionalBoundaryLayer() noexcept=default
Helper for constructing conditionally-executed subgraphs.
Definition: NvInfer.h:4582
IIfConditionalInputLayer * addInput(ITensor &input) noexcept
Add an If-conditional input.
Definition: NvInfer.h:4623
char const * getName() const noexcept
Return the name of the conditional.
Definition: NvInfer.h:4648
virtual ~IIfConditional() noexcept=default
IConditionLayer * setCondition(ITensor &condition) noexcept
Set the condition tensor for this If-Conditional construct.
Definition: NvInfer.h:4593
IIfConditionalOutputLayer * addOutput(ITensor &trueSubgraphOutput, ITensor &falseSubgraphOutput) noexcept
Add an If-conditional output.
Definition: NvInfer.h:4611
void setName(char const *name) noexcept
Set the name of the conditional.
Definition: NvInfer.h:4638
This layer represents an output of an IIfConditional.
Definition: NvInfer.h:4537
virtual ~IIfConditionalOutputLayer() noexcept=default
Application-implemented interface for calibration.
Definition: NvInfer.h:8215
virtual TRT_DEPRECATED int32_t getBatchSize() const noexcept=0
Get the batch size used for calibration batches.
A layer to do iterations.
Definition: NvInfer.h:4813
virtual ~IIteratorLayer() noexcept=default
void setReverse(bool reverse) noexcept
Set iteration order to be reverse.
Definition: NvInfer.h:4840
bool getReverse() const noexcept
Check if the iteration order is reverse.
Definition: NvInfer.h:4850
int32_t getAxis() const noexcept
Get axis being iterated over.
Definition: NvInfer.h:4826
void setAxis(int32_t axis) noexcept
Set axis to iterate over.
Definition: NvInfer.h:4818
A LRN layer in a network definition.
Definition: NvInfer.h:1747
int64_t getWindowSize() const noexcept
Get the LRN window size.
Definition: NvInfer.h:1768
float getAlpha() const noexcept
Get the LRN alpha value.
Definition: NvInfer.h:1790
void setWindowSize(int64_t windowSize) noexcept
Set the LRN window size.
Definition: NvInfer.h:1758
void setK(float k) noexcept
Set the LRN K value.
Definition: NvInfer.h:1824
void setAlpha(float alpha) noexcept
Set the LRN alpha value.
Definition: NvInfer.h:1780
void setBeta(float beta) noexcept
Set the LRN beta value.
Definition: NvInfer.h:1802
virtual ~ILRNLayer() noexcept=default
float getBeta() const noexcept
Get the LRN beta value.
Definition: NvInfer.h:1812
float getK() const noexcept
Get the LRN K value.
Definition: NvInfer.h:1834
Base class for all layer classes in a network definition.
Definition: NvInfer.h:577
TRT_DEPRECATED void setPrecision(DataType dataType) noexcept
Set the preferred or required computational precision of this layer in a weakly-typed network.
Definition: NvInfer.h:697
TRT_DEPRECATED void setOutputType(int32_t index, DataType dataType) noexcept
Set the output type of this layer in a weakly-typed network.
Definition: NvInfer.h:785
TRT_DEPRECATED bool precisionIsSet() const noexcept
whether the computational precision has been set for this layer
Definition: NvInfer.h:723
void setMetadata(char const *metadata) noexcept
Set the metadata for this layer.
Definition: NvInfer.h:848
TRT_DEPRECATED void resetOutputType(int32_t index) noexcept
reset the output type for this layer
Definition: NvInfer.h:830
void setName(char const *name) noexcept
Set the name of a layer.
Definition: NvInfer.h:598
int32_t getNbInputs() const noexcept
Get the number of inputs of a layer.
Definition: NvInfer.h:616
char const * getMetadata() const noexcept
Get the metadata of the layer.
Definition: NvInfer.h:861
DataType getOutputType(int32_t index) const noexcept
get the output type of this layer
Definition: NvInfer.h:800
DataType getPrecision() const noexcept
get the computational precision of this layer
Definition: NvInfer.h:709
TRT_DEPRECATED bool outputTypeIsSet(int32_t index) const noexcept
whether the output type has been set for this layer
Definition: NvInfer.h:816
char const * getName() const noexcept
Return the name of a layer.
Definition: NvInfer.h:608
int32_t getNbOutputs() const noexcept
Get the number of outputs of a layer.
Definition: NvInfer.h:637
ITensor * getOutput(int32_t index) const noexcept
Get the layer output corresponding to the given index.
Definition: NvInfer.h:647
void setInput(int32_t index, ITensor &tensor) noexcept
Replace an input of this layer with a specific tensor.
Definition: NvInfer.h:664
ITensor * getInput(int32_t index) const noexcept
Get the layer input corresponding to the given index.
Definition: NvInfer.h:629
LayerType getType() const noexcept
Return the type of a layer.
Definition: NvInfer.h:584
TRT_DEPRECATED void resetPrecision() noexcept
reset the computational precision for this layer
Definition: NvInfer.h:735
virtual ~ILayer() noexcept=default
Application-implemented logging interface for the builder, refitter and runtime.
Definition: NvInferRuntime.h:1542
This is a base class for Loop boundary layers.
Definition: NvInfer.h:4476
virtual ~ILoopBoundaryLayer() noexcept=default
ILoop * getLoop() const noexcept
Get a pointer to ILoop associated with this boundary layer.
Definition: NvInfer.h:4481
Helper for creating a recurrent subgraph.
Definition: NvInfer.h:4871
void setName(char const *name) noexcept
Set the name of the loop.
Definition: NvInfer.h:4941
ITripLimitLayer * addTripLimit(ITensor &tensor, TripLimit limit) noexcept
Add a trip-count limiter, based on the given tensor.
Definition: NvInfer.h:4900
IIteratorLayer * addIterator(ITensor &tensor, int32_t axis=0, bool reverse=false) noexcept
Return layer that subscripts tensor by loop iteration.
Definition: NvInfer.h:4913
ILoopOutputLayer * addLoopOutput(ITensor &tensor, LoopOutput outputKind, int32_t axis=0) noexcept
Make an output for this loop, based on the given tensor.
Definition: NvInfer.h:4926
virtual ~ILoop() noexcept=default
char const * getName() const noexcept
Return the name of the loop.
Definition: NvInfer.h:4951
IRecurrenceLayer * addRecurrence(ITensor &initialValue) noexcept
Create a recurrence layer for this loop with initialValue as its first input.
Definition: NvInfer.h:4879
An ILoopOutputLayer is the sole way to get output from a loop.
Definition: NvInfer.h:4713
virtual ~ILoopOutputLayer() noexcept=default
int32_t getAxis() const noexcept
Get axis being concatenated over.
Definition: NvInfer.h:4743
LoopOutput getLoopOutput() const noexcept
Get which kind a loop output has.
Definition: NvInfer.h:4718
void setAxis(int32_t axis) noexcept
Set where to insert the contenation axis. Ignored if getLoopOutput() is kLAST_VALUE.
Definition: NvInfer.h:4735
Layer that represents a Matrix Multiplication.
Definition: NvInfer.h:3719
apiv::VMatrixMultiplyLayer * mImpl
Definition: NvInfer.h:3747
virtual ~IMatrixMultiplyLayer() noexcept=default
MatrixOperation getOperation(int32_t index) const noexcept
Get the operation for an input tensor.
Definition: NvInfer.h:3741
void setOperation(int32_t index, MatrixOperation op) noexcept
Set the operation for an input tensor.
Definition: NvInfer.h:3729
A non-maximum suppression layer in a network definition.
Definition: NvInfer.h:6188
virtual ~INMSLayer() noexcept=default
void setTopKBoxLimit(int32_t limit) noexcept
Set the TopK box limit parameter for the layer.
Definition: NvInfer.h:6225
void setBoundingBoxFormat(BoundingBoxFormat fmt) noexcept
Set the bounding box format parameter for the layer.
Definition: NvInfer.h:6199
BoundingBoxFormat getBoundingBoxFormat() const noexcept
Get the bounding box format parameter for the layer.
Definition: NvInfer.h:6211
apiv::VNMSLayer * mImpl
Definition: NvInfer.h:6261
int32_t getTopKBoxLimit() const noexcept
Get the TopK box limit parameter for the layer.
Definition: NvInfer.h:6235
A network definition for input to the builder.
Definition: NvInfer.h:6685
IConcatenationLayer * addConcatenation(ITensor *const *inputs, int32_t nbInputs) noexcept
Add a concatenation layer to the network.
Definition: NvInfer.h:6913
IShuffleLayer * addShuffle(ITensor &input) noexcept
Add a shuffle layer to the network.
Definition: NvInfer.h:6976
INormalizationLayer * addNormalization(ITensor &input, ITensor &scale, ITensor &bias, uint32_t axesMask) noexcept
Add a normalization layer to the network.
Definition: NvInfer.h:8055
void setName(char const *name) noexcept
Sets the name of the network.
Definition: NvInfer.h:7386
bool markDebug(ITensor &tensor) noexcept
Mark a tensor as a debug tensor.
Definition: NvInfer.h:6756
ILRNLayer * addLRN(ITensor &input, int64_t window, float alpha, float beta, float k) noexcept
Add a LRN layer to the network.
Definition: NvInfer.h:6857
ITopKLayer * addTopK(ITensor &input, TopKOperation op, int32_t k, uint32_t reduceAxes) noexcept
Add a TopK layer to the network.
Definition: NvInfer.h:7136
ICumulativeLayer * addCumulative(ITensor &input, ITensor &axis, CumulativeOperation operation, bool exclusive, bool reverse) noexcept
Add a cumulative layer to the network.
Definition: NvInfer.h:8077
IAssertionLayer * addAssertion(ITensor &condition, char const *message) noexcept
Add an assertion layer to the network.
Definition: NvInfer.h:7702
IConvolutionLayer * addConvolutionNd(ITensor &input, int64_t nbOutputMaps, Dims const &kernelSize, Weights kernelWeights, Weights biasWeights) noexcept
Add a multi-dimension convolution layer to the network.
Definition: NvInfer.h:7521
ICastLayer * addCast(ITensor &input, DataType toType) noexcept
Add a cast layer.
Definition: NvInfer.h:7276
IScaleLayer * addScaleNd(ITensor &input, ScaleMode mode, Weights shift, Weights scale, Weights power, int32_t channelAxis) noexcept
Add a multi-dimension scale layer to the network.
Definition: NvInfer.h:7600
char const * getName() const noexcept
Returns the name associated with the network.
Definition: NvInfer.h:7400
IParametricReLULayer * addParametricReLU(ITensor &input, ITensor &slope) noexcept
Add a parametric ReLU layer to the network.
Definition: NvInfer.h:7499
ITensor * getOutput(int32_t index) const noexcept
Get the output tensor specified by the given index.
Definition: NvInfer.h:7077
ITensor * getInput(int32_t index) const noexcept
Get the input tensor specified by the given index.
Definition: NvInfer.h:7047
IDequantizeLayer * addDequantize(ITensor &input, ITensor &scale, DataType outputType) noexcept
Add a dequantization layer to the network.
Definition: NvInfer.h:7870
bool unmarkOutputForShapes(ITensor &tensor) noexcept
Undo markOutputForShapes.
Definition: NvInfer.h:7481
IFillLayer * addFill(Dims const &dimensions, FillOperation op, DataType outputType) noexcept
Add a fill layer to the network.
Definition: NvInfer.h:7753
ILoop * addLoop() noexcept
Add a loop to the network.
Definition: NvInfer.h:7631
IDynamicQuantizeLayer * addDynamicQuantize(ITensor &input, int32_t axis, int32_t blockSize, DataType outputType, DataType scaleType) noexcept
Add a dynamic quantization layer to the network.
Definition: NvInfer.h:7960
bool markUnfusedTensorsAsDebugTensors() noexcept
Mark unfused tensors as debug tensors.
Definition: NvInfer.h:6804
IActivationLayer * addActivation(ITensor &input, ActivationType type) noexcept
Add an activation layer to the network.
Definition: NvInfer.h:6838
TRT_DEPRECATED IFillLayer * addFill(Dims const &dimensions, FillOperation op) noexcept
Add a fill layer to the network.
Definition: NvInfer.h:7727
ISliceLayer * addSlice(ITensor &input, Dims const &start, Dims const &size, Dims const &stride) noexcept
Add a slice layer to the network.
Definition: NvInfer.h:7362
virtual ~INetworkDefinition() noexcept=default
TRT_DEPRECATED IQuantizeLayer * addQuantize(ITensor &input, ITensor &scale) noexcept
Add a quantization layer to the network.
Definition: NvInfer.h:7911
virtual IBuilder & getBuilder() const noexcept
Return the builder from which this INetworkDefinition was created.
Definition: NvInfer.h:8088
INMSLayer * addNMS(ITensor &boxes, ITensor &scores, ITensor &maxOutputBoxesPerClass) noexcept
Add a non-maximum suppression layer to the network.
Definition: NvInfer.h:8012
ILayer * getLayer(int32_t index) const noexcept
Get the layer specified by the given index.
Definition: NvInfer.h:7019
bool getFlag(NetworkDefinitionCreationFlag networkDefinitionCreationFlag) const noexcept
Returns true if the network definition creation flag is set.
Definition: NvInfer.h:7452
IIfConditional * addIfConditional() noexcept
Add an if-then-else to the network.
Definition: NvInfer.h:7646
IErrorRecorder * getErrorRecorder() const noexcept
get the ErrorRecorder assigned to this interface.
Definition: NvInfer.h:7827
ISqueezeLayer * addSqueeze(ITensor &input, ITensor &axes) noexcept
Add a squeeze layer to the network.
Definition: NvInfer.h:8145
IReverseSequenceLayer * addReverseSequence(ITensor &input, ITensor &sequenceLens) noexcept
Add a ReverseSequence layer to the network.
Definition: NvInfer.h:8029
int32_t getNbInputs() const noexcept
Get the number of inputs in the network.
Definition: NvInfer.h:7031
NetworkDefinitionCreationFlags getFlags() const noexcept
Get the network definition creation flags for this network definition object. Defaults to 0.
Definition: NvInfer.h:7440
IQuantizeLayer * addQuantize(ITensor &input, ITensor &scale, DataType outputType) noexcept
Add a quantization layer to the network.
Definition: NvInfer.h:7933
IReduceLayer * addReduce(ITensor &input, ReduceOperation operation, uint32_t reduceAxes, bool keepDimensions) noexcept
Add a reduce layer to the network.
Definition: NvInfer.h:7103
IUnaryLayer * addUnary(ITensor &input, UnaryOperation operation) noexcept
Add a unary layer to the network.
Definition: NvInfer.h:6962
IGridSampleLayer * addGridSample(ITensor &input, ITensor &grid) noexcept
Add a GridSample layer to the network.
Definition: NvInfer.h:7994
void removeTensor(ITensor &tensor) noexcept
remove a tensor from the network definition.
Definition: NvInfer.h:7291
bool areWeightsMarkedRefittable(char const *name) const noexcept
Whether the weight has been marked as refittable.
Definition: NvInfer.h:8126
ISelectLayer * addSelect(ITensor &condition, ITensor &thenInput, ITensor &elseInput) noexcept
Add a select layer to the network.
Definition: NvInfer.h:7685
IScatterLayer * addScatter(ITensor &data, ITensor &indices, ITensor &updates, ScatterMode mode) noexcept
Add a Scatter layer to the network with specified mode and axis=0.
Definition: NvInfer.h:7890
int32_t getNbLayers() const noexcept
Get the number of layers in the network.
Definition: NvInfer.h:7005
TRT_DEPRECATED bool hasImplicitBatchDimension() const noexcept
Query whether the network was created with an implicit batch dimension.
Definition: NvInfer.h:7430
apiv::VNetworkDefinition * mImpl
Definition: NvInfer.h:8172
bool markOutputForShapes(ITensor &tensor) noexcept
Enable tensor's value to be computed by IExecutionContext::getShapeBinding.
Definition: NvInfer.h:7469
IOneHotLayer * addOneHot(ITensor &indices, ITensor &values, ITensor &depth, int32_t axis) noexcept
Add a OneHot layer to the network.
Definition: NvInfer.h:6993
IScaleLayer * addScale(ITensor &input, ScaleMode mode, Weights shift, Weights scale, Weights power) noexcept
Add a Scale layer to the network.
Definition: NvInfer.h:6883
IPluginV3Layer * addPluginV3(ITensor *const *inputs, int32_t nbInputs, ITensor *const *shapeInputs, int32_t nbShapeInputs, IPluginV3 &plugin) noexcept
Add a plugin layer implementing the IPluginV3 interface to the network.
Definition: NvInfer.h:7342
void unmarkOutput(ITensor &tensor) noexcept
unmark a tensor as a network output.
Definition: NvInfer.h:7303
IIdentityLayer * addIdentity(ITensor &input) noexcept
Add an identity layer.
Definition: NvInfer.h:7261
IGatherLayer * addGatherV2(ITensor &data, ITensor &indices, GatherMode mode) noexcept
Add gather with specified mode, axis=0 and nbElementWiseDims=0.
Definition: NvInfer.h:7168
IElementWiseLayer * addElementWise(ITensor &input1, ITensor &input2, ElementWiseOperation op) noexcept
Add an elementwise layer to the network.
Definition: NvInfer.h:6940
IConstantLayer * addConstant(Dims const &dimensions, Weights weights) noexcept
Add a constant layer to the network.
Definition: NvInfer.h:7247
void setErrorRecorder(IErrorRecorder *recorder) noexcept
Set the ErrorRecorder for this interface.
Definition: NvInfer.h:7812
IPoolingLayer * addPoolingNd(ITensor &input, PoolingType type, Dims const &windowSize) noexcept
Add a multi-dimension pooling layer to the network.
Definition: NvInfer.h:7541
IRaggedSoftMaxLayer * addRaggedSoftMax(ITensor &input, ITensor &bounds) noexcept
Add a RaggedSoftMax layer to the network.
Definition: NvInfer.h:7187
IShapeLayer * addShape(ITensor &input) noexcept
Add a shape layer to the network.
Definition: NvInfer.h:7416
IGatherLayer * addGather(ITensor &data, ITensor &indices, int32_t axis) noexcept
Add gather with mode GatherMode::kDEFAULT and specified axis and nbElementWiseDims=0.
Definition: NvInfer.h:7152
bool unmarkWeightsRefittable(char const *name) noexcept
Unmark weights as refittable when the builder flag kREFIT_INDIVIDUAL is set.
Definition: NvInfer.h:8113
bool markWeightsRefittable(char const *name) noexcept
Mark weights as refittable when the builder flag kREFIT_INDIVIDUAL is set.
Definition: NvInfer.h:8101
IDeconvolutionLayer * addDeconvolutionNd(ITensor &input, int64_t nbOutputMaps, Dims kernelSize, Weights kernelWeights, Weights biasWeights) noexcept
Add a multi-dimension deconvolution layer to the network.
Definition: NvInfer.h:7563
IResizeLayer * addResize(ITensor &input) noexcept
Add a resize layer to the network.
Definition: NvInfer.h:7617
IUnsqueezeLayer * addUnsqueeze(ITensor &input, ITensor &axes) noexcept
Add an unsqueeze layer to the network.
Definition: NvInfer.h:8166
IMatrixMultiplyLayer * addMatrixMultiply(ITensor &input0, MatrixOperation op0, ITensor &input1, MatrixOperation op1) noexcept
Add a MatrixMultiply layer to the network.
Definition: NvInfer.h:7208
ISoftMaxLayer * addSoftMax(ITensor &input) noexcept
Add a SoftMax layer to the network.
Definition: NvInfer.h:6896
bool isDebugTensor(nvinfer1::ITensor const &tensor) const noexcept
Check if a tensor is marked as debug tensor.
Definition: NvInfer.h:6782
bool unmarkDebug(ITensor &tensor) noexcept
Unmark a tensor as a debug tensor.
Definition: NvInfer.h:6772
IEinsumLayer * addEinsum(ITensor *const *inputs, int32_t nbInputs, char const *equation) noexcept
Add an Einsum layer to the network.
Definition: NvInfer.h:7976
void markOutput(ITensor &tensor) noexcept
Mark a tensor as a network output.
Definition: NvInfer.h:6738
TRT_DEPRECATED IPluginV2Layer * addPluginV2(ITensor *const *inputs, int32_t nbInputs, IPluginV2 &plugin) noexcept
Add a plugin layer to the network using the IPluginV2 interface.
Definition: NvInfer.h:7324
IPaddingLayer * addPaddingNd(ITensor &input, Dims const &prePadding, Dims const &postPadding) noexcept
Add a padding layer to the network. Only 2D padding is currently supported.
Definition: NvInfer.h:7769
INonZeroLayer * addNonZero(ITensor &input) noexcept
Add a nonzero layer to the network.
Definition: NvInfer.h:7223
TRT_DEPRECATED IDequantizeLayer * addDequantize(ITensor &input, ITensor &scale) noexcept
Add a dequantization layer to the network.
Definition: NvInfer.h:7848
int32_t getNbOutputs() const noexcept
Get the number of outputs in the network.
Definition: NvInfer.h:7061
bool setWeightsName(Weights weights, char const *name) noexcept
Associate a name with all current uses of the given weights.
Definition: NvInfer.h:7793
bool unmarkUnfusedTensorsAsDebugTensors() noexcept
Undo the marking of unfused tensors as debug tensors.
Definition: NvInfer.h:6818
Forward declaration of IEngineInspector for use by other interfaces.
Definition: NvInferRuntime.h:51
Definition: NvInfer.h:3773
virtual ~INonZeroLayer() noexcept=default
A normalization layer in a network definition.
Definition: NvInfer.h:6350
float getEpsilon() const noexcept
Get the epsilon value used for the normalization calculation.
Definition: NvInfer.h:6369
uint32_t getAxes() const noexcept
Get the axes value used for the normalization calculation.
Definition: NvInfer.h:6389
virtual ~INormalizationLayer() noexcept=default
void setEpsilon(float eps) noexcept
Set the epsilon value used for the normalization calculation.
Definition: NvInfer.h:6359
DataType getComputePrecision() const noexcept
Get the compute precision of this layer.
Definition: NvInfer.h:6456
apiv::VNormalizationLayer * mImpl
Definition: NvInfer.h:6462
int64_t getNbGroups() const noexcept
Get the number of groups used to split the channels for the normalization calculation.
Definition: NvInfer.h:6420
void setAxes(uint32_t axesMask) noexcept
Set the reduction axes for the normalization calculation.
Definition: NvInfer.h:6379
void setComputePrecision(DataType type) noexcept
Set the compute precision of this layer.
Definition: NvInfer.h:6446
void setNbGroups(int64_t nbGroups) noexcept
Set the number of groups used to split the channels in the normalization calculation.
Definition: NvInfer.h:6410
A OneHot layer in a network definition.
Definition: NvInfer.h:6003
virtual ~IOneHotLayer() noexcept=default
apiv::VOneHotLayer * mImpl
Definition: NvInfer.h:6024
void setAxis(int32_t axis) noexcept
Set the axis parameter.
Definition: NvInfer.h:6010
int32_t getAxis() const noexcept
Get the value of the axis parameter.
Definition: NvInfer.h:6018
Optimization profile for dynamic input dimensions and shape tensors.
Definition: NvInferRuntime.h:2627
Layer that represents a padding operation.
Definition: NvInfer.h:2997
Dims getPostPaddingNd() const noexcept
Get the padding that is applied at the end of the tensor.
Definition: NvInfer.h:3046
void setPrePaddingNd(Dims const &padding) noexcept
Set the padding that is applied at the start of the tensor.
Definition: NvInfer.h:3008
virtual ~IPaddingLayer() noexcept=default
void setPostPaddingNd(Dims const &padding) noexcept
Set the padding that is applied at the end of the tensor.
Definition: NvInfer.h:3034
Dims getPrePaddingNd() const noexcept
Get the padding that is applied at the start of the tensor.
Definition: NvInfer.h:3020
apiv::VPaddingLayer * mImpl
Definition: NvInfer.h:3052
Layer that represents a parametric ReLU operation.
Definition: NvInfer.h:3960
apiv::VParametricReLULayer * mImpl
Definition: NvInfer.h:3962
virtual ~IParametricReLULayer() noexcept=default
Single registration point for all plugins in an application. It is used to find plugin implementation...
Definition: NvInferRuntimeCommon.h:56
Plugin class for user-implemented layers.
Definition: NvInferRuntimePlugin.h:139
Layer type for pluginV2.
Definition: NvInfer.h:2734
virtual ~IPluginV2Layer() noexcept=default
apiv::VPluginV2Layer * mImpl
Definition: NvInfer.h:2747
IPluginV2 & getPlugin() noexcept
Get the plugin for the layer.
Definition: NvInfer.h:2741
Layer type for V3 plugins.
Definition: NvInfer.h:2761
virtual ~IPluginV3Layer() noexcept=default
IPluginV3 & getPlugin() noexcept
Get the plugin for the layer.
Definition: NvInfer.h:2768
apiv::VPluginV3Layer * mImpl
Definition: NvInfer.h:2774
A Pooling layer in a network definition.
Definition: NvInfer.h:1496
PoolingType getPoolingType() const noexcept
Get the type of activation to be performed.
Definition: NvInfer.h:1515
PaddingMode getPaddingMode() const noexcept
Get the padding mode.
Definition: NvInfer.h:1648
Dims getPostPadding() const noexcept
Get the padding.
Definition: NvInfer.h:1624
bool getAverageCountExcludesPadding() const noexcept
Get whether average pooling uses as a denominator the overlap area between the window and the unpadde...
Definition: NvInfer.h:1568
Dims getPrePadding() const noexcept
Get the pre-padding.
Definition: NvInfer.h:1596
void setPoolingType(PoolingType type) noexcept
Set the type of activation to be performed.
Definition: NvInfer.h:1505
void setWindowSizeNd(Dims const &windowSize) noexcept
Set the multi-dimension window size for pooling.
Definition: NvInfer.h:1661
void setPaddingMode(PaddingMode paddingMode) noexcept
Set the padding mode.
Definition: NvInfer.h:1637
Dims getWindowSizeNd() const noexcept
Get the multi-dimension window size for pooling.
Definition: NvInfer.h:1671
void setAverageCountExcludesPadding(bool exclusive) noexcept
Set whether average pooling uses as a denominator the overlap area between the window and the unpadde...
Definition: NvInfer.h:1557
void setPaddingNd(Dims const &padding) noexcept
Set the multi-dimension padding for pooling.
Definition: NvInfer.h:1715
float getBlendFactor() const noexcept
Get the blending factor for the max_average_blend mode: max_average_blendPool = (1-blendFactor)*maxPo...
Definition: NvInfer.h:1543
void setStrideNd(Dims const &stride) noexcept
Set the multi-dimension stride for pooling.
Definition: NvInfer.h:1686
Dims getStrideNd() const noexcept
Get the multi-dimension stride for pooling.
Definition: NvInfer.h:1696
virtual ~IPoolingLayer() noexcept=default
Dims getPaddingNd() const noexcept
Get the multi-dimension padding for pooling.
Definition: NvInfer.h:1727
void setPostPadding(Dims const &padding) noexcept
Set the multi-dimension post-padding for pooling.
Definition: NvInfer.h:1614
void setPrePadding(Dims const &padding) noexcept
Set the multi-dimension pre-padding for pooling.
Definition: NvInfer.h:1586
void setBlendFactor(float blendFactor) noexcept
Set the blending factor for the max_average_blend mode: max_average_blendPool = (1-blendFactor)*maxPo...
Definition: NvInfer.h:1530
A Quantize layer in a network definition.
Definition: NvInfer.h:5453
void setToType(DataType toType) noexcept
Set the Quantize layer output type.
Definition: NvInfer.h:5490
void setAxis(int32_t axis) noexcept
Set the quantization axis.
Definition: NvInfer.h:5474
int32_t getAxis() const noexcept
Get the quantization axis.
Definition: NvInfer.h:5463
virtual ~IQuantizeLayer() noexcept=default
DataType getToType() const noexcept
Return the Quantize layer output type.
Definition: NvInfer.h:5502
A RaggedSoftmax layer in a network definition.
Definition: NvInfer.h:3794
apiv::VRaggedSoftMaxLayer * mImpl
Definition: NvInfer.h:3796
virtual ~IRaggedSoftMaxLayer() noexcept=default
A recurrence layer in a network definition.
Definition: NvInfer.h:4666
virtual ~IRecurrenceLayer() noexcept=default
Layer that represents a reduction across a non-bool tensor.
Definition: NvInfer.h:2917
void setKeepDimensions(bool keepDimensions) noexcept
Set the boolean that specifies whether or not to keep the reduced dimensions for the layer.
Definition: NvInfer.h:2964
void setOperation(ReduceOperation op) noexcept
Set the reduce operation for the layer.
Definition: NvInfer.h:2924
ReduceOperation getOperation() const noexcept
Get the reduce operation for the layer.
Definition: NvInfer.h:2934
virtual ~IReduceLayer() noexcept=default
uint32_t getReduceAxes() const noexcept
Get the axes over which to reduce for the layer.
Definition: NvInfer.h:2954
void setReduceAxes(uint32_t reduceAxes) noexcept
Set the axes over which to reduce.
Definition: NvInfer.h:2944
apiv::VReduceLayer * mImpl
Definition: NvInfer.h:2980
bool getKeepDimensions() const noexcept
Get the boolean that specifies whether or not to keep the reduced dimensions for the layer.
Definition: NvInfer.h:2974
A resize layer in a network definition.
Definition: NvInfer.h:4149
void setSelectorForSinglePixel(ResizeSelector selector) noexcept
Set coordinate selector function when resized to single pixel.
Definition: NvInfer.h:4310
void setNearestRounding(ResizeRoundMode value) noexcept
Set rounding mode for nearest neighbor resize.
Definition: NvInfer.h:4334
virtual ~IResizeLayer() noexcept=default
int32_t getScales(int32_t size, float *scales) const noexcept
Copies resize scales to scales[0, ..., nbScales-1], where nbScales is the number of scales that were ...
Definition: NvInfer.h:4228
void setOutputDimensions(Dims const &dimensions) noexcept
Set the output dimensions.
Definition: NvInfer.h:4169
void setCubicCoeff(float A) noexcept
Set the coefficient 'A' used in cubic interpolation.
Definition: NvInfer.h:4366
void setScales(float const *scales, int32_t nbScales) noexcept
Set the resize scales.
Definition: NvInfer.h:4209
float getCubicCoeff() const noexcept
Get the coefficient 'A' used in cubic interpolation.
Definition: NvInfer.h:4376
ResizeSelector getSelectorForSinglePixel() const noexcept
Get the coordinate selector function when resized to single pixel.
Definition: NvInfer.h:4320
InterpolationMode getResizeMode() const noexcept
Get resize mode for an input tensor.
Definition: NvInfer.h:4250
void setCoordinateTransformation(ResizeCoordinateTransformation coordTransform) noexcept
Set coordinate transformation function.
Definition: NvInfer.h:4285
void setExcludeOutside(bool excludeFlag) noexcept
Set the state for excluding outside pixels.
Definition: NvInfer.h:4389
void setResizeMode(InterpolationMode interpolationMode) noexcept
Set resize mode for an input tensor.
Definition: NvInfer.h:4240
Dims getOutputDimensions() const noexcept
Get the output dimensions.
Definition: NvInfer.h:4179
ResizeRoundMode getNearestRounding() const noexcept
Get rounding mode for nearest neighbor resize.
Definition: NvInfer.h:4344
bool getExcludeOutside() const noexcept
Get the state for excluding outside pixels.
Definition: NvInfer.h:4399
ResizeCoordinateTransformation getCoordinateTransformation() const noexcept
Get coordinate transformation function.
Definition: NvInfer.h:4295
A ReverseSequence layer in a network definition.
Definition: NvInfer.h:6278
void setSequenceAxis(int32_t sequenceAxis) noexcept
Set the sequence axis. Default is 0.
Definition: NvInfer.h:6311
int32_t getBatchAxis() const noexcept
Return the batch axis. Return 1 if no batch axis was set.
Definition: NvInfer.h:6298
apiv::VReverseSequenceLayer * mImpl
Definition: NvInfer.h:6327
int32_t getSequenceAxis() const noexcept
Return the sequence axis. Return 0 if no sequence axis was set.
Definition: NvInfer.h:6321
void setBatchAxis(int32_t batchAxis) noexcept
Set the batch axis. Default is 1.
Definition: NvInfer.h:6288
virtual ~IReverseSequenceLayer() noexcept=default
A Scale layer in a network definition.
Definition: NvInfer.h:1893
Weights getScale() const noexcept
Get the scale value.
Definition: NvInfer.h:1950
Weights getPower() const noexcept
Get the power value.
Definition: NvInfer.h:1970
void setScale(Weights scale) noexcept
Set the scale value.
Definition: NvInfer.h:1940
void setPower(Weights power) noexcept
Set the power value.
Definition: NvInfer.h:1960
ScaleMode getMode() const noexcept
Get the scale mode.
Definition: NvInfer.h:1910
void setShift(Weights shift) noexcept
Set the shift value.
Definition: NvInfer.h:1920
void setChannelAxis(int32_t channelAxis) noexcept
Set the channel axis.
Definition: NvInfer.h:2006
Weights getShift() const noexcept
Get the shift value.
Definition: NvInfer.h:1930
virtual ~IScaleLayer() noexcept=default
void setMode(ScaleMode mode) noexcept
Set the scale mode.
Definition: NvInfer.h:1900
int32_t getChannelAxis() const noexcept
Get the channel axis.
Definition: NvInfer.h:1985
A scatter layer in a network definition. Supports several kinds of scattering.
Definition: NvInfer.h:5931
void setMode(ScatterMode mode) noexcept
Set the scatter mode.
Definition: NvInfer.h:5938
apiv::VScatterLayer * mImpl
Definition: NvInfer.h:5972
void setAxis(int32_t axis) noexcept
Set the axis used by ScatterMode::kELEMENTS.
Definition: NvInfer.h:5958
int32_t getAxis() const noexcept
Get the axis.
Definition: NvInfer.h:5966
ScatterMode getMode() const noexcept
Get the scatter mode.
Definition: NvInfer.h:5948
virtual ~IScatterLayer() noexcept=default
Select elements from two data tensors based on a condition tensor.
Definition: NvInfer.h:4974
virtual ~ISelectLayer() noexcept=default
Layer type for getting shape of a tensor.
Definition: NvInfer.h:3522
virtual ~IShapeLayer() noexcept=default
apiv::VShapeLayer * mImpl
Definition: NvInfer.h:3524
Layer type for shuffling data.
Definition: NvInfer.h:3085
apiv::VShuffleLayer * mImpl
Definition: NvInfer.h:3243
void setFirstTranspose(Permutation permutation) noexcept
Set the permutation applied by the first transpose operation.
Definition: NvInfer.h:3096
void setSecondTranspose(Permutation permutation) noexcept
Set the permutation applied by the second transpose operation.
Definition: NvInfer.h:3196
Dims getReshapeDimensions() const noexcept
Get the reshaped dimensions.
Definition: NvInfer.h:3149
void setReshapeDimensions(Dims const &dimensions) noexcept
Set the reshaped dimensions.
Definition: NvInfer.h:3136
Permutation getFirstTranspose() const noexcept
Get the permutation applied by the first transpose operation.
Definition: NvInfer.h:3108
virtual ~IShuffleLayer() noexcept=default
Permutation getSecondTranspose() const noexcept
Get the permutation applied by the second transpose operation.
Definition: NvInfer.h:3208
bool getZeroIsPlaceholder() const noexcept
Get meaning of 0 in reshape dimensions.
Definition: NvInfer.h:3237
void setZeroIsPlaceholder(bool zeroIsPlaceholder) noexcept
Set meaning of 0 in reshape dimensions.
Definition: NvInfer.h:3224
Slices an input tensor into an output tensor based on the offset and strides.
Definition: NvInfer.h:3337
void setStride(Dims const &stride) noexcept
Set the stride for computing the output slice data.
Definition: NvInfer.h:3406
apiv::VSliceLayer * mImpl
Definition: NvInfer.h:3505
virtual ~ISliceLayer() noexcept=default
void setSize(Dims const &size) noexcept
Set the dimensions of the output slice.
Definition: NvInfer.h:3377
void setAxes(Dims const &axes) noexcept
Set the axes for this ISliceLayer.
Definition: NvInfer.h:3484
void setStart(Dims const &start) noexcept
Set the start offset that the slice layer uses to create the output slice.
Definition: NvInfer.h:3348
Dims getStart() const noexcept
Get the start offset for the slice layer.
Definition: NvInfer.h:3363
void setMode(SampleMode mode) noexcept
Set the slice mode.
Definition: NvInfer.h:3431
Dims getSize() const noexcept
Get dimensions of the output slice.
Definition: NvInfer.h:3392
SampleMode getMode() const noexcept
Get the slice mode.
Definition: NvInfer.h:3441
Dims getStride() const noexcept
Get the stride for the output slice.
Definition: NvInfer.h:3421
Dims getAxes() const noexcept
Get the axes for this ISliceLayer.
Definition: NvInfer.h:3499
A Softmax layer in a network definition.
Definition: NvInfer.h:2037
void setAxes(uint32_t axes) noexcept
Set the axis along which softmax is computed. Currently, only one axis can be set.
Definition: NvInfer.h:2059
uint32_t getAxes() const noexcept
Get the axis along which softmax occurs.
Definition: NvInfer.h:2069
virtual ~ISoftMaxLayer() noexcept=default
Layer that represents a squeeze operation, removing unit dimensions of the input tensor on a set of a...
Definition: NvInfer.h:6476
virtual ~ISqueezeLayer() noexcept=default
apiv::VSqueezeLayer * mImpl
Definition: NvInfer.h:6493
A tensor in a network definition.
Definition: NvInfer.h:183
void setAllowedFormats(TensorFormats formats) noexcept
Set allowed formats for an input or output tensor. By default all formats are allowed....
Definition: NvInfer.h:453
TensorLocation getLocation() const noexcept
Get the storage location of a tensor.
Definition: NvInfer.h:372
void setDimensions(Dims const &dimensions) noexcept
Set the dimensions of a tensor.
Definition: NvInfer.h:231
void resetDynamicRange() noexcept
Undo effect of setDynamicRange.
Definition: NvInfer.h:411
void setName(char const *name) noexcept
Set the tensor name.
Definition: NvInfer.h:200
bool isExecutionTensor() const noexcept
Whether the tensor is an execution tensor.
Definition: NvInfer.h:518
TRT_DEPRECATED bool dynamicRangeIsSet() const noexcept
Query whether dynamic range is set.
Definition: NvInfer.h:403
char const * getName() const noexcept
Get the tensor name.
Definition: NvInfer.h:212
bool isShapeTensor() const noexcept
Whether the tensor is a shape tensor.
Definition: NvInfer.h:497
float getDynamicRangeMax() const noexcept
Get maximum of dynamic range.
Definition: NvInfer.h:431
bool isNetworkInput() const noexcept
Whether the tensor is a network input.
Definition: NvInfer.h:321
TRT_DEPRECATED void setBroadcastAcrossBatch(bool broadcastAcrossBatch) noexcept
Set whether to enable broadcast of tensor across the implicit batch dimension.
Definition: NvInfer.h:346
TRT_DEPRECATED bool setDynamicRange(float min, float max) noexcept
Set dynamic range for the tensor.
Definition: NvInfer.h:313
TRT_DEPRECATED void setType(DataType type) noexcept
Set the data type of a tensor.
Definition: NvInfer.h:281
TRT_DEPRECATED bool getBroadcastAcrossBatch() const noexcept
Check if tensor is broadcast across the implicit batch dimension.
Definition: NvInfer.h:360
bool isNetworkOutput() const noexcept
Whether the tensor is a network output.
Definition: NvInfer.h:329
DataType getType() const noexcept
Get the data type of a tensor.
Definition: NvInfer.h:296
apiv::VTensor * mImpl
Definition: NvInfer.h:565
float getDynamicRangeMin() const noexcept
Get minimum of dynamic range.
Definition: NvInfer.h:421
virtual ~ITensor() noexcept=default
void setDimensionName(int32_t index, char const *name) noexcept
Name a dimension of an input tensor.
Definition: NvInfer.h:544
char const * getDimensionName(int32_t index) const noexcept
Get the name of an input dimension.
Definition: NvInfer.h:559
TRT_DEPRECATED void setLocation(TensorLocation location) noexcept
Set the storage location of a tensor.
Definition: NvInfer.h:391
Dims getDimensions() const noexcept
Get the dimensions of a tensor.
Definition: NvInfer.h:245
TensorFormats getAllowedFormats() const noexcept
Get a bitmask of TensorFormat values that the tensor supports. For a shape tensor,...
Definition: NvInfer.h:466
Class to handle tactic timing info collected from builder.
Definition: NvInfer.h:9083
int64_t queryKeys(TimingCacheKey *keyBuffer, int64_t capacity) const noexcept
Query cache keys from Timing Cache.
Definition: NvInfer.h:9149
bool combine(ITimingCache const &inputCache, bool ignoreMismatch) noexcept
Combine input timing cache into local instance.
Definition: NvInfer.h:9120
TimingCacheValue query(TimingCacheKey const &key) const noexcept
Query value in a cache entry.
Definition: NvInfer.h:9166
virtual ~ITimingCache() noexcept=default
bool update(TimingCacheKey const &key, TimingCacheValue const &value) noexcept
Update values in a cache entry.
Definition: NvInfer.h:9188
apiv::VTimingCache * mImpl
Definition: NvInfer.h:9194
bool reset() noexcept
Empty the timing cache.
Definition: NvInfer.h:9130
Layer that represents a TopK reduction.
Definition: NvInfer.h:3562
void setK(int32_t k) noexcept
Set the static k value for the layer.
Definition: NvInfer.h:3593
void setReduceAxes(uint32_t reduceAxes) noexcept
Set which axes to reduce for the layer.
Definition: NvInfer.h:3617
TopKOperation getOperation() const noexcept
Get the operation for the layer.
Definition: NvInfer.h:3579
apiv::VTopKLayer * mImpl
Definition: NvInfer.h:3649
void setOperation(TopKOperation op) noexcept
Set the operation for the layer.
Definition: NvInfer.h:3569
int32_t getK() const noexcept
Get the k value for the layer.
Definition: NvInfer.h:3607
uint32_t getReduceAxes() const noexcept
Get the axes to reduce for the layer.
Definition: NvInfer.h:3627
virtual ~ITopKLayer() noexcept=default
A layer that represents a trip-count limiter.
Definition: NvInfer.h:4787
TripLimit getTripLimit() const noexcept
Get a trip limiter type.
Definition: NvInfer.h:4792
virtual ~ITripLimitLayer() noexcept=default
Layer that represents an unary operation.
Definition: NvInfer.h:2842
void setOperation(UnaryOperation op) noexcept
Set the unary operation for the layer.
Definition: NvInfer.h:2851
apiv::VUnaryLayer * mImpl
Definition: NvInfer.h:2867
UnaryOperation getOperation() const noexcept
Get the unary operation for the layer.
Definition: NvInfer.h:2861
virtual ~IUnaryLayer() noexcept=default
Layer that represents an unsqueeze operation, which reshapes the input tensor by inserting unit-lengt...
Definition: NvInfer.h:6505
virtual ~IUnsqueezeLayer() noexcept=default
apiv::VUnsqueezeLayer * mImpl
Definition: NvInfer.h:6522
An Interface class for version control.
Definition: NvInferRuntimeBase.h:276
Version information associated with a TRT interface.
Definition: NvInferRuntimeBase.h:241
An array of weights used as a layer parameter.
Definition: NvInferRuntime.h:124
Definition: NvInfer.h:8694
virtual int32_t selectAlgorithms(IAlgorithmContext const &context, IAlgorithm const *const *choices, int32_t nbChoices, int32_t *selection) noexcept=0
Select Algorithms for a layer from the given list of algorithm choices.
virtual void reportAlgorithms(IAlgorithmContext const *const *algoContexts, IAlgorithm const *const *algoChoices, int32_t nbAlgorithms) noexcept=0
Called by TensorRT to report choices it made.
InterfaceInfo getInterfaceInfo() const noexcept override
Return version information associated with this interface. Applications must not override this method...
Definition: NvInfer.h:8699
virtual ~IAlgorithmSelector() noexcept=default
Definition: NvInferRuntimeBase.h:413
Definition: NvInferRuntime.h:1610
Definition: NvInfer.h:8321
~IInt8EntropyCalibrator2() noexcept override=default
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:8334
InterfaceInfo getInterfaceInfo() const noexcept override
Return version information associated with this interface. Applications must not override this method...
Definition: NvInfer.h:8326
Definition: NvInfer.h:8281
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:8294
InterfaceInfo getInterfaceInfo() const noexcept override
Return version information associated with this interface. Applications must not override this method...
Definition: NvInfer.h:8286
~IInt8EntropyCalibrator() noexcept override=default
Definition: NvInfer.h:8400
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:8413
InterfaceInfo getInterfaceInfo() const noexcept override
Return version information associated with this interface. Applications must not override this method...
Definition: NvInfer.h:8405
virtual double getQuantile() const noexcept=0
The quantile (between 0 and 1) that will be used to select the region maximum when the quantile metho...
Definition: NvInfer.h:8361
~IInt8MinMaxCalibrator() noexcept override=default
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:8374
InterfaceInfo getInterfaceInfo() const noexcept override
Return version information associated with this interface. Applications must not override this method...
Definition: NvInfer.h:8366
Definition: NvInferPluginBase.h:206
Definition: NvInfer.h:9410
virtual bool stepComplete(char const *phaseName, int32_t step) noexcept=0
Signal that a step of an optimizer phase has finished.
virtual ~IProgressMonitor() noexcept=default
IProgressMonitor()=default
virtual void phaseFinish(char const *phaseName) noexcept=0
Signal that a phase of the optimizer has finished.
virtual void phaseStart(char const *phaseName, char const *parentPhase, int32_t nbSteps) noexcept=0
Signal that a phase of the optimizer has started.
IBuilder * createInferBuilder(ILogger &logger) noexcept
Create an instance of an IBuilder class.
Definition: NvInfer.h:10791
The TensorRT API version 1 namespace.
uint32_t TacticSources
Represents a collection of one or more TacticSource values combine using bitwise-OR operations.
Definition: NvInferRuntime.h:2913
ResizeSelector
The coordinate selector when resize to single pixel output.
Definition: NvInfer.h:4054
@ kFORMULA
Use formula to map the original index.
@ kUPPER
Select the upper left pixel.
EngineCapability
List of supported engine capability flows.
Definition: NvInferRuntime.h:76
nvinfer1::IPluginRegistry * getBuilderPluginRegistry(nvinfer1::EngineCapability capability) noexcept
Return the plugin registry for building a Standard engine, or nullptr if no registry exists.
MemoryPoolType
The type for memory pools used by TensorRT.
Definition: NvInfer.h:9205
ScaleMode
Controls how shift, scale and power are applied in a Scale layer.
Definition: NvInfer.h:1850
@ kUNIFORM
Identical coefficients across all elements of the tensor.
@ kCHANNEL
Per-channel coefficients.
RuntimePlatform
Describes the intended runtime platform (operating system and CPU architecture) for the execution of ...
Definition: NvInfer.h:8805
uint32_t QuantizationFlags
Represents one or more QuantizationFlag values using binary OR operations.
Definition: NvInfer.h:8757
HardwareCompatibilityLevel
Describes requirements of compatibility with GPU architectures other than that of the GPU on which th...
Definition: NvInfer.h:9324
@ kSAME_COMPUTE_CAPABILITY
CumulativeOperation
Enumerates the cumulative operations that may be performed by a Cumulative layer.
Definition: NvInfer.h:6538
BoundingBoxFormat
Representation of bounding box data used for the Boxes input tensor in INMSLayer.
Definition: NvInfer.h:6123
@ kCENTER_SIZES
(x_center, y_center, width, height) where (x_center, y_center) is the center point of the box
@ kCORNER_PAIRS
(x1, y1, x2, y2) where (x1, y1) and (x2, y2) are any pair of diagonal corners
constexpr int32_t EnumMax< BuilderFlag >() noexcept
Definition: NvInfer.h:9027
constexpr int32_t EnumMax< LayerType >() noexcept
Definition: NvInfer.h:118
constexpr int32_t EnumMax< CalibrationAlgoType >() noexcept
Definition: NvInfer.h:8196
UnaryOperation
Enumerates the unary operations that may be performed by a Unary layer.
Definition: NvInfer.h:2795
@ kISINF
Return true if input value equals +/- infinity for floating-point data type.
@ kCOSH
Hyperbolic cosine.
@ kACOSH
Inverse hyperbolic cosine.
@ kERF
Gauss error function.
@ kISNAN
Return true if input value is a NaN for floating-point data type.
@ kROUND
Round to nearest even for floating-point data type.
@ kATANH
Inverse hyperbolic tangent.
@ kASINH
Inverse hyperbolic sine.
@ kSIGN
Sign, If input > 0, output 1; if input < 0, output -1; if input == 0, output 0.
constexpr int32_t EnumMax< ReduceOperation >() noexcept
Definition: NvInfer.h:2904
constexpr int32_t EnumMax< TripLimit >() noexcept
Definition: NvInfer.h:4455
ActivationType
Enumerates the types of activation to perform in an activation layer.
Definition: NvInfer.h:137
@ kSELU
Selu activation: x>0 ? beta * x : beta * (alpha*exp(x) - alpha)
@ kSCALED_TANH
Scaled tanh activation: alpha*tanh(beta*x)
@ kRELU
Rectified linear activation.
@ kELU
Elu activation: x>=0 ? x : alpha * (exp(x) - 1).
@ kLEAKY_RELU
LeakyRelu activation: x>=0 ? x : alpha * x.
@ kSOFTSIGN
Softsign activation: x / (1+|x|)
@ kHARD_SIGMOID
Hard sigmoid activation: max(0, min(1, alpha*x+beta))
@ kTHRESHOLDED_RELU
Thresholded ReLU activation: x>alpha ? x : 0.
@ kSIGMOID
Sigmoid activation.
@ kCLIP
Clip activation: max(alpha, min(beta, x))
@ kGELU_TANH
GELU tanh activation: 0.5 * x * (1 + tanh(sqrt(2/pi) * (0.044715F * pow(x, 3) + x)))
@ kGELU_ERF
GELU erf activation: 0.5 * x * (1 + erf(sqrt(0.5) * x))
@ kSOFTPLUS
Parametric softplus activation: alpha*log(exp(beta*x)+1)
FillOperation
Enumerates the tensor fill operations that may performed by a fill layer.
Definition: NvInfer.h:5035
@ kRANDOM_UNIFORM
Randomly draw values from a uniform distribution.
@ kRANDOM_NORMAL
Randomly draw values from a normal distribution.
ResizeRoundMode
The rounding mode for nearest neighbor resize.
Definition: NvInfer.h:4084
@ kHALF_DOWN
Round half down.
nvinfer1::safe::IPluginRegistry * getBuilderSafePluginRegistry(nvinfer1::EngineCapability capability) noexcept
Return the plugin registry for building a Safety engine, or nullptr if no registry exists.
PaddingMode
Enumerates the modes of padding to perform in convolution, deconvolution and pooling layer,...
Definition: NvInfer.h:1028
@ kSAME_LOWER
Use SAME padding, with prePadding >= postPadding.
@ kEXPLICIT_ROUND_DOWN
Use explicit padding, rounding output size down.
@ kEXPLICIT_ROUND_UP
Use explicit padding, rounding output size up.
@ kSAME_UPPER
Use SAME padding, with prePadding <= postPadding.
TripLimit
Enum that describes kinds of trip limits.
Definition: NvInfer.h:4443
@ kWHILE
Tensor is a scalar of type kBOOL. Loop terminates when value is false.
@ kCOUNT
Tensor is a scalar of type kINT32 or kINT64 that contains the trip count.
uint32_t NetworkDefinitionCreationFlags
Represents one or more NetworkDefinitionCreationFlag flags using binary OR operations....
Definition: NvInfer.h:10435
PreviewFeature
Define preview features.
Definition: NvInfer.h:9280
@ kRUNTIME_ACTIVATION_RESIZE_10_10
@ kALIASED_PLUGIN_IO_10_03
TilingOptimizationLevel
Define the optimization levels for Tiling.
Definition: NvInfer.h:9377
@ kFAST
Use a fast algorithm and heuristic based strategy. Slightly increases engine build time.
@ kFULL
Increase search space even wider. Significantly increases engine build time.
constexpr int32_t EnumMax< GatherMode >() noexcept
Definition: NvInfer.h:2554
DataType
The type of weights and tensors.
Definition: NvInferRuntimeBase.h:143
uint32_t BuilderFlags
Represents one or more BuilderFlag values using binary OR operations, e.g., 1U << BuilderFlag::kFP16 ...
Definition: NvInfer.h:8835
DeviceType
The device that this layer/network will execute on.
Definition: NvInferRuntime.h:1304
constexpr int32_t EnumMax< ScaleMode >() noexcept
Definition: NvInfer.h:1862
CalibrationAlgoType
Version of calibration algorithm to use.
Definition: NvInfer.h:8183
@ kENTROPY_CALIBRATION_2
Entropy calibration.
@ kLEGACY_CALIBRATION
Legacy calibration.
@ kENTROPY_CALIBRATION
Legacy entropy calibration.
@ kMINMAX_CALIBRATION
Minmax calibration.
LayerType
The type values of layer classes.
Definition: NvInfer.h:58
@ kGRID_SAMPLE
Grid sample layer.
@ kRAGGED_SOFTMAX
Ragged softmax layer.
@ kDECONVOLUTION
Deconvolution layer.
@ kASSERTION
Assertion layer.
@ kMATRIX_MULTIPLY
Matrix multiply layer.
@ kCONDITION
Condition layer.
@ kCUMULATIVE
Cumulative layer.
@ kCONDITIONAL_INPUT
Conditional Input layer.
@ kIDENTITY
Identity layer.
@ kNORMALIZATION
Normalization layer.
@ kQUANTIZE
Quantize layer.
@ kCONVOLUTION
Convolution layer.
@ kPARAMETRIC_RELU
Parametric ReLU layer.
@ kUNSQUEEZE
Unsqueeze Layer.
@ kCONCATENATION
Concatenation layer.
@ kREVERSE_SEQUENCE
Reverse sequence layer.
@ kRECURRENCE
Loop Recurrence layer.
@ kDEQUANTIZE
Dequantize layer.
@ kPLUGIN_V3
PluginV3 layer.
@ kITERATOR
Loop Iterator layer.
@ kTRIP_LIMIT
Loop Trip limit layer.
@ kDYNAMIC_QUANTIZE
Dynamic Quantize layer.
@ kUNARY
UnaryOp operation Layer.
@ kACTIVATION
Activation layer.
@ kELEMENTWISE
Elementwise layer.
@ kPLUGIN_V2
PluginV2 layer.
@ kLOOP_OUTPUT
Loop output layer.
@ kCONDITIONAL_OUTPUT
Conditional Output layer.
@ kCONSTANT
Constant layer.
@ kNON_ZERO
NonZero layer.
constexpr int32_t EnumMax< QuantizationFlag >() noexcept
Definition: NvInfer.h:8782
SampleMode
Controls how ISliceLayer and IGridSample handle out-of-bounds coordinates.
Definition: NvInfer.h:3253
@ kCLAMP
Out of bounds indices are clamped to bounds.
@ kSTRICT_BOUNDS
Fail with error when the coordinates are out of bounds.
@ kWRAP
Coordinates wrap around periodically.
GatherMode
Control form of IGatherLayer.
Definition: NvInfer.h:2542
@ kDEFAULT
Similar to ONNX Gather.
@ kELEMENT
Similar to ONNX GatherElements.
@ kND
Similar to ONNX GatherND.
uint32_t TensorFormats
It is capable of representing one or more TensorFormat by binary OR operations, e....
Definition: NvInfer.h:129
ProfilingVerbosity
List of verbosity levels of layer information exposed in NVTX annotations and in IEngineInspector.
Definition: NvInferRuntime.h:2925
NetworkDefinitionCreationFlag
List of immutable network properties expressed at network creation time. NetworkDefinitionCreationFla...
Definition: NvInfer.h:10446
@ kPREFER_JIT_PYTHON_PLUGINS
@ kPREFER_AOT_PYTHON_PLUGINS
ElementWiseOperation
Enumerates the binary operations that may be performed by an ElementWise layer.
Definition: NvInfer.h:2452
@ kSUB
Subtract the second element from the first.
@ kSUM
Sum of the two elements.
@ kPROD
Product of the two elements.
@ kFLOOR_DIV
Floor division of the first element by the second.
@ kEQUAL
Check if two elements are equal.
@ kAND
Logical AND of two elements.
@ kOR
Logical OR of two elements.
@ kMIN
Minimum of the two elements.
@ kPOW
The first element to the power of the second element.
@ kLESS
Check if element in first tensor is less than corresponding element in second tensor.
@ kGREATER
Check if element in first tensor is greater than corresponding element in second tensor.
@ kXOR
Logical XOR of two elements.
@ kDIV
Divide the first element by the second.
QuantizationFlag
List of valid flags for quantizing the network to int8.
Definition: NvInfer.h:8769
@ kCALIBRATE_BEFORE_FUSION
constexpr int32_t EnumMax< SampleMode >() noexcept
Definition: NvInfer.h:3269
InterpolationMode
Enumerates various modes of interpolation.
Definition: NvInfer.h:3972
@ kNEAREST
ND (0 < N <= 8) nearest neighbor resizing.
@ kCUBIC
Supports bicubic (2D) interpolation.
@ kLINEAR
Supports linear (1D), bilinear (2D), and trilinear (3D) interpolation.
BuilderFlag
List of valid modes that the builder can enable when creating an engine from a network definition.
Definition: NvInfer.h:8845
@ kWEIGHT_STREAMING
Enable weight streaming for the current engine.
@ kDEBUG
Enable debugging of layers via synchronizing after every layer.
@ kGPU_FALLBACK
Enable layers marked to execute on GPU if layer cannot execute on DLA.
@ kSPARSE_WEIGHTS
Allow the builder to examine weights and use optimized functions when weights have suitable sparsity.
@ kERROR_ON_TIMING_CACHE_MISS
@ kEDITABLE_TIMING_CACHE
Enable editable timing cache.
@ kPREFER_PRECISION_CONSTRAINTS
@ kDISTRIBUTIVE_INDEPENDENCE
@ kSTRIP_PLAN
Strip the refittable weights from the engine plan file.
@ kMONITOR_MEMORY
Enable memory monitor during build time.
@ kDISABLE_TIMING_CACHE
Disable reuse of timing information across identical layers.
@ kDISABLE_COMPILATION_CACHE
@ kREFIT
Enable building a refittable engine.
@ kOBEY_PRECISION_CONSTRAINTS
@ kREJECT_EMPTY_ALGORITHMS
constexpr int32_t EnumMax< TopKOperation >() noexcept
Definition: NvInfer.h:3545
constexpr int32_t EnumMax< MemoryPoolType >() noexcept
Definition: NvInfer.h:9266
TopKOperation
Enumerates the operations that may be performed by a TopK layer.
Definition: NvInfer.h:3534
ReduceOperation
Enumerates the reduce operations that may be performed by a Reduce layer.
Definition: NvInfer.h:2890
constexpr int32_t EnumMax< LoopOutput >() noexcept
Definition: NvInfer.h:4432
constexpr int32_t EnumMax< NetworkDefinitionCreationFlag >() noexcept
Definition: NvInfer.h:10474
ScatterMode
Control form of IScatterLayer.
Definition: NvInfer.h:5857
MatrixOperation
Enumerates the operations that may be performed on a tensor by IMatrixMultiplyLayer before multiplica...
Definition: NvInfer.h:3660
@ kTRANSPOSE
Like kNONE, but transpose the matrix dimensions.
ResizeCoordinateTransformation
The resize coordinate transformation function.
Definition: NvInfer.h:4000
constexpr int32_t EnumMax< UnaryOperation >() noexcept
Definition: NvInfer.h:2829
LoopOutput
Enum that describes kinds of loop outputs.
Definition: NvInfer.h:4415
@ kLAST_VALUE
Output value is value of tensor for last iteration.
@ kCONCATENATE
Output value is concatenation of values of tensor for each iteration, in forward order.
@ kREVERSE
Output value is concatenation of values of tensor for each iteration, in reverse order.
constexpr int32_t EnumMax< BoundingBoxFormat >() noexcept
Definition: NvInfer.h:6136
constexpr int32_t EnumMax< MatrixOperation >() noexcept
Definition: NvInfer.h:3688
PoolingType
The type of pooling to perform in a pooling layer.
Definition: NvInfer.h:1464
@ kAVERAGE
Average over elements. If the tensor is padded, the count includes the padding.
@ kMAX
Maximum over elements.
@ kMAX_AVERAGE_BLEND
Blending between max and average pooling: (1-blendFactor)*maxPool + blendFactor*avgPool.
v_1_0::IProgressMonitor IProgressMonitor
Definition: NvInfer.h:9493
constexpr int32_t EnumMax< FillOperation >() noexcept
Definition: NvInfer.h:5066
TensorLocation
The location for tensor data storage, device or host.
Definition: NvInferRuntime.h:204
OptProfileSelector
When setting or querying optimization profile parameters (such as shape tensor inputs or dynamic dime...
Definition: NvInferRuntime.h:2587
constexpr int32_t EnumMax< ScatterMode >() noexcept
Definition: NvInfer.h:5868
Represents a permutation of dimensions.
Definition: NvInfer.h:3062
Declaration of EnumMaxImpl struct to store maximum number of elements in an enumeration type.
Definition: NvInferRuntimeBase.h:128
The key to retrieve timing cache entries.
Definition: NvInfer.h:9047
Definition: NvInfer.h:9059
uint64_t tacticHash
Hash of the selected tactic.
Definition: NvInfer.h:9061
float timingMSec
Timing of this tactic in milliseconds. Negative numbers and NaN are invalid values.
Definition: NvInfer.h:9063