176 static constexpr int32_t kVALUE = 12;
212 mImpl->setName(name);
224 return mImpl->getName();
243 mImpl->setDimensions(dimensions);
256 return mImpl->getDimensions();
271 mImpl->setType(type);
283 return mImpl->getType();
298 return mImpl->setDynamicRange(min, max);
306 return mImpl->isNetworkInput();
314 return mImpl->isNetworkOutput();
336 mImpl->setBroadcastAcrossBatch(broadcastAcrossBatch);
352 return mImpl->getBroadcastAcrossBatch();
362 return mImpl->getLocation();
377 mImpl->setLocation(location);
387 return mImpl->dynamicRangeIsSet();
395 mImpl->resetDynamicRange();
405 return mImpl->getDynamicRangeMin();
415 return mImpl->getDynamicRangeMax();
434 mImpl->setAllowedFormats(formats);
447 return mImpl->getAllowedFormats();
481 return mImpl->isShapeTensor();
504 return mImpl->isExecutionTensor();
508 apiv::VTensor* mImpl;
509 virtual ~ITensor() noexcept = default;
529 return mLayer->getType();
541 mLayer->setName(name);
552 return mLayer->getName();
560 return mLayer->getNbInputs();
573 return mLayer->getInput(index);
581 return mLayer->getNbOutputs();
592 return mLayer->getOutput(index);
609 return mLayer->setInput(index, tensor);
635 mLayer->setPrecision(dataType);
647 return mLayer->getPrecision();
659 return mLayer->precisionIsSet();
669 mLayer->resetPrecision();
707 mLayer->setOutputType(index, dataType);
721 return mLayer->getOutputType(index);
734 return mLayer->outputTypeIsSet(index);
746 return mLayer->resetOutputType(index);
750 virtual ~ILayer() noexcept = default;
751 apiv::VLayer* mLayer;
992 static constexpr int32_t kVALUE = 6;
1022 mImpl->setKernelSize(kernelSize);
1034 return mImpl->getKernelSize();
1046 mImpl->setNbOutputMaps(nbOutputMaps);
1056 return mImpl->getNbOutputMaps();
1072 mImpl->setStride(stride);
1082 return mImpl->getStride();
1102 return mImpl->setPadding(padding);
1114 return mImpl->getPadding();
1134 mImpl->setNbGroups(nbGroups);
1144 return mImpl->getNbGroups();
1158 mImpl->setKernelWeights(weights);
1168 return mImpl->getKernelWeights();
1183 mImpl->setBiasWeights(weights);
1193 return mImpl->getBiasWeights();
1209 return mImpl->setDilation(dilation);
1221 return mImpl->getDilation();
1238 mImpl->setPrePadding(padding);
1248 return mImpl->getPrePadding();
1265 mImpl->setPostPadding(padding);
1275 return mImpl->getPostPadding();
1289 mImpl->setPaddingMode(paddingMode);
1301 return mImpl->getPaddingMode();
1314 mImpl->setKernelSizeNd(kernelSize);
1324 return mImpl->getKernelSizeNd();
1339 mImpl->setStrideNd(stride);
1349 return mImpl->getStrideNd();
1367 mImpl->setPaddingNd(padding);
1379 return mImpl->getPaddingNd();
1393 mImpl->setDilationNd(dilation);
1403 return mImpl->getDilationNd();
1431 apiv::VConvolutionLayer* mImpl;
1475 mImpl->setNbOutputChannels(nbOutputs);
1485 return mImpl->getNbOutputChannels();
1495 mImpl->setKernelWeights(weights);
1505 return mImpl->getKernelWeights();
1517 mImpl->setBiasWeights(weights);
1527 return mImpl->getBiasWeights();
1555 apiv::VFullyConnectedLayer* mImpl;
1581 mImpl->setActivationType(type);
1591 return mImpl->getActivationType();
1606 mImpl->setAlpha(alpha);
1620 mImpl->setBeta(beta);
1629 return mImpl->getAlpha();
1638 return mImpl->getBeta();
1643 apiv::VActivationLayer* mImpl;
1655 kMAX_AVERAGE_BLEND = 2
1664 static constexpr int32_t kVALUE = 3;
1691 mImpl->setPoolingType(type);
1701 return mImpl->getPoolingType();
1715 mImpl->setWindowSize(windowSize);
1727 return mImpl->getWindowSize();
1743 mImpl->setStride(stride);
1755 return mImpl->getStride();
1771 mImpl->setPadding(padding);
1785 return mImpl->getPadding();
1800 mImpl->setBlendFactor(blendFactor);
1813 return mImpl->getBlendFactor();
1830 mImpl->setAverageCountExcludesPadding(exclusive);
1841 return mImpl->getAverageCountExcludesPadding();
1859 mImpl->setPrePadding(padding);
1869 return mImpl->getPrePadding();
1887 mImpl->setPostPadding(padding);
1897 return mImpl->getPostPadding();
1910 mImpl->setPaddingMode(paddingMode);
1921 return mImpl->getPaddingMode();
1934 mImpl->setWindowSizeNd(windowSize);
1944 return mImpl->getWindowSizeNd();
1959 mImpl->setStrideNd(stride);
1969 return mImpl->getStrideNd();
1988 mImpl->setPaddingNd(padding);
2000 return mImpl->getPaddingNd();
2005 apiv::VPoolingLayer* mImpl;
2031 mImpl->setWindowSize(windowSize);
2041 return mImpl->getWindowSize();
2052 mImpl->setAlpha(alpha);
2062 return mImpl->getAlpha();
2073 mImpl->setBeta(beta);
2083 return mImpl->getBeta();
2104 return mImpl->getK();
2109 apiv::VLRNLayer* mImpl;
2167 mImpl->setMode(mode);
2177 return mImpl->getMode();
2187 mImpl->setShift(shift);
2197 return mImpl->getShift();
2207 mImpl->setScale(scale);
2217 return mImpl->getScale();
2227 mImpl->setPower(power);
2237 return mImpl->getPower();
2252 return mImpl->getChannelAxis();
2273 mImpl->setChannelAxis(channelAxis);
2278 apiv::VScaleLayer* mImpl;
2327 mImpl->setAxes(axes);
2337 return mImpl->getAxes();
2342 apiv::VSoftMaxLayer* mImpl;
2372 mImpl->setAxis(axis);
2382 return mImpl->getAxis();
2387 apiv::VConcatenationLayer* mImpl;
2413 mImpl->setKernelSize(kernelSize);
2425 return mImpl->getKernelSize();
2437 mImpl->setNbOutputMaps(nbOutputMaps);
2447 return mImpl->getNbOutputMaps();
2464 mImpl->setStride(stride);
2476 return mImpl->getStride();
2496 mImpl->setPadding(padding);
2510 return mImpl->getPadding();
2530 mImpl->setNbGroups(nbGroups);
2540 return mImpl->getNbGroups();
2554 mImpl->setKernelWeights(weights);
2564 return mImpl->getKernelWeights();
2579 mImpl->setBiasWeights(weights);
2589 return mImpl->getBiasWeights();
2607 mImpl->setPrePadding(padding);
2617 return mImpl->getPrePadding();
2635 mImpl->setPostPadding(padding);
2645 return mImpl->getPostPadding();
2659 mImpl->setPaddingMode(paddingMode);
2671 return mImpl->getPaddingMode();
2686 mImpl->setKernelSizeNd(kernelSize);
2696 return mImpl->getKernelSizeNd();
2714 mImpl->setStrideNd(stride);
2724 return mImpl->getStrideNd();
2742 mImpl->setPaddingNd(padding);
2754 return mImpl->getPaddingNd();
2788 mImpl->setDilationNd(dilation);
2798 return mImpl->getDilationNd();
2803 apiv::VDeconvolutionLayer* mImpl;
2837 static constexpr int32_t kVALUE = 14;
2874 return mImpl->setOperation(op);
2886 return mImpl->getOperation();
2890 apiv::VElementWiseLayer* mImpl;
3005 mImpl->setGatherAxis(axis);
3016 return mImpl->getGatherAxis();
3037 mImpl->setNbElementWiseDims(elementWiseDims);
3047 return mImpl->getNbElementWiseDims();
3057 mImpl->setMode(mode);
3067 return mImpl->getMode();
3071 apiv::VGatherLayer* mImpl;
3257 return mImpl->getLayerCount();
3261 return mImpl->getHiddenSize();
3265 return mImpl->getMaxSeqLength();
3269 return mImpl->getDataLength();
3288 return mImpl->setSequenceLengths(seqLengths);
3300 return mImpl->getSequenceLengths();
3309 mImpl->setOperation(op);
3318 return mImpl->getOperation();
3327 mImpl->setInputMode(op);
3336 return mImpl->getInputMode();
3351 mImpl->setDirection(op);
3360 return mImpl->getDirection();
3419 mImpl->setWeightsForGate(layerIndex, gate, isW, weights);
3428 return mImpl->getWeightsForGate(layerIndex, gate, isW);
3453 mImpl->setBiasForGate(layerIndex, gate, isW, bias);
3462 return mImpl->getBiasForGate(layerIndex, gate, isW);
3479 mImpl->setHiddenState(hidden);
3488 return mImpl->getHiddenState();
3507 mImpl->setCellState(cell);
3516 return mImpl->getCellState();
3520 apiv::VRNNv2Layer* mImpl;
3543 return mImpl->getPlugin();
3547 apiv::VPluginV2Layer* mImpl;
3609 mImpl->setOperation(op);
3619 return mImpl->getOperation();
3623 apiv::VUnaryLayer* mImpl;
3678 mImpl->setOperation(op);
3688 return mImpl->getOperation();
3698 mImpl->setReduceAxes(reduceAxes);
3708 return mImpl->getReduceAxes();
3718 mImpl->setKeepDimensions(keepDimensions);
3728 return mImpl->getKeepDimensions();
3732 apiv::VReduceLayer* mImpl;
3760 mImpl->setPrePadding(padding);
3772 return mImpl->getPrePadding();
3786 mImpl->setPostPadding(padding);
3798 return mImpl->getPostPadding();
3812 mImpl->setPrePaddingNd(padding);
3824 return mImpl->getPrePaddingNd();
3838 mImpl->setPostPaddingNd(padding);
3850 return mImpl->getPostPaddingNd();
3854 apiv::VPaddingLayer* mImpl;
3895 mImpl->setFirstTranspose(permutation);
3907 return mImpl->getFirstTranspose();
3932 mImpl->setReshapeDimensions(dimensions);
3945 return mImpl->getReshapeDimensions();
3992 mImpl->setSecondTranspose(permutation);
4004 return mImpl->getSecondTranspose();
4020 return mImpl->setZeroIsPlaceholder(zeroIsPlaceholder);
4033 return mImpl->getZeroIsPlaceholder();
4037 apiv::VShuffleLayer* mImpl;
4111 mImpl->setStart(start);
4126 return mImpl->getStart();
4140 return mImpl->setSize(size);
4155 return mImpl->getSize();
4169 mImpl->setStride(stride);
4184 return mImpl->getStride();
4194 mImpl->setMode(mode);
4204 return mImpl->getMode();
4231 apiv::VSliceLayer* mImpl;
4250 apiv::VShapeLayer* mImpl;
4289 mImpl->setOperation(op);
4299 return mImpl->getOperation();
4321 return mImpl->getK();
4331 mImpl->setReduceAxes(reduceAxes);
4341 return mImpl->getReduceAxes();
4345 apiv::VTopKLayer* mImpl;
4420 mImpl->setOperation(index, op);
4430 return mImpl->getOperation(index);
4434 apiv::VMatrixMultiplyLayer* mImpl;
4455 apiv::VRaggedSoftMaxLayer* mImpl;
4474 apiv::VIdentityLayer* mImpl;
4499 mImpl->setWeights(weights);
4509 return mImpl->getWeights();
4521 mImpl->setDimensions(dimensions);
4533 return mImpl->getDimensions();
4537 apiv::VConstantLayer* mImpl;
4551 apiv::VParametricReLULayer* mImpl;
4572 static constexpr int32_t kVALUE = 2;
4622 static constexpr int32_t kVALUE = 3;
4648 static constexpr int32_t kVALUE = 2;
4680 static constexpr int32_t kVALUE = 4;
4728 return mImpl->setOutputDimensions(dimensions);
4738 return mImpl->getOutputDimensions();
4766 void setScales(
const float* scales, int32_t nbScales)
noexcept
4768 mImpl->setScales(scales, nbScales);
4785 int32_t
getScales(int32_t size,
float* scales)
const noexcept
4787 return mImpl->getScales(size, scales);
4801 mImpl->setResizeMode(resizeMode);
4811 return mImpl->getResizeMode();
4828 mImpl->setAlignCorners(alignCorners);
4841 return mImpl->getAlignCorners();
4878 mImpl->setCoordinateTransformation(coordTransform);
4888 return mImpl->getCoordinateTransformation();
4905 mImpl->setSelectorForSinglePixel(selector);
4915 return mImpl->getSelectorForSinglePixel();
4931 mImpl->setNearestRounding(value);
4941 return mImpl->getNearestRounding();
4946 apiv::VResizeLayer* mImpl;
4992 return mBoundary->getLoop();
4997 apiv::VLoopBoundaryLayer* mBoundary;
5011 return mBoundary->getConditional();
5016 apiv::VConditionalBoundaryLayer* mBoundary;
5027 apiv::VConditionLayer* mImpl;
5040 apiv::VConditionalOutputLayer* mImpl;
5051 apiv::VConditionalInputLayer* mImpl;
5089 return mImpl->setCondition(condition);
5105 return mImpl->addOutput(trueSubgraphOutput, falseSubgraphOutput);
5117 return mImpl->addInput(input);
5130 mImpl->setName(name);
5140 return mImpl->getName();
5145 apiv::VIfConditional* mImpl;
5174 apiv::VRecurrenceLayer* mImpl;
5199 return mImpl->getLoopOutput();
5216 mImpl->setAxis(axis);
5222 return mImpl->getAxis();
5249 apiv::VLoopOutputLayer* mImpl;
5257 return mImpl->getTripLimit();
5262 apiv::VTripLimitLayer* mImpl;
5271 mImpl->setAxis(axis);
5277 return mImpl->getAxis();
5287 mImpl->setReverse(reverse);
5293 return mImpl->getReverse();
5298 apiv::VIteratorLayer* mImpl;
5317 return mImpl->addRecurrence(initialValue);
5338 return mImpl->addTripLimit(tensor, limit);
5351 return mImpl->addIterator(tensor, axis, reverse);
5363 return mImpl->addLoopOutput(tensor, outputKind, axis);
5376 mImpl->setName(name);
5386 return mImpl->getName();
5390 virtual ~ILoop() noexcept = default;
5401 apiv::VSelectLayer* mImpl;
5431 mImpl->setMessage(message);
5441 return mImpl->getMessage();
5447 apiv::VAssertionLayer* mImpl;
5509 mImpl->setDimensions(dimensions);
5524 return mImpl->getDimensions();
5534 mImpl->setOperation(op);
5544 return mImpl->getOperation();
5562 mImpl->setAlpha(alpha);
5577 return mImpl->getAlpha();
5595 mImpl->setBeta(beta);
5610 return mImpl->getBeta();
5643 apiv::VFillLayer* mImpl;
5716 return mImpl->getAxis();
5727 mImpl->setAxis(axis);
5732 apiv::VQuantizeLayer* mImpl;
5803 return mImpl->getAxis();
5814 mImpl->setAxis(axis);
5819 apiv::VDequantizeLayer* mImpl;
5872 return mImpl->setEquation(equation);
5882 return mImpl->getEquation();
5887 apiv::VEinsumLayer* mImpl;
5974 mImpl->setMode(mode);
5984 return mImpl->getMode();
5994 mImpl->setAxis(axis);
6002 return mImpl->getAxis();
6006 apiv::VScatterLayer* mImpl;
6071 return mImpl->addInput(name, type, dimensions);
6085 mImpl->markOutput(tensor);
6109 return mImpl->addConvolution(input, nbOutputMaps, kernelSize, kernelWeights, biasWeights);
6130 return mImpl->addFullyConnected(input, nbOutputs, kernelWeights, biasWeights);
6149 return mImpl->addActivation(input, type);
6168 return mImpl->addPooling(input, type, windowSize);
6187 return mImpl->addLRN(input, window, alpha, beta, k);
6214 return mImpl->addScale(input, mode, shift, scale, power);
6227 return mImpl->addSoftMax(input);
6244 return mImpl->addConcatenation(inputs, nbInputs);
6268 return mImpl->addDeconvolution(input, nbOutputMaps, kernelSize, kernelWeights, biasWeights);
6293 return mImpl->addElementWise(input1, input2, op);
6313 return mImpl->addUnary(input, operation);
6330 return mImpl->addPadding(input, prePadding, postPadding);
6344 return mImpl->addShuffle(input);
6356 return mImpl->getNbLayers();
6370 return mImpl->getLayer(index);
6382 return mImpl->getNbInputs();
6398 return mImpl->getInput(index);
6412 return mImpl->getNbOutputs();
6428 return mImpl->getOutput(index);
6469 return mImpl->addReduce(input, operation, reduceAxes, keepDimensions);
6502 return mImpl->addTopK(input, op, k, reduceAxes);
6518 return mImpl->addGather(data, indices, axis);
6534 return mImpl->addGatherV2(data, indices, mode);
6552 return mImpl->addRaggedSoftMax(input, bounds);
6572 return mImpl->addMatrixMultiply(input0, op0, input1, op1);
6597 return mImpl->addConstant(dimensions, weights);
6664 ITensor& input, int32_t layerCount, int32_t hiddenSize, int32_t maxSeqLen,
RNNOperation op)
noexcept
6666 return mImpl->addRNNv2(input, layerCount, hiddenSize, maxSeqLen, op);
6680 return mImpl->addIdentity(input);
6695 mImpl->removeTensor(tensor);
6707 mImpl->unmarkOutput(tensor);
6726 return mImpl->addPluginV2(inputs, nbInputs, plugin);
6745 return mImpl->addSlice(input, start, size, stride);
6767 mImpl->setName(name);
6781 return mImpl->getName();
6799 return mImpl->addShape(input);
6818 return mImpl->hasImplicitBatchDimension();
6836 return mImpl->markOutputForShapes(tensor);
6848 return mImpl->unmarkOutputForShapes(tensor);
6866 return mImpl->addParametricReLU(input, slope);
6889 return mImpl->addConvolutionNd(input, nbOutputMaps, kernelSize, kernelWeights, biasWeights);
6908 return mImpl->addPoolingNd(input, type, windowSize);
6931 return mImpl->addDeconvolutionNd(input, nbOutputMaps, kernelSize, kernelWeights, biasWeights);
6967 return mImpl->addScaleNd(input, mode, shift, scale, power, channelAxis);
6982 return mImpl->addResize(input);
6999 return mImpl->hasExplicitPrecision();
7015 return mImpl->addLoop();
7053 return mImpl->addSelect(condition, thenInput, elseInput);
7070 return mImpl->addAssertion(condition, message);
7088 return mImpl->addFill(dimensions, op);
7105 return mImpl->addPaddingNd(input, prePadding, postPadding);
7125 return mImpl->setWeightsName(weights, name);
7144 mImpl->setErrorRecorder(recorder);
7159 return mImpl->getErrorRecorder();
7178 return mImpl->addDequantize(input, scale);
7198 return mImpl->addScatter(data, indices, updates, mode);
7217 return mImpl->addQuantize(input, scale);
7232 return mImpl->addIfConditional();
7246 return mImpl->addEinsum(inputs, nbInputs, equation);
7250 apiv::VNetworkDefinition* mImpl;
7260 kLEGACY_CALIBRATION = 0,
7261 kENTROPY_CALIBRATION = 1,
7262 kENTROPY_CALIBRATION_2 = 2,
7263 kMINMAX_CALIBRATION = 3,
7307 virtual
bool getBatch(
void* bindings[], const
char* names[], int32_t nbBindings) noexcept = 0;
7323 virtual const
void* readCalibrationCache(std::
size_t& length) noexcept = 0;
7333 virtual
void writeCalibrationCache(const
void* ptr, std::
size_t length) noexcept = 0;
7357 return CalibrationAlgoType::kENTROPY_CALIBRATION;
7375 return CalibrationAlgoType::kENTROPY_CALIBRATION_2;
7392 return CalibrationAlgoType::kMINMAX_CALIBRATION;
7410 return CalibrationAlgoType::kLEGACY_CALIBRATION;
7427 virtual
double getRegressionCutoff() const noexcept = 0;
7441 virtual const
void* readHistogramCache(std::
size_t& length) noexcept = 0;
7451 virtual
void writeHistogramCache(const
void* ptr, std::
size_t length) noexcept = 0;
7474 return mImpl->getTensorFormat();
7482 return mImpl->getDataType();
7490 return mImpl->getStrides();
7495 apiv::VAlgorithmIOInfo* mImpl;
7517 return mImpl->getImplementation();
7525 return mImpl->getTactic();
7530 apiv::VAlgorithmVariant* mImpl;
7550 return mImpl->getName();
7561 return mImpl->getDimensions(index, select);
7569 return mImpl->getNbInputs();
7577 return mImpl->getNbOutputs();
7582 apiv::VAlgorithmContext* mImpl;
7609 return mImpl->getAlgorithmIOInfo(index);
7617 return mImpl->getAlgorithmVariant();
7625 return mImpl->getTimingMSec();
7633 return mImpl->getWorkspaceSize();
7646 return mImpl->getAlgorithmIOInfoByIndex(index);
7651 apiv::VAlgorithm* mImpl;
7680 int32_t nbChoices, int32_t* selection)
noexcept
7693 int32_t nbAlgorithms)
noexcept
7830 return mImpl->serialize();
7854 return mImpl->combine(inputCache, ignoreMismatch);
7864 return mImpl->reset();
7868 apiv::VTimingCache* mImpl;
7893 mImpl->setMinTimingIterations(minTiming);
7905 return mImpl->getMinTimingIterations();
7918 mImpl->setAvgTimingIterations(avgTiming);
7930 return mImpl->getAvgTimingIterations();
7943 mImpl->setEngineCapability(capability);
7955 return mImpl->getEngineCapability();
7965 mImpl->setInt8Calibrator(calibrator);
7973 return mImpl->getInt8Calibrator();
7985 mImpl->setMaxWorkspaceSize(workspaceSize);
7999 return mImpl->getMaxWorkspaceSize();
8016 mImpl->setFlags(builderFlags);
8028 return mImpl->getFlags();
8040 mImpl->clearFlag(builderFlag);
8052 mImpl->setFlag(builderFlag);
8064 return mImpl->getFlag(builderFlag);
8079 mImpl->setDeviceType(layer, deviceType);
8088 return mImpl->getDeviceType(layer);
8098 return mImpl->isDeviceTypeSet(layer);
8108 mImpl->resetDeviceType(layer);
8117 return mImpl->canRunOnDLA(layer);
8132 mImpl->setDLACore(dlaCore);
8143 return mImpl->getDLACore();
8153 mImpl->setDefaultDeviceType(deviceType);
8163 return mImpl->getDefaultDeviceType();
8199 return mImpl->setProfileStream(stream);
8211 return mImpl->getProfileStream();
8227 return mImpl->addOptimizationProfile(profile);
8240 return mImpl->getNbOptimizationProfiles();
8252 mImpl->setProfilingVerbosity(verbosity);
8265 return mImpl->getProfilingVerbosity();
8274 mImpl->setAlgorithmSelector(selector);
8282 return mImpl->getAlgorithmSelector();
8297 return mImpl->setCalibrationProfile(profile);
8307 return mImpl->getCalibrationProfile();
8324 mImpl->setQuantizationFlags(flags);
8336 return mImpl->getQuantizationFlags();
8348 mImpl->clearQuantizationFlag(flag);
8360 mImpl->setQuantizationFlag(flag);
8372 return mImpl->getQuantizationFlag(flag);
8397 return mImpl->setTacticSources(tacticSources);
8412 return mImpl->getTacticSources();
8431 return mImpl->createTimingCache(blob, size);
8454 return mImpl->setTimingCache(cache, ignoreMismatch);
8464 return mImpl->getTimingCache();
8468 apiv::VBuilderConfig* mImpl;
8540 mImpl->setMaxBatchSize(batchSize);
8553 return mImpl->getMaxBatchSize();
8561 return mImpl->platformHasFastFp16();
8569 return mImpl->platformHasFastInt8();
8593 return mImpl->getMaxDLABatchSize();
8601 return mImpl->getNbDLACores();
8617 mImpl->setGpuAllocator(allocator);
8627 return mImpl->createBuilderConfig();
8643 return mImpl->buildEngineWithConfig(network, config);
8659 return mImpl->createNetworkV2(flags);
8673 return mImpl->createOptimizationProfile();
8692 mImpl->setErrorRecorder(recorder);
8707 return mImpl->getErrorRecorder();
8723 return mImpl->platformHasTf32();
8742 return mImpl->buildSerializedNetwork(network, config);
8766 return mImpl->isNetworkSupported(network, config);
8776 return mImpl->getLogger();
8780 apiv::VBuilder* mImpl;
8789extern "C" TENSORRTAPI
void* createInferBuilder_INTERNAL(
void* logger, int32_t version)
noexcept;
8805 return static_cast<IBuilder*
>(createInferBuilder_INTERNAL(&logger, NV_TENSORRT_VERSION));
#define TRT_DEPRECATED
< Items that are marked as deprecated will be removed in a future release.
Definition: NvInferRuntimeCommon.h:77
Definition: NvInferRuntimeCommon.h:190
static constexpr int32_t MAX_DIMS
The maximum rank (number of dimensions) supported for a tensor.
Definition: NvInferRuntimeCommon.h:193
Descriptor for two-dimensional spatial data.
Definition: NvInferLegacyDims.h:101
An Activation layer in a network definition.
Definition: NvInfer.h:1570
void setBeta(float beta) noexcept
Set the beta parameter (must be finite).
Definition: NvInfer.h:1618
void setActivationType(ActivationType type) noexcept
Set the type of activation to be performed.
Definition: NvInfer.h:1579
ActivationType getActivationType() const noexcept
Get the type of activation to be performed.
Definition: NvInfer.h:1589
float getAlpha() const noexcept
Get the alpha parameter.
Definition: NvInfer.h:1627
float getBeta() const noexcept
Get the beta parameter.
Definition: NvInfer.h:1636
void setAlpha(float alpha) noexcept
Set the alpha parameter (must be finite).
Definition: NvInfer.h:1604
Describes the context and requirements, that could be fulfilled by one or more instances of IAlgorith...
Definition: NvInfer.h:7542
const char * getName() const noexcept
Return name of the algorithm node. This is a unique identifier for the IAlgorithmContext.
Definition: NvInfer.h:7548
int32_t getNbOutputs() const noexcept
Return number of outputs of the algorithm.
Definition: NvInfer.h:7575
int32_t getNbInputs() const noexcept
Return number of inputs of the algorithm.
Definition: NvInfer.h:7567
Dims getDimensions(int32_t index, OptProfileSelector select) const noexcept
Get the minimum / optimum / maximum dimensions for input or output tensor.
Definition: NvInfer.h:7559
Describes a variation of execution of a layer. An algorithm is represented by IAlgorithmVariant and t...
Definition: NvInfer.h:7595
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:7631
float getTimingMSec() const noexcept
The time in milliseconds to execute the algorithm.
Definition: NvInfer.h:7623
TRT_DEPRECATED const IAlgorithmIOInfo & getAlgorithmIOInfo(int32_t index) const noexcept
Returns the format of an Algorithm input or output. Algorithm inputs are incrementally numbered first...
Definition: NvInfer.h:7607
const IAlgorithmVariant & getAlgorithmVariant() const noexcept
Returns the algorithm variant.
Definition: NvInfer.h:7615
const IAlgorithmIOInfo * getAlgorithmIOInfoByIndex(int32_t index) const noexcept
Returns the format of an Algorithm input or output. Algorithm inputs are incrementally numbered first...
Definition: NvInfer.h:7644
Carries information about input or output of the algorithm. IAlgorithmIOInfo for all the input and ou...
Definition: NvInfer.h:7467
Dims getStrides() const noexcept
Return strides of the input/output tensor of algorithm.
Definition: NvInfer.h:7488
DataType getDataType() const noexcept
Return DataType of the input/output of algorithm.
Definition: NvInfer.h:7480
TensorFormat getTensorFormat() const noexcept
Return TensorFormat of the input/output of algorithm.
Definition: NvInfer.h:7472
Interface implemented by application for selecting and reporting algorithms of a layer provided by th...
Definition: NvInfer.h:7663
virtual int32_t selectAlgorithms(const IAlgorithmContext &context, const IAlgorithm *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(const IAlgorithmContext *const *algoContexts, const IAlgorithm *const *algoChoices, int32_t nbAlgorithms) noexcept=0
Called by TensorRT to report choices it made.
provides a unique 128-bit identifier, which along with the input and output information denotes the v...
Definition: NvInfer.h:7510
int64_t getTactic() const noexcept
Return tactic of the algorithm.
Definition: NvInfer.h:7523
int64_t getImplementation() const noexcept
Return implementation of the algorithm.
Definition: NvInfer.h:7515
An assertion layer in a network.
Definition: NvInfer.h:5419
void setMessage(const char *message) noexcept
Set the message to print if the assertion fails.
Definition: NvInfer.h:5429
const char * getMessage() const noexcept
Return the assertion message.
Definition: NvInfer.h:5439
Holds properties for configuring a builder to produce an engine.
Definition: NvInfer.h:7877
DeviceType getDeviceType(const ILayer *layer) const noexcept
Get the device that this layer executes on.
Definition: NvInfer.h:8086
void setQuantizationFlag(QuantizationFlag flag) noexcept
Set a single quantization flag.
Definition: NvInfer.h:8358
const IOptimizationProfile * getCalibrationProfile() noexcept
Get the current calibration profile.
Definition: NvInfer.h:8305
virtual void setMinTimingIterations(int32_t minTiming) noexcept
Set the number of minimization iterations used when timing layers.
Definition: NvInfer.h:7891
void setInt8Calibrator(IInt8Calibrator *calibrator) noexcept
Set Int8 Calibration interface.
Definition: NvInfer.h:7963
void clearQuantizationFlag(QuantizationFlag flag) noexcept
clear a quantization flag.
Definition: NvInfer.h:8346
void setDeviceType(const ILayer *layer, DeviceType deviceType) noexcept
Set the device that this layer must execute on.
Definition: NvInfer.h:8077
bool setTacticSources(TacticSources tacticSources) noexcept
Set tactic sources.
Definition: NvInfer.h:8395
bool getQuantizationFlag(QuantizationFlag flag) const noexcept
Returns true if the quantization flag is set.
Definition: NvInfer.h:8370
int32_t getDLACore() const noexcept
Get the DLA core that the engine executes on.
Definition: NvInfer.h:8141
void setEngineCapability(EngineCapability capability) noexcept
Configure the builder to target specified EngineCapability flow.
Definition: NvInfer.h:7941
bool getFlag(BuilderFlag builderFlag) const noexcept
Returns true if the build mode flag is set.
Definition: NvInfer.h:8062
void setQuantizationFlags(QuantizationFlags flags) noexcept
Set the quantization flags.
Definition: NvInfer.h:8322
virtual void setAvgTimingIterations(int32_t avgTiming) noexcept
Set the number of averaging iterations used when timing layers.
Definition: NvInfer.h:7916
std::size_t getMaxWorkspaceSize() const noexcept
Get the maximum workspace size.
Definition: NvInfer.h:7997
void setProfilingVerbosity(ProfilingVerbosity verbosity) noexcept
Set verbosity level of layer information exposed in NVTX annotations and IEngineInspector.
Definition: NvInfer.h:8250
nvinfer1::ITimingCache * createTimingCache(const void *blob, std::size_t size) const noexcept
Create timing cache.
Definition: NvInfer.h:8429
IAlgorithmSelector * getAlgorithmSelector() const noexcept
Get Algorithm Selector.
Definition: NvInfer.h:8280
int32_t getNbOptimizationProfiles() const noexcept
Get number of optimization profiles.
Definition: NvInfer.h:8238
QuantizationFlags getQuantizationFlags() const noexcept
Get the quantization flags.
Definition: NvInfer.h:8334
bool setCalibrationProfile(const IOptimizationProfile *profile) noexcept
Add a calibration profile.
Definition: NvInfer.h:8295
virtual int32_t getMinTimingIterations() const noexcept
Query the number of minimization iterations.
Definition: NvInfer.h:7903
void reset() noexcept
Resets the builder configuration to defaults.
Definition: NvInfer.h:8171
TRT_DEPRECATED void destroy() noexcept
Delete this IBuilderConfig.
Definition: NvInfer.h:8185
EngineCapability getEngineCapability() const noexcept
Query EngineCapability flow configured for the builder.
Definition: NvInfer.h:7953
void setAlgorithmSelector(IAlgorithmSelector *selector) noexcept
Set Algorithm Selector.
Definition: NvInfer.h:8272
DeviceType getDefaultDeviceType() const noexcept
Get the default DeviceType which was set by setDefaultDeviceType.
Definition: NvInfer.h:8161
BuilderFlags getFlags() const noexcept
Get the build mode flags for this builder config. Defaults to 0.
Definition: NvInfer.h:8026
void setFlags(BuilderFlags builderFlags) noexcept
Set the build mode flags to turn on builder options for this network.
Definition: NvInfer.h:8014
TacticSources getTacticSources() const noexcept
Get tactic sources.
Definition: NvInfer.h:8410
bool setTimingCache(const ITimingCache &cache, bool ignoreMismatch) noexcept
Attach a timing cache to IBuilderConfig.
Definition: NvInfer.h:8452
bool canRunOnDLA(const ILayer *layer) const noexcept
Checks if a layer can run on DLA.
Definition: NvInfer.h:8115
void setDLACore(int32_t dlaCore) noexcept
Sets the DLA core used by the network.
Definition: NvInfer.h:8130
void setMaxWorkspaceSize(std::size_t workspaceSize) noexcept
Set the maximum workspace size.
Definition: NvInfer.h:7983
int32_t addOptimizationProfile(const IOptimizationProfile *profile) noexcept
Add an optimization profile.
Definition: NvInfer.h:8225
const nvinfer1::ITimingCache * getTimingCache() const noexcept
Get the pointer to the timing cache from current IBuilderConfig.
Definition: NvInfer.h:8462
bool isDeviceTypeSet(const ILayer *layer) const noexcept
whether the DeviceType has been explicitly set for this layer
Definition: NvInfer.h:8096
void clearFlag(BuilderFlag builderFlag) noexcept
clear a single build mode flag.
Definition: NvInfer.h:8038
int32_t getAvgTimingIterations() const noexcept
Query the number of averaging iterations.
Definition: NvInfer.h:7928
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:8151
void setFlag(BuilderFlag builderFlag) noexcept
Set a single build mode flag.
Definition: NvInfer.h:8050
cudaStream_t getProfileStream() const noexcept
Get the cuda stream that is used to profile this network.
Definition: NvInfer.h:8209
IInt8Calibrator * getInt8Calibrator() const noexcept
Get Int8 Calibration interface.
Definition: NvInfer.h:7971
ProfilingVerbosity getProfilingVerbosity() const noexcept
Get verbosity level of layer information exposed in NVTX annotations and IEngineInspector.
Definition: NvInfer.h:8263
void resetDeviceType(const ILayer *layer) noexcept
reset the DeviceType for this layer
Definition: NvInfer.h:8106
void setProfileStream(const cudaStream_t stream) noexcept
Set the cuda stream that is used to profile this network.
Definition: NvInfer.h:8197
Builds an engine from a network definition.
Definition: NvInfer.h:8526
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:8591
int32_t getNbDLACores() const noexcept
Return the number of DLA engines available to this builder.
Definition: NvInfer.h:8599
IErrorRecorder * getErrorRecorder() const noexcept
get the ErrorRecorder assigned to this interface.
Definition: NvInfer.h:8705
ILogger * getLogger() const noexcept
get the logger with which the builder was created
Definition: NvInfer.h:8774
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:8764
bool platformHasTf32() const noexcept
Determine whether the platform has TF32 support.
Definition: NvInfer.h:8721
TRT_DEPRECATED void destroy() noexcept
Destroy this object.
Definition: NvInfer.h:8579
int32_t getMaxBatchSize() const noexcept
Get the maximum batch size.
Definition: NvInfer.h:8551
nvinfer1::IOptimizationProfile * createOptimizationProfile() noexcept
Create a new optimization profile.
Definition: NvInfer.h:8671
bool platformHasFastFp16() const noexcept
Determine whether the platform has fast native fp16.
Definition: NvInfer.h:8559
void setGpuAllocator(IGpuAllocator *allocator) noexcept
Set the GPU allocator.
Definition: NvInfer.h:8615
nvinfer1::INetworkDefinition * createNetworkV2(NetworkDefinitionCreationFlags flags) noexcept
Create a network definition object.
Definition: NvInfer.h:8657
nvinfer1::IBuilderConfig * createBuilderConfig() noexcept
Create a builder configuration object.
Definition: NvInfer.h:8625
void reset() noexcept
Resets the builder state to default values.
Definition: NvInfer.h:8713
void setErrorRecorder(IErrorRecorder *recorder) noexcept
Set the ErrorRecorder for this interface.
Definition: NvInfer.h:8690
bool platformHasFastInt8() const noexcept
Determine whether the platform has fast native int8.
Definition: NvInfer.h:8567
nvinfer1::IHostMemory * buildSerializedNetwork(INetworkDefinition &network, IBuilderConfig &config) noexcept
Builds and serializes a network for the given INetworkDefinition and IBuilderConfig.
Definition: NvInfer.h:8740
TRT_DEPRECATED nvinfer1::ICudaEngine * buildEngineWithConfig(INetworkDefinition &network, IBuilderConfig &config) noexcept
Builds an engine for the given INetworkDefinition and given IBuilderConfig.
Definition: NvInfer.h:8640
void setMaxBatchSize(int32_t batchSize) noexcept
Set the maximum batch size.
Definition: NvInfer.h:8538
A concatenation layer in a network definition.
Definition: NvInfer.h:2358
void setAxis(int32_t axis) noexcept
Set the axis along which concatenation occurs.
Definition: NvInfer.h:2370
int32_t getAxis() const noexcept
Get the axis along which concatenation occurs.
Definition: NvInfer.h:2380
Definition: NvInfer.h:5023
Layer that represents a constant value.
Definition: NvInfer.h:4486
void setWeights(Weights weights) noexcept
Set the weights for the layer.
Definition: NvInfer.h:4497
Weights getWeights() const noexcept
Get the weights for the layer.
Definition: NvInfer.h:4507
void setDimensions(Dims dimensions) noexcept
Set the dimensions for the layer.
Definition: NvInfer.h:4519
Dims getDimensions() const noexcept
Get the dimensions for the layer.
Definition: NvInfer.h:4531
A convolution layer in a network definition.
Definition: NvInfer.h:1009
TRT_DEPRECATED DimsHW getPadding() const noexcept
Get the padding of the convolution. If the padding is asymmetric, the pre-padding is returned.
Definition: NvInfer.h:1112
TRT_DEPRECATED DimsHW getStride() const noexcept
Get the stride of the convolution.
Definition: NvInfer.h:1080
TRT_DEPRECATED DimsHW getDilation() const noexcept
Get the dilation for a convolution.
Definition: NvInfer.h:1219
void setStrideNd(Dims stride) noexcept
Set the multi-dimension stride of the convolution.
Definition: NvInfer.h:1337
void setKernelSizeNd(Dims kernelSize) noexcept
Set the multi-dimension kernel size of the convolution.
Definition: NvInfer.h:1312
Dims getPrePadding() const noexcept
Get the pre-padding.
Definition: NvInfer.h:1246
Weights getBiasWeights() const noexcept
Get the bias weights for the convolution.
Definition: NvInfer.h:1191
int32_t getNbGroups() const noexcept
Get the number of groups of the convolution.
Definition: NvInfer.h:1142
void setPaddingMode(PaddingMode paddingMode) noexcept
Set the padding mode.
Definition: NvInfer.h:1287
TRT_DEPRECATED void setStride(DimsHW stride) noexcept
Get the stride of the convolution.
Definition: NvInfer.h:1070
Dims getPaddingNd() const noexcept
Get the multi-dimension padding of the convolution.
Definition: NvInfer.h:1377
Dims getStrideNd() const noexcept
Get the multi-dimension stride of the convolution.
Definition: NvInfer.h:1347
TRT_DEPRECATED void setPadding(DimsHW padding) noexcept
Set the padding of the convolution.
Definition: NvInfer.h:1100
Weights getKernelWeights() const noexcept
Get the kernel weights of the convolution.
Definition: NvInfer.h:1166
void setPaddingNd(Dims padding) noexcept
Set the multi-dimension padding of the convolution.
Definition: NvInfer.h:1365
Dims getDilationNd() const noexcept
Get the multi-dimension dilation of the convolution.
Definition: NvInfer.h:1401
void setKernelWeights(Weights weights) noexcept
Set the kernel weights for the convolution.
Definition: NvInfer.h:1156
void setDilationNd(Dims dilation) noexcept
Set the multi-dimension dilation of the convolution.
Definition: NvInfer.h:1391
Dims getPostPadding() const noexcept
Get the post-padding.
Definition: NvInfer.h:1273
TRT_DEPRECATED void setDilation(DimsHW dilation) noexcept
Set the dilation for a convolution.
Definition: NvInfer.h:1207
void setNbGroups(int32_t nbGroups) noexcept
Set the number of groups for a convolution.
Definition: NvInfer.h:1132
PaddingMode getPaddingMode() const noexcept
Get the padding mode.
Definition: NvInfer.h:1299
int32_t getNbOutputMaps() const noexcept
Get the number of output maps for the convolution.
Definition: NvInfer.h:1054
void setPrePadding(Dims padding) noexcept
Set the multi-dimension pre-padding of the convolution.
Definition: NvInfer.h:1236
void setBiasWeights(Weights weights) noexcept
Set the bias weights for the convolution.
Definition: NvInfer.h:1181
Dims getKernelSizeNd() const noexcept
Get the multi-dimension kernel size of the convolution.
Definition: NvInfer.h:1322
void setNbOutputMaps(int32_t nbOutputMaps) noexcept
Set the number of output maps for the convolution.
Definition: NvInfer.h:1044
TRT_DEPRECATED void setKernelSize(DimsHW kernelSize) noexcept
Set the HW kernel size of the convolution.
Definition: NvInfer.h:1020
void setPostPadding(Dims padding) noexcept
Set the multi-dimension post-padding of the convolution.
Definition: NvInfer.h:1263
TRT_DEPRECATED DimsHW getKernelSize() const noexcept
Get the HW kernel size of the convolution.
Definition: NvInfer.h:1032
An engine for executing inference on a built network, with functionally unsafe features.
Definition: NvInferRuntime.h:1310
A deconvolution layer in a network definition.
Definition: NvInfer.h:2398
void setBiasWeights(Weights weights) noexcept
Set the bias weights for the deconvolution.
Definition: NvInfer.h:2577
TRT_DEPRECATED void setStride(DimsHW stride) noexcept
Set the stride of the deconvolution.
Definition: NvInfer.h:2462
void setNbOutputMaps(int32_t nbOutputMaps) noexcept
Set the number of output feature maps for the deconvolution.
Definition: NvInfer.h:2435
Weights getKernelWeights() const noexcept
Get the kernel weights for the deconvolution.
Definition: NvInfer.h:2562
void setPrePadding(Dims padding) noexcept
Set the multi-dimension pre-padding of the deconvolution.
Definition: NvInfer.h:2605
TRT_DEPRECATED DimsHW getPadding() const noexcept
Get the padding of the deconvolution.
Definition: NvInfer.h:2508
Dims getStrideNd() const noexcept
Get the multi-dimension stride of the deconvolution.
Definition: NvInfer.h:2722
int32_t getNbGroups() const noexcept
Get the number of groups for a deconvolution.
Definition: NvInfer.h:2538
Dims getDilationNd() const noexcept
Get the multi-dimension dilation of the deconvolution.
Definition: NvInfer.h:2796
Weights getBiasWeights() const noexcept
Get the bias weights for the deconvolution.
Definition: NvInfer.h:2587
void setKernelWeights(Weights weights) noexcept
Set the kernel weights for the deconvolution.
Definition: NvInfer.h:2552
void setDilationNd(Dims dilation) noexcept
Set the multi-dimension dilation of the deconvolution.
Definition: NvInfer.h:2786
TRT_DEPRECATED DimsHW getStride() const noexcept
Get the stride of the deconvolution.
Definition: NvInfer.h:2474
TRT_DEPRECATED void setPadding(DimsHW padding) noexcept
Set the padding of the deconvolution.
Definition: NvInfer.h:2494
Dims getPostPadding() const noexcept
Get the padding.
Definition: NvInfer.h:2643
Dims getKernelSizeNd() const noexcept
Get the multi-dimension kernel size of the deconvolution.
Definition: NvInfer.h:2694
void setKernelSizeNd(Dims kernelSize) noexcept
Set the multi-dimension kernel size of the deconvolution.
Definition: NvInfer.h:2684
TRT_DEPRECATED void setKernelSize(DimsHW kernelSize) noexcept
Set the HW kernel size of the convolution.
Definition: NvInfer.h:2411
void setStrideNd(Dims stride) noexcept
Set the multi-dimension stride of the deconvolution.
Definition: NvInfer.h:2712
Dims getPaddingNd() const noexcept
Get the multi-dimension padding of the deconvolution.
Definition: NvInfer.h:2752
void setNbGroups(int32_t nbGroups) noexcept
Set the number of groups for a deconvolution.
Definition: NvInfer.h:2528
void setPaddingMode(PaddingMode paddingMode) noexcept
Set the padding mode.
Definition: NvInfer.h:2657
void setPaddingNd(Dims padding) noexcept
Set the multi-dimension padding of the deconvolution.
Definition: NvInfer.h:2740
TRT_DEPRECATED DimsHW getKernelSize() const noexcept
Get the HW kernel size of the deconvolution.
Definition: NvInfer.h:2423
void setPostPadding(Dims padding) noexcept
Set the multi-dimension post-padding of the deconvolution.
Definition: NvInfer.h:2633
int32_t getNbOutputMaps() const noexcept
Get the number of output feature maps for the deconvolution.
Definition: NvInfer.h:2445
Dims getPrePadding() const noexcept
Get the pre-padding.
Definition: NvInfer.h:2615
PaddingMode getPaddingMode() const noexcept
Get the padding mode.
Definition: NvInfer.h:2669
A Dequantize layer in a network definition.
Definition: NvInfer.h:5791
int32_t getAxis() const noexcept
Get the quantization axis.
Definition: NvInfer.h:5801
void setAxis(int32_t axis) noexcept
Set the quantization axis.
Definition: NvInfer.h:5812
An Einsum layer in a network.
Definition: NvInfer.h:5859
bool setEquation(const char *equation) noexcept
Set the equation. The equation is a comma-separated list of subscript labels, where each label refers...
Definition: NvInfer.h:5870
const char * getEquation() const noexcept
Return the equation.
Definition: NvInfer.h:5880
A elementwise layer in a network definition.
Definition: NvInfer.h:2861
ElementWiseOperation getOperation() const noexcept
Get the binary operation for the layer.
Definition: NvInfer.h:2884
void setOperation(ElementWiseOperation op) noexcept
Set the binary operation for the layer.
Definition: NvInfer.h:2872
Reference counted application-implemented error reporting interface for TensorRT objects.
Definition: NvInferRuntimeCommon.h:1693
Generate an output tensor with specified mode.
Definition: NvInfer.h:5496
FillOperation getOperation() const noexcept
Get the fill operation for the layer.
Definition: NvInfer.h:5542
void setOperation(FillOperation op) noexcept
Set the fill operation for the layer.
Definition: NvInfer.h:5532
void setDimensions(Dims dimensions) noexcept
Set the output tensor's dimensions.
Definition: NvInfer.h:5507
void setBeta(double beta) noexcept
Set the beta parameter.
Definition: NvInfer.h:5593
double getAlpha() const noexcept
Get the value of alpha parameter.
Definition: NvInfer.h:5575
void setAlpha(double alpha) noexcept
Set the alpha parameter.
Definition: NvInfer.h:5560
Dims getDimensions() const noexcept
Get the output tensor's dimensions.
Definition: NvInfer.h:5522
double getBeta() const noexcept
Get the value of beta parameter.
Definition: NvInfer.h:5608
A fully connected layer in a network definition. This layer expects an input tensor of three or more ...
Definition: NvInfer.h:1464
void setNbOutputChannels(int32_t nbOutputs) noexcept
Set the number of output channels K from the fully connected layer.
Definition: NvInfer.h:1473
Weights getKernelWeights() const noexcept
Get the kernel weights.
Definition: NvInfer.h:1503
void setBiasWeights(Weights weights) noexcept
Set the bias weights.
Definition: NvInfer.h:1515
void setKernelWeights(Weights weights) noexcept
Set the kernel weights, given as a KxC matrix in row-major order.
Definition: NvInfer.h:1493
int32_t getNbOutputChannels() const noexcept
Get the number of output channels K from the fully connected layer.
Definition: NvInfer.h:1483
Weights getBiasWeights() const noexcept
Get the bias weights.
Definition: NvInfer.h:1525
A Gather layer in a network definition. Supports several kinds of gathering.
Definition: NvInfer.h:2992
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:3003
void setNbElementWiseDims(int32_t elementWiseDims) noexcept
Set the number of leading dimensions of indices tensor to be handled elementwise. The gathering of in...
Definition: NvInfer.h:3035
int32_t getNbElementWiseDims() const noexcept
Get the number of leading dimensions of indices tensor to be handled elementwise.
Definition: NvInfer.h:3045
void setMode(GatherMode mode) noexcept
Set the gather mode.
Definition: NvInfer.h:3055
int32_t getGatherAxis() const noexcept
Get the axis to gather on.
Definition: NvInfer.h:3014
GatherMode getMode() const noexcept
Get the gather mode.
Definition: NvInfer.h:3065
Application-implemented class for controlling allocation on the GPU.
Definition: NvInferRuntimeCommon.h:1372
Class to handle library allocated memory that is accessible to the user.
Definition: NvInferRuntime.h:175
A layer that represents the identity function.
Definition: NvInfer.h:4472
Definition: NvInfer.h:5006
IIfConditional * getConditional() const noexcept
Return pointer to the IIfConditional associated with this boundary layer.
Definition: NvInfer.h:5009
Definition: NvInfer.h:5076
IIfConditionalInputLayer * addInput(ITensor &input) noexcept
Add an If-conditional input.
Definition: NvInfer.h:5115
void setName(const char *name) noexcept
Set the name of the conditional.
Definition: NvInfer.h:5128
IConditionLayer * setCondition(ITensor &condition) noexcept
Set the condition tensor for this If-Conditional construct.
Definition: NvInfer.h:5087
IIfConditionalOutputLayer * addOutput(ITensor &trueSubgraphOutput, ITensor &falseSubgraphOutput) noexcept
Add an If-conditional output.
Definition: NvInfer.h:5103
const char * getName() const noexcept
Return the name of the conditional.
Definition: NvInfer.h:5138
Definition: NvInfer.h:5036
Application-implemented interface for calibration.
Definition: NvInfer.h:7285
virtual int32_t getBatchSize() const noexcept=0
Get the batch size used for calibration batches.
Definition: NvInfer.h:7368
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:7373
Definition: NvInfer.h:7350
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:7355
Definition: NvInfer.h:7403
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:7408
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:7385
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:7390
Definition: NvInfer.h:5266
void setReverse(bool reverse) noexcept
Definition: NvInfer.h:5285
bool getReverse() const noexcept
True if and only if reversing input.
Definition: NvInfer.h:5291
int32_t getAxis() const noexcept
Get axis being iterated over.
Definition: NvInfer.h:5275
void setAxis(int32_t axis) noexcept
Set axis to iterate over.
Definition: NvInfer.h:5269
A LRN layer in a network definition.
Definition: NvInfer.h:2018
int32_t getWindowSize() const noexcept
Get the LRN window size.
Definition: NvInfer.h:2039
float getAlpha() const noexcept
Get the LRN alpha value.
Definition: NvInfer.h:2060
void setWindowSize(int32_t windowSize) noexcept
Set the LRN window size.
Definition: NvInfer.h:2029
void setK(float k) noexcept
Set the LRN K value.
Definition: NvInfer.h:2092
void setAlpha(float alpha) noexcept
Set the LRN alpha value.
Definition: NvInfer.h:2050
void setBeta(float beta) noexcept
Set the LRN beta value.
Definition: NvInfer.h:2071
float getBeta() const noexcept
Get the LRN beta value.
Definition: NvInfer.h:2081
float getK() const noexcept
Get the LRN K value.
Definition: NvInfer.h:2102
Base class for all layer classes in a network definition.
Definition: NvInfer.h:520
bool precisionIsSet() const noexcept
whether the computational precision has been set for this layer
Definition: NvInfer.h:657
void setPrecision(DataType dataType) noexcept
Set the computational precision of this layer.
Definition: NvInfer.h:633
void resetPrecision() noexcept
reset the computational precision for this layer
Definition: NvInfer.h:667
int32_t getNbInputs() const noexcept
Get the number of inputs of a layer.
Definition: NvInfer.h:558
DataType getOutputType(int32_t index) const noexcept
get the output type of this layer
Definition: NvInfer.h:719
DataType getPrecision() const noexcept
get the computational precision of this layer
Definition: NvInfer.h:645
int32_t getNbOutputs() const noexcept
Get the number of outputs of a layer.
Definition: NvInfer.h:579
void setName(const char *name) noexcept
Set the name of a layer.
Definition: NvInfer.h:539
bool outputTypeIsSet(int32_t index) const noexcept
whether the output type has been set for this layer
Definition: NvInfer.h:732
ITensor * getOutput(int32_t index) const noexcept
Get the layer output corresponding to the given index.
Definition: NvInfer.h:590
void setInput(int32_t index, ITensor &tensor) noexcept
Replace an input of this layer with a specific tensor.
Definition: NvInfer.h:607
void resetOutputType(int32_t index) noexcept
reset the output type for this layer
Definition: NvInfer.h:744
ITensor * getInput(int32_t index) const noexcept
Get the layer input corresponding to the given index.
Definition: NvInfer.h:571
void setOutputType(int32_t index, DataType dataType) noexcept
Set the output type of this layer.
Definition: NvInfer.h:705
LayerType getType() const noexcept
Return the type of a layer.
Definition: NvInfer.h:527
const char * getName() const noexcept
Return the name of a layer.
Definition: NvInfer.h:550
Application-implemented logging interface for the builder, refitter and runtime.
Definition: NvInferRuntimeCommon.h:1506
Definition: NvInfer.h:4987
ILoop * getLoop() const noexcept
Return pointer to ILoop associated with this boundary layer.
Definition: NvInfer.h:4990
Definition: NvInfer.h:5307
ITripLimitLayer * addTripLimit(ITensor &tensor, TripLimit limit) noexcept
Add a trip-count limiter, based on the given tensor.
Definition: NvInfer.h:5336
IIteratorLayer * addIterator(ITensor &tensor, int32_t axis=0, bool reverse=false) noexcept
Return layer that subscripts tensor by loop iteration.
Definition: NvInfer.h:5349
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:5361
const char * getName() const noexcept
Return the name of the loop.
Definition: NvInfer.h:5384
void setName(const char *name) noexcept
Set the name of the loop.
Definition: NvInfer.h:5374
IRecurrenceLayer * addRecurrence(ITensor &initialValue) noexcept
Create a recurrence layer for this loop with initialValue as its first input.
Definition: NvInfer.h:5315
Definition: NvInfer.h:5195
int32_t getAxis() const noexcept
Get axis being concatenated over.
Definition: NvInfer.h:5220
void setAxis(int32_t axis) noexcept
Set where to insert the contenation axis. Ignored if getLoopOutput() is kLAST_VALUE.
Definition: NvInfer.h:5214
Layer that represents a Matrix Multiplication.
Definition: NvInfer.h:4410
MatrixOperation getOperation(int32_t index) const noexcept
Get the operation for an input tensor.
Definition: NvInfer.h:4428
void setOperation(int32_t index, MatrixOperation op) noexcept
Set the operation for an input tensor.
Definition: NvInfer.h:4418
A network definition for input to the builder.
Definition: NvInfer.h:6030
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:6724
IConvolutionLayer * addConvolutionNd(ITensor &input, int32_t nbOutputMaps, Dims kernelSize, Weights kernelWeights, Weights biasWeights) noexcept
Add a multi-dimension convolution layer to the network.
Definition: NvInfer.h:6886
IDequantizeLayer * addDequantize(ITensor &input, ITensor &scale) noexcept
Add a dequantization layer to the network.
Definition: NvInfer.h:7176
IConcatenationLayer * addConcatenation(ITensor *const *inputs, int32_t nbInputs) noexcept
Add a concatenation layer to the network.
Definition: NvInfer.h:6242
TRT_DEPRECATED IDeconvolutionLayer * addDeconvolution(ITensor &input, int32_t nbOutputMaps, DimsHW kernelSize, Weights kernelWeights, Weights biasWeights) noexcept
Add a deconvolution layer to the network.
Definition: NvInfer.h:6265
ITensor * addInput(const char *name, DataType type, Dims dimensions) noexcept
Add an input tensor to the network.
Definition: NvInfer.h:6069
IGatherLayer * addGatherV2(ITensor &data, ITensor &indices, GatherMode mode)
Add gather with specified mode, axis=0 and nbElementWiseDims=0.
Definition: NvInfer.h:6532
IShuffleLayer * addShuffle(ITensor &input) noexcept
Add a shuffle layer to the network.
Definition: NvInfer.h:6342
ILRNLayer * addLRN(ITensor &input, int32_t window, float alpha, float beta, float k) noexcept
Add a LRN layer to the network.
Definition: NvInfer.h:6185
ITopKLayer * addTopK(ITensor &input, TopKOperation op, int32_t k, uint32_t reduceAxes) noexcept
Add a TopK layer to the network.
Definition: NvInfer.h:6500
IAssertionLayer * addAssertion(ITensor &condition, const char *message) noexcept
Add an assertion layer to the network.
Definition: NvInfer.h:7068
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:6964
TRT_DEPRECATED bool hasExplicitPrecision() const noexcept
True if network is an explicit precision network.
Definition: NvInfer.h:6997
IParametricReLULayer * addParametricReLU(ITensor &input, ITensor &slope) noexcept
Add a parametric ReLU layer to the network.
Definition: NvInfer.h:6864
ITensor * getOutput(int32_t index) const noexcept
Get the output tensor specified by the given index.
Definition: NvInfer.h:6426
ITensor * getInput(int32_t index) const noexcept
Get the input tensor specified by the given index.
Definition: NvInfer.h:6396
bool unmarkOutputForShapes(ITensor &tensor) noexcept
Undo markOutputForShapes.
Definition: NvInfer.h:6846
IFillLayer * addFill(Dims dimensions, FillOperation op) noexcept
Add a fill layer to the network.
Definition: NvInfer.h:7086
IFullyConnectedLayer * addFullyConnected(ITensor &input, int32_t nbOutputs, Weights kernelWeights, Weights biasWeights) noexcept
Add a fully connected layer to the network.
Definition: NvInfer.h:6127
ILoop * addLoop() noexcept
Add a loop to the network.
Definition: NvInfer.h:7013
IDeconvolutionLayer * addDeconvolutionNd(ITensor &input, int32_t nbOutputMaps, Dims kernelSize, Weights kernelWeights, Weights biasWeights) noexcept
Add a multi-dimension deconvolution layer to the network.
Definition: NvInfer.h:6928
IActivationLayer * addActivation(ITensor &input, ActivationType type) noexcept
Add an activation layer to the network.
Definition: NvInfer.h:6147
void setName(const char *name) noexcept
Sets the name of the network.
Definition: NvInfer.h:6765
ILayer * getLayer(int32_t index) const noexcept
Get the layer specified by the given index.
Definition: NvInfer.h:6368
TRT_DEPRECATED IRNNv2Layer * addRNNv2(ITensor &input, int32_t layerCount, int32_t hiddenSize, int32_t maxSeqLen, RNNOperation op) noexcept
Add an layerCount deep RNN layer to the network with hiddenSize internal states that can take a batch...
Definition: NvInfer.h:6663
IIfConditional * addIfConditional() noexcept
Add an If-conditional layer to the network.
Definition: NvInfer.h:7230
IErrorRecorder * getErrorRecorder() const noexcept
get the ErrorRecorder assigned to this interface.
Definition: NvInfer.h:7157
int32_t getNbInputs() const noexcept
Get the number of inputs in the network.
Definition: NvInfer.h:6380
bool hasImplicitBatchDimension() const noexcept
Query whether the network was created with an implicit batch dimension.
Definition: NvInfer.h:6816
IReduceLayer * addReduce(ITensor &input, ReduceOperation operation, uint32_t reduceAxes, bool keepDimensions) noexcept
Add a reduce layer to the network.
Definition: NvInfer.h:6466
IUnaryLayer * addUnary(ITensor &input, UnaryOperation operation) noexcept
Add a unary layer to the network.
Definition: NvInfer.h:6311
const char * getName() const noexcept
Returns the name associated with the network.
Definition: NvInfer.h:6779
void removeTensor(ITensor &tensor) noexcept
remove a tensor from the network definition.
Definition: NvInfer.h:6693
ISelectLayer * addSelect(ITensor &condition, ITensor &thenInput, ITensor &elseInput) noexcept
Add a select layer to the network.
Definition: NvInfer.h:7051
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:7196
int32_t getNbLayers() const noexcept
Get the number of layers in the network.
Definition: NvInfer.h:6354
IPoolingLayer * addPoolingNd(ITensor &input, PoolingType type, Dims windowSize) noexcept
Add a multi-dimension pooling layer to the network.
Definition: NvInfer.h:6906
bool markOutputForShapes(ITensor &tensor) noexcept
Enable tensor's value to be computed by IExecutionContext::getShapeBinding.
Definition: NvInfer.h:6834
TRT_DEPRECATED IPaddingLayer * addPadding(ITensor &input, DimsHW prePadding, DimsHW postPadding) noexcept
Add a padding layer to the network.
Definition: NvInfer.h:6328
IScaleLayer * addScale(ITensor &input, ScaleMode mode, Weights shift, Weights scale, Weights power) noexcept
Add a Scale layer to the network.
Definition: NvInfer.h:6212
void unmarkOutput(ITensor &tensor) noexcept
unmark a tensor as a network output.
Definition: NvInfer.h:6705
IIdentityLayer * addIdentity(ITensor &input) noexcept
Add an identity layer.
Definition: NvInfer.h:6678
TRT_DEPRECATED IConvolutionLayer * addConvolution(ITensor &input, int32_t nbOutputMaps, DimsHW kernelSize, Weights kernelWeights, Weights biasWeights) noexcept
Add a convolution layer to the network.
Definition: NvInfer.h:6106
IQuantizeLayer * addQuantize(ITensor &input, ITensor &scale) noexcept
Add a quantization layer to the network.
Definition: NvInfer.h:7215
IElementWiseLayer * addElementWise(ITensor &input1, ITensor &input2, ElementWiseOperation op) noexcept
Add an elementwise layer to the network.
Definition: NvInfer.h:6291
void setErrorRecorder(IErrorRecorder *recorder) noexcept
Set the ErrorRecorder for this interface.
Definition: NvInfer.h:7142
ISliceLayer * addSlice(ITensor &input, Dims start, Dims size, Dims stride) noexcept
Add a slice layer to the network.
Definition: NvInfer.h:6743
IConstantLayer * addConstant(Dims dimensions, Weights weights) noexcept
Add a constant layer to the network.
Definition: NvInfer.h:6595
IRaggedSoftMaxLayer * addRaggedSoftMax(ITensor &input, ITensor &bounds) noexcept
Add a RaggedSoftMax layer to the network.
Definition: NvInfer.h:6550
IShapeLayer * addShape(ITensor &input) noexcept
Add a shape layer to the network.
Definition: NvInfer.h:6797
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:6516
IEinsumLayer * addEinsum(ITensor *const *inputs, int32_t nbInputs, const char *equation) noexcept
Add an Einsum layer to the network.
Definition: NvInfer.h:7244
TRT_DEPRECATED IPoolingLayer * addPooling(ITensor &input, PoolingType type, DimsHW windowSize) noexcept
Add a pooling layer to the network.
Definition: NvInfer.h:6166
TRT_DEPRECATED void destroy() noexcept
Destroy this INetworkDefinition object.
Definition: NvInfer.h:6438
IResizeLayer * addResize(ITensor &input) noexcept
Add a resize layer to the network.
Definition: NvInfer.h:6980
IMatrixMultiplyLayer * addMatrixMultiply(ITensor &input0, MatrixOperation op0, ITensor &input1, MatrixOperation op1) noexcept
Add a MatrixMultiply layer to the network.
Definition: NvInfer.h:6569
ISoftMaxLayer * addSoftMax(ITensor &input) noexcept
Add a SoftMax layer to the network.
Definition: NvInfer.h:6225
void markOutput(ITensor &tensor) noexcept
Mark a tensor as a network output.
Definition: NvInfer.h:6083
bool setWeightsName(Weights weights, const char *name) noexcept
Associate a name with all current uses of the given weights.
Definition: NvInfer.h:7123
int32_t getNbOutputs() const noexcept
Get the number of outputs in the network.
Definition: NvInfer.h:6410
TRT_DEPRECATED IPaddingLayer * addPaddingNd(ITensor &input, Dims prePadding, Dims postPadding) noexcept
Add a padding layer to the network. Only 2D padding is currently supported.
Definition: NvInfer.h:7103
Forward declaration of IEngineInspector for use by other interfaces.
Definition: NvInferRuntime.h:80
Optimization profile for dynamic input dimensions and shape tensors.
Definition: NvInferRuntime.h:1094
Layer that represents a padding operation.
Definition: NvInfer.h:3747
Dims getPostPaddingNd() const noexcept
Get the padding that is applied at the end of the tensor.
Definition: NvInfer.h:3848
TRT_DEPRECATED DimsHW getPrePadding() const noexcept
Get the padding that is applied at the start of the tensor.
Definition: NvInfer.h:3770
TRT_DEPRECATED void setPostPadding(DimsHW padding) noexcept
Set the padding that is applied at the end of the tensor.
Definition: NvInfer.h:3784
TRT_DEPRECATED void setPrePadding(DimsHW padding) noexcept
Set the padding that is applied at the start of the tensor.
Definition: NvInfer.h:3758
Dims getPrePaddingNd() const noexcept
Get the padding that is applied at the start of the tensor.
Definition: NvInfer.h:3822
TRT_DEPRECATED DimsHW getPostPadding() const noexcept
Get the padding that is applied at the end of the tensor.
Definition: NvInfer.h:3796
void setPostPaddingNd(Dims padding) noexcept
Set the padding that is applied at the end of the tensor.
Definition: NvInfer.h:3836
void setPrePaddingNd(Dims padding) noexcept
Set the padding that is applied at the start of the tensor.
Definition: NvInfer.h:3810
Layer that represents a parametric ReLU operation.
Definition: NvInfer.h:4549
Plugin class for user-implemented layers.
Definition: NvInferRuntimeCommon.h:411
Layer type for pluginV2.
Definition: NvInfer.h:3534
IPluginV2 & getPlugin() noexcept
Get the plugin for the layer.
Definition: NvInfer.h:3541
A Pooling layer in a network definition.
Definition: NvInfer.h:1680
TRT_DEPRECATED DimsHW getStride() const noexcept
Get the stride for pooling.
Definition: NvInfer.h:1753
PoolingType getPoolingType() const noexcept
Get the type of activation to be performed.
Definition: NvInfer.h:1699
void setWindowSizeNd(Dims windowSize) noexcept
Set the multi-dimension window size for pooling.
Definition: NvInfer.h:1932
PaddingMode getPaddingMode() const noexcept
Get the padding mode.
Definition: NvInfer.h:1919
Dims getPostPadding() const noexcept
Get the padding.
Definition: NvInfer.h:1895
TRT_DEPRECATED void setStride(DimsHW stride) noexcept
Set the stride for pooling.
Definition: NvInfer.h:1741
bool getAverageCountExcludesPadding() const noexcept
Get whether average pooling uses as a denominator the overlap area between the window and the unpadde...
Definition: NvInfer.h:1839
Dims getPrePadding() const noexcept
Get the pre-padding.
Definition: NvInfer.h:1867
void setPoolingType(PoolingType type) noexcept
Set the type of activation to be performed.
Definition: NvInfer.h:1689
void setPaddingMode(PaddingMode paddingMode) noexcept
Set the padding mode.
Definition: NvInfer.h:1908
TRT_DEPRECATED void setWindowSize(DimsHW windowSize) noexcept
Set the window size for pooling.
Definition: NvInfer.h:1713
Dims getWindowSizeNd() const noexcept
Get the multi-dimension window size for pooling.
Definition: NvInfer.h:1942
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:1828
void setPaddingNd(Dims padding) noexcept
Set the multi-dimension padding for pooling.
Definition: NvInfer.h:1986
void setPrePadding(Dims padding) noexcept
Set the multi-dimension pre-padding for pooling.
Definition: NvInfer.h:1857
float getBlendFactor() const noexcept
Get the blending factor for the max_average_blend mode: max_average_blendPool = (1-blendFactor)*maxPo...
Definition: NvInfer.h:1811
Dims getStrideNd() const noexcept
Get the multi-dimension stride for pooling.
Definition: NvInfer.h:1967
Dims getPaddingNd() const noexcept
Get the multi-dimension padding for pooling.
Definition: NvInfer.h:1998
TRT_DEPRECATED DimsHW getWindowSize() const noexcept
Get the window size for pooling.
Definition: NvInfer.h:1725
TRT_DEPRECATED DimsHW getPadding() const noexcept
Get the padding for pooling.
Definition: NvInfer.h:1783
void setStrideNd(Dims stride) noexcept
Set the multi-dimension stride for pooling.
Definition: NvInfer.h:1957
void setPostPadding(Dims padding) noexcept
Set the multi-dimension post-padding for pooling.
Definition: NvInfer.h:1885
TRT_DEPRECATED void setPadding(DimsHW padding) noexcept
Set the padding for pooling.
Definition: NvInfer.h:1769
void setBlendFactor(float blendFactor) noexcept
Set the blending factor for the max_average_blend mode: max_average_blendPool = (1-blendFactor)*maxPo...
Definition: NvInfer.h:1798
A Quantize layer in a network definition.
Definition: NvInfer.h:5704
void setAxis(int32_t axis) noexcept
Set the quantization axis.
Definition: NvInfer.h:5725
int32_t getAxis() const noexcept
Get the quantization axis.
Definition: NvInfer.h:5714
An RNN layer in a network definition, version 2.
Definition: NvInfer.h:3253
void setDirection(RNNDirection op) noexcept
Set the direction of the RNN layer.
Definition: NvInfer.h:3349
void setBiasForGate(int32_t layerIndex, RNNGateType gate, bool isW, Weights bias) noexcept
Set the bias parameters for an individual gate in the RNN.
Definition: NvInfer.h:3451
void setCellState(ITensor &cell) noexcept
Set the initial cell state of the LSTM with the provided cell ITensor.
Definition: NvInfer.h:3505
int32_t getDataLength() const noexcept
Get the maximum data length of the RNN.
Definition: NvInfer.h:3267
void setInputMode(RNNInputMode op) noexcept
Set the input mode of the RNN layer.
Definition: NvInfer.h:3325
Weights getBiasForGate(int32_t layerIndex, RNNGateType gate, bool isW) const noexcept
Get the bias parameters for an individual gate in the RNN.
Definition: NvInfer.h:3460
void setWeightsForGate(int32_t layerIndex, RNNGateType gate, bool isW, Weights weights) noexcept
Set the weight parameters for an individual gate in the RNN.
Definition: NvInfer.h:3417
void setOperation(RNNOperation op) noexcept
Set the operation of the RNN layer.
Definition: NvInfer.h:3307
Weights getWeightsForGate(int32_t layerIndex, RNNGateType gate, bool isW) const noexcept
Get the weight parameters for an individual gate in the RNN.
Definition: NvInfer.h:3426
RNNDirection getDirection() const noexcept
Get the direction of the RNN layer.
Definition: NvInfer.h:3358
void setSequenceLengths(ITensor &seqLengths) noexcept
Specify individual sequence lengths in the batch with the ITensor pointed to by seqLengths.
Definition: NvInfer.h:3286
RNNInputMode getInputMode() const noexcept
Get the input mode of the RNN layer.
Definition: NvInfer.h:3334
ITensor * getHiddenState() const noexcept
Get the initial hidden state of the RNN.
Definition: NvInfer.h:3486
ITensor * getCellState() const noexcept
Get the initial cell state of the RNN.
Definition: NvInfer.h:3514
int32_t getMaxSeqLength() const noexcept
Get the maximum sequence length of the RNN.
Definition: NvInfer.h:3263
int32_t getLayerCount() const noexcept
Get the layer count of the RNN.
Definition: NvInfer.h:3255
ITensor * getSequenceLengths() const noexcept
Get the sequence lengths specified for the RNN.
Definition: NvInfer.h:3298
RNNOperation getOperation() const noexcept
Get the operation of the RNN layer.
Definition: NvInfer.h:3316
int32_t getHiddenSize() const noexcept
Get the hidden size of the RNN.
Definition: NvInfer.h:3259
void setHiddenState(ITensor &hidden) noexcept
Set the initial hidden state of the RNN with the provided hidden ITensor.
Definition: NvInfer.h:3477
A RaggedSoftmax layer in a network definition.
Definition: NvInfer.h:4453
Definition: NvInfer.h:5150
Layer that represents a reduction operator across Shape, Int32, Float, Half, and Int8 tensors.
Definition: NvInfer.h:3669
void setKeepDimensions(bool keepDimensions) noexcept
Set the boolean that specifies whether or not to keep the reduced dimensions for the layer.
Definition: NvInfer.h:3716
void setOperation(ReduceOperation op) noexcept
Set the reduce operation for the layer.
Definition: NvInfer.h:3676
ReduceOperation getOperation() const noexcept
Get the reduce operation for the layer.
Definition: NvInfer.h:3686
uint32_t getReduceAxes() const noexcept
Get the axes over which to reduce for the layer.
Definition: NvInfer.h:3706
void setReduceAxes(uint32_t reduceAxes) noexcept
Set the axes over which to reduce.
Definition: NvInfer.h:3696
bool getKeepDimensions() const noexcept
Get the boolean that specifies whether or not to keep the reduced dimensions for the layer.
Definition: NvInfer.h:3726
A resize layer in a network definition.
Definition: NvInfer.h:4706
void setSelectorForSinglePixel(ResizeSelector selector) noexcept
Set coordinate selector function when resized to single pixel.
Definition: NvInfer.h:4903
void setOutputDimensions(Dims dimensions) noexcept
Set the output dimensions.
Definition: NvInfer.h:4726
void setNearestRounding(ResizeRoundMode value) noexcept
Set rounding mode for nearest neighbor resize.
Definition: NvInfer.h:4929
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:4785
ResizeMode getResizeMode() const noexcept
Get resize mode for an input tensor.
Definition: NvInfer.h:4809
void setResizeMode(ResizeMode resizeMode) noexcept
Set resize mode for an input tensor.
Definition: NvInfer.h:4799
ResizeSelector getSelectorForSinglePixel() const noexcept
Get the coordinate selector function when resized to single pixel.
Definition: NvInfer.h:4913
TRT_DEPRECATED bool getAlignCorners() const noexcept
True if align corners has been set.
Definition: NvInfer.h:4839
void setCoordinateTransformation(ResizeCoordinateTransformation coordTransform) noexcept
Set coordinate transformation function.
Definition: NvInfer.h:4876
void setScales(const float *scales, int32_t nbScales) noexcept
Set the resize scales.
Definition: NvInfer.h:4766
Dims getOutputDimensions() const noexcept
Get the output dimensions.
Definition: NvInfer.h:4736
ResizeRoundMode getNearestRounding() const noexcept
Get rounding mode for nearest neighbor resize.
Definition: NvInfer.h:4939
TRT_DEPRECATED void setAlignCorners(bool alignCorners) noexcept
Set whether to align corners while resizing.
Definition: NvInfer.h:4826
ResizeCoordinateTransformation getCoordinateTransformation() const noexcept
Get coordinate transformation function.
Definition: NvInfer.h:4886
A Scale layer in a network definition.
Definition: NvInfer.h:2158
Weights getScale() const noexcept
Get the scale value.
Definition: NvInfer.h:2215
Weights getPower() const noexcept
Get the power value.
Definition: NvInfer.h:2235
void setScale(Weights scale) noexcept
Set the scale value.
Definition: NvInfer.h:2205
void setPower(Weights power) noexcept
Set the power value.
Definition: NvInfer.h:2225
ScaleMode getMode() const noexcept
Get the scale mode.
Definition: NvInfer.h:2175
void setShift(Weights shift) noexcept
Set the shift value.
Definition: NvInfer.h:2185
void setChannelAxis(int32_t channelAxis) noexcept
Set the channel axis.
Definition: NvInfer.h:2271
Weights getShift() const noexcept
Get the shift value.
Definition: NvInfer.h:2195
void setMode(ScaleMode mode) noexcept
Set the scale mode.
Definition: NvInfer.h:2165
int32_t getChannelAxis() const noexcept
Get the channel axis.
Definition: NvInfer.h:2250
A scatter layer in a network definition. Supports several kinds of scattering.
Definition: NvInfer.h:5965
void setMode(ScatterMode mode) noexcept
Set the scatter mode.
Definition: NvInfer.h:5972
void setAxis(int32_t axis) noexcept
Set the axis used by ScatterMode::kELEMENTS.
Definition: NvInfer.h:5992
int32_t getAxis() const noexcept
Get the axis.
Definition: NvInfer.h:6000
ScatterMode getMode() const noexcept
Get the scatter mode.
Definition: NvInfer.h:5982
Definition: NvInfer.h:5398
Layer type for getting shape of a tensor.
Definition: NvInfer.h:4248
Layer type for shuffling data.
Definition: NvInfer.h:3882
void setReshapeDimensions(Dims dimensions) noexcept
Set the reshaped dimensions.
Definition: NvInfer.h:3930
void setFirstTranspose(Permutation permutation) noexcept
Set the permutation applied by the first transpose operation.
Definition: NvInfer.h:3893
void setSecondTranspose(Permutation permutation) noexcept
Set the permutation applied by the second transpose operation.
Definition: NvInfer.h:3990
Dims getReshapeDimensions() const noexcept
Get the reshaped dimensions.
Definition: NvInfer.h:3943
Permutation getFirstTranspose() const noexcept
Get the permutation applied by the first transpose operation.
Definition: NvInfer.h:3905
Permutation getSecondTranspose() const noexcept
Get the permutation applied by the second transpose operation.
Definition: NvInfer.h:4002
bool getZeroIsPlaceholder() const noexcept
Get meaning of 0 in reshape dimensions.
Definition: NvInfer.h:4031
void setZeroIsPlaceholder(bool zeroIsPlaceholder) noexcept
Set meaning of 0 in reshape dimensions.
Definition: NvInfer.h:4018
Slices an input tensor into an output tensor based on the offset and strides.
Definition: NvInfer.h:4098
void setMode(SliceMode mode) noexcept
Set the slice mode.
Definition: NvInfer.h:4192
void setStride(Dims stride) noexcept
Set the stride for computing the output slice data.
Definition: NvInfer.h:4167
void setStart(Dims start) noexcept
Set the start offset that the slice layer uses to create the output slice.
Definition: NvInfer.h:4109
Dims getStart() const noexcept
Get the start offset for the slice layer.
Definition: NvInfer.h:4124
void setSize(Dims size) noexcept
Set the dimensions of the output slice.
Definition: NvInfer.h:4138
Dims getSize() const noexcept
Get dimensions of the output slice.
Definition: NvInfer.h:4153
SliceMode getMode() const noexcept
Get the slice mode.
Definition: NvInfer.h:4202
Dims getStride() const noexcept
Get the stride for the output slice.
Definition: NvInfer.h:4182
A Softmax layer in a network definition.
Definition: NvInfer.h:2293
void setAxes(uint32_t axes) noexcept
Set the axis along which softmax is computed. Currently, only one axis can be set.
Definition: NvInfer.h:2325
uint32_t getAxes() const noexcept
Get the axis along which softmax occurs.
Definition: NvInfer.h:2335
A tensor in a network definition.
Definition: NvInfer.h:196
bool setDynamicRange(float min, float max) noexcept
Set dynamic range for the tensor.
Definition: NvInfer.h:296
void setAllowedFormats(TensorFormats formats) noexcept
Set allowed formats for this tensor. By default all formats are allowed. Shape tensors (for which isS...
Definition: NvInfer.h:432
TensorLocation getLocation() const noexcept
Get the storage location of a tensor.
Definition: NvInfer.h:360
void setName(const char *name) noexcept
Set the tensor name.
Definition: NvInfer.h:210
void resetDynamicRange() noexcept
Undo effect of setDynamicRange.
Definition: NvInfer.h:393
bool isExecutionTensor() const noexcept
Whether the tensor is an execution tensor.
Definition: NvInfer.h:502
void setType(DataType type) noexcept
Set the data type of a tensor.
Definition: NvInfer.h:269
bool dynamicRangeIsSet() const noexcept
Query whether dynamic range is set.
Definition: NvInfer.h:385
void setLocation(TensorLocation location) noexcept
Set the storage location of a tensor.
Definition: NvInfer.h:375
bool isShapeTensor() const noexcept
Whether the tensor is a shape tensor.
Definition: NvInfer.h:479
float getDynamicRangeMax() const noexcept
Get maximum of dynamic range.
Definition: NvInfer.h:413
bool isNetworkInput() const noexcept
Whether the tensor is a network input.
Definition: NvInfer.h:304
bool isNetworkOutput() const noexcept
Whether the tensor is a network output.
Definition: NvInfer.h:312
bool getBroadcastAcrossBatch() const noexcept
Check if tensor is broadcast across the batch.
Definition: NvInfer.h:350
void setBroadcastAcrossBatch(bool broadcastAcrossBatch) noexcept
Set whether to enable broadcast of tensor across the batch.
Definition: NvInfer.h:334
DataType getType() const noexcept
Get the data type of a tensor.
Definition: NvInfer.h:281
float getDynamicRangeMin() const noexcept
Get minimum of dynamic range.
Definition: NvInfer.h:403
void setDimensions(Dims dimensions) noexcept
Set the dimensions of a tensor.
Definition: NvInfer.h:241
const char * getName() const noexcept
Get the tensor name.
Definition: NvInfer.h:222
Dims getDimensions() const noexcept
Get the dimensions of a tensor.
Definition: NvInfer.h:254
TensorFormats getAllowedFormats() const noexcept
Get a bitmask of TensorFormat values that the tensor supports. For a shape tensor,...
Definition: NvInfer.h:445
Class to handle tactic timing info collected from builder.
Definition: NvInfer.h:7815
nvinfer1::IHostMemory * serialize() const noexcept
Serialize a timing cache to IHostMemory object.
Definition: NvInfer.h:7828
bool combine(const ITimingCache &inputCache, bool ignoreMismatch) noexcept
Combine input timing cache into local instance.
Definition: NvInfer.h:7852
bool reset() noexcept
Empty the timing cache.
Definition: NvInfer.h:7862
Layer that represents a TopK reduction.
Definition: NvInfer.h:4280
void setK(int32_t k) noexcept
Set the k value for the layer.
Definition: NvInfer.h:4309
void setReduceAxes(uint32_t reduceAxes) noexcept
Set which axes to reduce for the layer.
Definition: NvInfer.h:4329
TopKOperation getOperation() const noexcept
Get the operation for the layer.
Definition: NvInfer.h:4297
void setOperation(TopKOperation op) noexcept
Set the operation for the layer.
Definition: NvInfer.h:4287
int32_t getK() const noexcept
Get the k value for the layer.
Definition: NvInfer.h:4319
uint32_t getReduceAxes() const noexcept
Get the axes to reduce for the layer.
Definition: NvInfer.h:4339
Definition: NvInfer.h:5253
Layer that represents an unary operation.
Definition: NvInfer.h:3600
void setOperation(UnaryOperation op) noexcept
Set the unary operation for the layer.
Definition: NvInfer.h:3607
UnaryOperation getOperation() const noexcept
Get the unary operation for the layer.
Definition: NvInfer.h:3617
An array of weights used as a layer parameter.
Definition: NvInferRuntime.h:157
IBuilder * createInferBuilder(ILogger &logger) noexcept
Create an instance of an IBuilder class.
Definition: NvInfer.h:8803
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:1275
ResizeSelector
The coordinate selector when resize to single pixel output.
Definition: NvInfer.h:4634
@ kFORMULA
Use formula to map the original index.
@ kUPPER
Select the upper left pixel.
EngineCapability
List of supported engine capability flows.
Definition: NvInferRuntime.h:106
ScaleMode
Controls how shift, scale and power are applied in a Scale layer.
Definition: NvInfer.h:2118
@ kUNIFORM
Identical coefficients across all elements of the tensor.
@ kCHANNEL
Per-channel coefficients.
uint32_t QuantizationFlags
Represents one or more QuantizationFlag values using binary OR operations.
Definition: NvInfer.h:7705
constexpr int32_t EnumMax< RNNDirection >() noexcept
Maximum number of elements in RNNDirection enum.
Definition: NvInfer.h:3184
constexpr int32_t EnumMax< BuilderFlag >() noexcept
Maximum number of builder flags in BuilderFlag enum.
Definition: NvInfer.h:7799
constexpr int32_t EnumMax< LayerType >() noexcept
Maximum number of elements in LayerType enum.
Definition: NvInfer.h:136
constexpr int32_t EnumMax< RNNGateType >() noexcept
Maximum number of elements in RNNGateType enum.
Definition: NvInfer.h:3236
constexpr int32_t EnumMax< CalibrationAlgoType >() noexcept
Maximum number of elements in CalibrationAlgoType enum.
Definition: NvInfer.h:7268
UnaryOperation
Enumerates the unary operations that may be performed by a Unary layer.
Definition: NvInfer.h:3559
@ kCOSH
Hyperbolic cosine.
@ kACOSH
Inverse hyperbolic cosine.
@ kERF
Gauss error function.
@ kROUND
Round to nearest even for float datatype.
@ 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
Maximum number of elements in ReduceOperation enum.
Definition: NvInfer.h:3656
constexpr int32_t EnumMax< TripLimit >() noexcept
Maximum number of elements in TripLimit enum.
Definition: NvInfer.h:4979
constexpr int32_t EnumMax< RNNInputMode >() noexcept
Maximum number of elements in RNNInputMode enum.
Definition: NvInfer.h:3212
RNNInputMode
Enumerates the RNN input modes that may occur with an RNN layer.
Definition: NvInfer.h:3205
@ kSKIP
No operation is performed on the first recurrent layer.
@ kLINEAR
Perform the normal matrix multiplication in the first recurrent layer.
ActivationType
Enumerates the types of activation to perform in an activation layer.
Definition: NvInfer.h:155
@ 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))
@ 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:5458
@ kLINSPACE
Generate evenly spaced numbers over a specified interval.
@ kRANDOM_UNIFORM
Generate a tensor with random values drawn from a uniform distribution.
ResizeRoundMode
The rounding mode for nearest neighbor resize.
Definition: NvInfer.h:4660
@ kHALF_DOWN
Round half down.
RNNGateType
Identifies an individual gate within an RNN cell.
Definition: NvInfer.h:3225
@ kUPDATE
Update gate (z).
@ kHIDDEN
Hidden gate (h).
@ kFORGET
Forget gate (f).
@ kOUTPUT
Output gate (o).
PaddingMode
Enumerates the modes of padding to perform in convolution, deconvolution and pooling layer,...
Definition: NvInfer.h:977
@ kSAME_LOWER
Use SAME padding, with prePadding >= postPadding.
@ kEXPLICIT_ROUND_DOWN
Use explicit padding, rounding output size down.
@ kCAFFE_ROUND_DOWN
Use CAFFE padding, rounding output size down, uses prePadding value.
@ kEXPLICIT_ROUND_UP
Use explicit padding, rounding output size up.
@ kCAFFE_ROUND_UP
Use CAFFE padding, rounding output size up, uses prePadding value.
@ kSAME_UPPER
Use SAME padding, with prePadding <= postPadding.
TripLimit
Enum that describes kinds of trip limits.
Definition: NvInfer.h:4971
@ kWHILE
Tensor is a scalar of type kBOOL. Loop terminates when value is false.
@ kCOUNT
Tensor is scalar of type kINT32 that contains the trip count.
uint32_t NetworkDefinitionCreationFlags
Represents one or more NetworkDefinitionCreationFlag flags using binary OR operations....
Definition: NvInfer.h:8477
constexpr int32_t EnumMax< GatherMode >() noexcept
Maximum number of elements in GatherMode enum.
Definition: NvInfer.h:2908
DataType
The type of weights and tensors.
Definition: NvInferRuntimeCommon.h:151
uint32_t BuilderFlags
Represents one or more QuantizationFlag values using binary OR operations, e.g., 1U << BuilderFlag::k...
Definition: NvInfer.h:7735
DeviceType
The device that this layer/network will execute on.
Definition: NvInferRuntime.h:633
constexpr int32_t EnumMax< ScaleMode >() noexcept
Maximum number of elements in ScaleMode enum.
Definition: NvInfer.h:2126
CalibrationAlgoType
Version of calibration algorithm to use.
Definition: NvInfer.h:7259
LayerType
The type values of layer classes.
Definition: NvInfer.h:90
@ kRAGGED_SOFTMAX
Ragged softmax layer.
@ kDECONVOLUTION
Deconvolution layer.
@ kASSERTION
Assertion layer.
@ kMATRIX_MULTIPLY
Matrix multiply layer.
@ kCONDITION
Condition layer.
@ kCONDITIONAL_INPUT
Conditional Input layer.
@ kIDENTITY
Identity layer.
@ kQUANTIZE
Quantize layer.
@ kCONVOLUTION
Convolution layer.
@ kPARAMETRIC_RELU
Parametric ReLU layer.
@ kCONCATENATION
Concatenation layer.
@ kFULLY_CONNECTED
Fully connected layer.
@ kRECURRENCE
Loop Recurrence layer.
@ kDEQUANTIZE
Dequantize layer.
@ kITERATOR
Loop Iterator layer.
@ kTRIP_LIMIT
Loop Trip limit 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.
SliceMode
Controls how ISliceLayer handles out of bounds coordinates.
Definition: NvInfer.h:4047
@ kCLAMP
Out of bounds indices are clamped to bounds.
@ kWRAP
Coordinates wrap around periodically.
constexpr int32_t EnumMax< QuantizationFlag >() noexcept
Maximum number of quantization flags in QuantizationFlag enum.
Definition: NvInfer.h:7724
GatherMode
Control form of IGatherLayer.
Definition: NvInfer.h:2900
@ 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:147
ProfilingVerbosity
List of verbosity levels of layer information exposed in NVTX annotations and in IEngineInspector.
Definition: NvInferRuntime.h:1287
ResizeMode
Enumerates various modes of resize in the resize layer. Resize mode set using setResizeMode().
Definition: NvInfer.h:4561
@ kNEAREST
ND (0 < N <= 8) nearest neighbor resizing.
constexpr int32_t EnumMax< SliceMode >() noexcept
Maximum number of elements in SliceMode enum.
Definition: NvInfer.h:4059
NetworkDefinitionCreationFlag
List of immutable network properties expressed at network creation time. NetworkDefinitionCreationFla...
Definition: NvInfer.h:8489
@ kEXPLICIT_BATCH
Mark the network to be an explicit batch network.
ElementWiseOperation
Enumerates the binary operations that may be performed by an ElementWise layer.
Definition: NvInfer.h:2814
@ kSUB
Substract 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:7715
@ kCALIBRATE_BEFORE_FUSION
RNNDirection
Enumerates the RNN direction that may be performed by an RNN layer.
Definition: NvInfer.h:3177
@ kBIDIRECTION
Network iterates from first to last and vice versa and outputs concatenated.
@ kUNIDIRECTION
Network iterations from first input to last input.
BuilderFlag
List of valid modes that the builder can enable when creating an engine from a network definition.
Definition: NvInfer.h:7745
@ 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.
@ kFP16
Enable FP16 layer selection, with FP32 fallback.
@ kINT8
Enable Int8 layer selection, with FP32 fallback with FP16 fallback if kFP16 also specified.
@ kPREFER_PRECISION_CONSTRAINTS
@ kDISABLE_TIMING_CACHE
Disable reuse of timing information across identical layers.
@ kREFIT
Enable building a refittable engine.
@ kOBEY_PRECISION_CONSTRAINTS
Require that layers execute in specified precisions. Build fails otherwise.
@ kREJECT_EMPTY_ALGORITHMS
Fail if IAlgorithmSelector::selectAlgorithms returns an empty set of algorithms.
constexpr int32_t EnumMax< TopKOperation >() noexcept
Maximum number of elements in TopKOperation enum.
Definition: NvInfer.h:4267
TensorFormat
Format of the input/output tensors.
Definition: NvInferRuntimeCommon.h:221
TopKOperation
Enumerates the operations that may be performed by a TopK layer.
Definition: NvInfer.h:4260
ReduceOperation
Enumerates the reduce operations that may be performed by a Reduce layer.
Definition: NvInfer.h:3646
constexpr int32_t EnumMax< LoopOutput >() noexcept
Maximum number of elements in LoopOutput enum.
Definition: NvInfer.h:4964
RNNOperation
Enumerates the RNN operations that may be performed by an RNN layer.
Definition: NvInfer.h:3155
@ kGRU
Three-gate network consisting of Gated Recurrent Units.
@ kLSTM
Four-gate LSTM network w/o peephole connections.
constexpr int32_t EnumMax< NetworkDefinitionCreationFlag >() noexcept
Maximum number of elements in NetworkDefinitionCreationFlag enum.
Definition: NvInfer.h:8513
ScatterMode
Control form of IScatterLayer.
Definition: NvInfer.h:5896
MatrixOperation
Enumerates the operations that may be performed on a tensor by IMatrixMultiplyLayer before multiplica...
Definition: NvInfer.h:4356
@ kTRANSPOSE
Like kNONE, but transpose the matrix dimensions.
ResizeCoordinateTransformation
The resize coordinate transformation function.
Definition: NvInfer.h:4584
constexpr int32_t EnumMax< UnaryOperation >() noexcept
Maximum number of elements in UnaryOperation enum.
Definition: NvInfer.h:3587
LoopOutput
Enum that describes kinds of loop outputs.
Definition: NvInfer.h:4951
@ 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< MatrixOperation >() noexcept
Maximum number of elements in MatrixOperation enum.
Definition: NvInfer.h:4379
constexpr int32_t EnumMax< RNNOperation >() noexcept
Maximum number of elements in RNNOperation enum.
Definition: NvInfer.h:3164
PoolingType
The type of pooling to perform in a pooling layer.
Definition: NvInfer.h:1652
constexpr int32_t EnumMax< FillOperation >() noexcept
Maximum number of elements in FillOperation enum.
Definition: NvInfer.h:5465
TensorLocation
The location for tensor data storage, device or host.
Definition: NvInferRuntime.h:247
OptProfileSelector
When setting or querying optimization profile parameters (such as shape tensor inputs or dynamic dime...
Definition: NvInferRuntime.h:1058
constexpr int32_t EnumMax< ScatterMode >() noexcept
Maximum number of elements in ScatterMode enum.
Definition: NvInfer.h:5903
Definition: NvInfer.h:3859
Declaration of EnumMaxImpl struct to store maximum number of elements in an enumeration type.
Definition: NvInferRuntimeCommon.h:136