184 static constexpr int32_t kVALUE = 12;
220 mImpl->setName(name);
232 return mImpl->getName();
251 mImpl->setDimensions(dimensions);
264 return mImpl->getDimensions();
279 mImpl->setType(type);
291 return mImpl->getType();
306 return mImpl->setDynamicRange(min, max);
314 return mImpl->isNetworkInput();
322 return mImpl->isNetworkOutput();
344 mImpl->setBroadcastAcrossBatch(broadcastAcrossBatch);
360 return mImpl->getBroadcastAcrossBatch();
370 return mImpl->getLocation();
385 mImpl->setLocation(location);
395 return mImpl->dynamicRangeIsSet();
403 mImpl->resetDynamicRange();
413 return mImpl->getDynamicRangeMin();
423 return mImpl->getDynamicRangeMax();
442 mImpl->setAllowedFormats(formats);
455 return mImpl->getAllowedFormats();
490 return mImpl->isShapeTensor();
513 return mImpl->isExecutionTensor();
538 return mLayer->getType();
550 mLayer->setName(name);
561 return mLayer->getName();
569 return mLayer->getNbInputs();
582 return mLayer->getInput(index);
590 return mLayer->getNbOutputs();
601 return mLayer->getOutput(index);
618 return mLayer->setInput(index, tensor);
646 mLayer->setPrecision(dataType);
658 return mLayer->getPrecision();
670 return mLayer->precisionIsSet();
680 mLayer->resetPrecision();
718 mLayer->setOutputType(index, dataType);
732 return mLayer->getOutputType(index);
745 return mLayer->outputTypeIsSet(index);
757 return mLayer->resetOutputType(index);
762 apiv::VLayer* mLayer;
1007 static constexpr int32_t kVALUE = 6;
1037 mImpl->setKernelSize(kernelSize);
1049 return mImpl->getKernelSize();
1061 mImpl->setNbOutputMaps(nbOutputMaps);
1071 return mImpl->getNbOutputMaps();
1087 mImpl->setStride(stride);
1097 return mImpl->getStride();
1117 return mImpl->setPadding(padding);
1129 return mImpl->getPadding();
1149 mImpl->setNbGroups(nbGroups);
1159 return mImpl->getNbGroups();
1173 mImpl->setKernelWeights(weights);
1183 return mImpl->getKernelWeights();
1198 mImpl->setBiasWeights(weights);
1208 return mImpl->getBiasWeights();
1224 return mImpl->setDilation(dilation);
1236 return mImpl->getDilation();
1253 mImpl->setPrePadding(padding);
1263 return mImpl->getPrePadding();
1280 mImpl->setPostPadding(padding);
1290 return mImpl->getPostPadding();
1304 mImpl->setPaddingMode(paddingMode);
1316 return mImpl->getPaddingMode();
1329 mImpl->setKernelSizeNd(kernelSize);
1339 return mImpl->getKernelSizeNd();
1354 mImpl->setStrideNd(stride);
1364 return mImpl->getStrideNd();
1382 mImpl->setPaddingNd(padding);
1394 return mImpl->getPaddingNd();
1408 mImpl->setDilationNd(dilation);
1418 return mImpl->getDilationNd();
1492 mImpl->setNbOutputChannels(nbOutputs);
1502 return mImpl->getNbOutputChannels();
1512 mImpl->setKernelWeights(weights);
1522 return mImpl->getKernelWeights();
1534 mImpl->setBiasWeights(weights);
1544 return mImpl->getBiasWeights();
1600 mImpl->setActivationType(type);
1610 return mImpl->getActivationType();
1625 mImpl->setAlpha(alpha);
1639 mImpl->setBeta(beta);
1648 return mImpl->getAlpha();
1657 return mImpl->getBeta();
1687 static constexpr int32_t kVALUE = 3;
1714 mImpl->setPoolingType(type);
1724 return mImpl->getPoolingType();
1738 mImpl->setWindowSize(windowSize);
1750 return mImpl->getWindowSize();
1766 mImpl->setStride(stride);
1778 return mImpl->getStride();
1794 mImpl->setPadding(padding);
1808 return mImpl->getPadding();
1823 mImpl->setBlendFactor(blendFactor);
1836 return mImpl->getBlendFactor();
1853 mImpl->setAverageCountExcludesPadding(exclusive);
1864 return mImpl->getAverageCountExcludesPadding();
1882 mImpl->setPrePadding(padding);
1892 return mImpl->getPrePadding();
1910 mImpl->setPostPadding(padding);
1920 return mImpl->getPostPadding();
1933 mImpl->setPaddingMode(paddingMode);
1944 return mImpl->getPaddingMode();
1957 mImpl->setWindowSizeNd(windowSize);
1967 return mImpl->getWindowSizeNd();
1982 mImpl->setStrideNd(stride);
1992 return mImpl->getStrideNd();
2011 mImpl->setPaddingNd(padding);
2023 return mImpl->getPaddingNd();
2054 mImpl->setWindowSize(windowSize);
2064 return mImpl->getWindowSize();
2075 mImpl->setAlpha(alpha);
2085 return mImpl->getAlpha();
2096 mImpl->setBeta(beta);
2106 return mImpl->getBeta();
2127 return mImpl->getK();
2194 mImpl->setMode(mode);
2204 return mImpl->getMode();
2214 mImpl->setShift(shift);
2224 return mImpl->getShift();
2234 mImpl->setScale(scale);
2244 return mImpl->getScale();
2254 mImpl->setPower(power);
2264 return mImpl->getPower();
2279 return mImpl->getChannelAxis();
2300 mImpl->setChannelAxis(channelAxis);
2361 mImpl->setAxes(axes);
2371 return mImpl->getAxes();
2406 mImpl->setAxis(axis);
2416 return mImpl->getAxis();
2447 mImpl->setKernelSize(kernelSize);
2459 return mImpl->getKernelSize();
2471 mImpl->setNbOutputMaps(nbOutputMaps);
2481 return mImpl->getNbOutputMaps();
2497 mImpl->setStride(stride);
2509 return mImpl->getStride();
2529 mImpl->setPadding(padding);
2543 return mImpl->getPadding();
2563 mImpl->setNbGroups(nbGroups);
2573 return mImpl->getNbGroups();
2587 mImpl->setKernelWeights(weights);
2597 return mImpl->getKernelWeights();
2612 mImpl->setBiasWeights(weights);
2622 return mImpl->getBiasWeights();
2640 mImpl->setPrePadding(padding);
2650 return mImpl->getPrePadding();
2668 mImpl->setPostPadding(padding);
2678 return mImpl->getPostPadding();
2692 mImpl->setPaddingMode(paddingMode);
2704 return mImpl->getPaddingMode();
2719 mImpl->setKernelSizeNd(kernelSize);
2729 return mImpl->getKernelSizeNd();
2746 mImpl->setStrideNd(stride);
2756 return mImpl->getStrideNd();
2774 mImpl->setPaddingNd(padding);
2786 return mImpl->getPaddingNd();
2820 mImpl->setDilationNd(dilation);
2830 return mImpl->getDilationNd();
2879 static constexpr int32_t kVALUE = 14;
2916 return mImpl->setOperation(op);
2928 return mImpl->getOperation();
3051 mImpl->setGatherAxis(axis);
3062 return mImpl->getGatherAxis();
3083 mImpl->setNbElementWiseDims(elementWiseDims);
3093 return mImpl->getNbElementWiseDims();
3103 mImpl->setMode(mode);
3113 return mImpl->getMode();
3320 return mImpl->getLayerCount();
3324 return mImpl->getHiddenSize();
3328 return mImpl->getMaxSeqLength();
3332 return mImpl->getDataLength();
3351 return mImpl->setSequenceLengths(seqLengths);
3363 return mImpl->getSequenceLengths();
3373 mImpl->setOperation(op);
3383 return mImpl->getOperation();
3393 mImpl->setInputMode(op);
3403 return mImpl->getInputMode();
3418 mImpl->setDirection(op);
3428 return mImpl->getDirection();
3487 mImpl->setWeightsForGate(layerIndex, gate, isW, weights);
3497 return mImpl->getWeightsForGate(layerIndex, gate, isW);
3522 mImpl->setBiasForGate(layerIndex, gate, isW, bias);
3532 return mImpl->getBiasForGate(layerIndex, gate, isW);
3549 mImpl->setHiddenState(hidden);
3559 return mImpl->getHiddenState();
3578 mImpl->setCellState(cell);
3588 return mImpl->getCellState();
3615 return mImpl->getPlugin();
3693 mImpl->setOperation(op);
3703 return mImpl->getOperation();
3766 mImpl->setOperation(op);
3776 return mImpl->getOperation();
3786 mImpl->setReduceAxes(reduceAxes);
3796 return mImpl->getReduceAxes();
3806 mImpl->setKeepDimensions(keepDimensions);
3816 return mImpl->getKeepDimensions();
3848 mImpl->setPrePadding(padding);
3860 return mImpl->getPrePadding();
3874 mImpl->setPostPadding(padding);
3886 return mImpl->getPostPadding();
3900 mImpl->setPrePaddingNd(padding);
3912 return mImpl->getPrePaddingNd();
3926 mImpl->setPostPaddingNd(padding);
3938 return mImpl->getPostPaddingNd();
3983 mImpl->setFirstTranspose(permutation);
3995 return mImpl->getFirstTranspose();
4020 mImpl->setReshapeDimensions(dimensions);
4033 return mImpl->getReshapeDimensions();
4080 mImpl->setSecondTranspose(permutation);
4092 return mImpl->getSecondTranspose();
4108 return mImpl->setZeroIsPlaceholder(zeroIsPlaceholder);
4121 return mImpl->getZeroIsPlaceholder();
4205 mImpl->setStart(start);
4220 return mImpl->getStart();
4234 return mImpl->setSize(size);
4249 return mImpl->getSize();
4263 mImpl->setStride(stride);
4278 return mImpl->getStride();
4288 mImpl->setMode(mode);
4298 return mImpl->getMode();
4387 mImpl->setOperation(op);
4397 return mImpl->getOperation();
4419 return mImpl->getK();
4429 mImpl->setReduceAxes(reduceAxes);
4439 return mImpl->getReduceAxes();
4523 mImpl->setOperation(index, op);
4535 return mImpl->getOperation(index);
4616 mImpl->setWeights(weights);
4626 return mImpl->getWeights();
4638 mImpl->setDimensions(dimensions);
4650 return mImpl->getDimensions();
4695 static constexpr int32_t kVALUE = 2;
4749 static constexpr int32_t kVALUE = 3;
4779 static constexpr int32_t kVALUE = 2;
4815 static constexpr int32_t kVALUE = 4;
4878 return mImpl->setOutputDimensions(dimensions);
4888 return mImpl->getOutputDimensions();
4916 void setScales(
const float* scales, int32_t nbScales)
noexcept
4918 mImpl->setScales(scales, nbScales);
4935 int32_t
getScales(int32_t size,
float* scales)
const noexcept
4937 return mImpl->getScales(size, scales);
4949 mImpl->setResizeMode(resizeMode);
4959 return mImpl->getResizeMode();
4975 mImpl->setAlignCorners(alignCorners);
4987 return mImpl->getAlignCorners();
5022 mImpl->setCoordinateTransformation(coordTransform);
5032 return mImpl->getCoordinateTransformation();
5047 mImpl->setSelectorForSinglePixel(selector);
5057 return mImpl->getSelectorForSinglePixel();
5071 mImpl->setNearestRounding(value);
5081 return mImpl->getNearestRounding();
5140 return mBoundary->getLoop();
5145 apiv::VLoopBoundaryLayer* mBoundary;
5159 return mBoundary->getConditional();
5164 apiv::VConditionalBoundaryLayer* mBoundary;
5237 return mImpl->setCondition(condition);
5253 return mImpl->addOutput(trueSubgraphOutput, falseSubgraphOutput);
5265 return mImpl->addInput(input);
5278 mImpl->setName(name);
5288 return mImpl->getName();
5347 return mImpl->getLoopOutput();
5364 mImpl->setAxis(axis);
5370 return mImpl->getAxis();
5405 return mImpl->getTripLimit();
5419 mImpl->setAxis(axis);
5425 return mImpl->getAxis();
5435 mImpl->setReverse(reverse);
5441 return mImpl->getReverse();
5465 return mImpl->addRecurrence(initialValue);
5486 return mImpl->addTripLimit(tensor, limit);
5499 return mImpl->addIterator(tensor, axis, reverse);
5511 return mImpl->addLoopOutput(tensor, outputKind, axis);
5524 mImpl->setName(name);
5534 return mImpl->getName();
5579 mImpl->setMessage(message);
5589 return mImpl->getMessage();
5661 mImpl->setDimensions(dimensions);
5676 return mImpl->getDimensions();
5686 mImpl->setOperation(op);
5696 return mImpl->getOperation();
5714 mImpl->setAlpha(alpha);
5729 return mImpl->getAlpha();
5747 mImpl->setBeta(beta);
5762 return mImpl->getBeta();
5868 return mImpl->getAxis();
5879 mImpl->setAxis(axis);
5955 return mImpl->getAxis();
5966 mImpl->setAxis(axis);
6023 return mImpl->setEquation(equation);
6033 return mImpl->getEquation();
6129 mImpl->setMode(mode);
6139 return mImpl->getMode();
6149 mImpl->setAxis(axis);
6157 return mImpl->getAxis();
6226 return mImpl->addInput(name, type, dimensions);
6240 mImpl->markOutput(tensor);
6264 return mImpl->addConvolution(input, nbOutputMaps, kernelSize, kernelWeights, biasWeights);
6287 return mImpl->addFullyConnected(input, nbOutputs, kernelWeights, biasWeights);
6306 return mImpl->addActivation(input, type);
6325 return mImpl->addPooling(input, type, windowSize);
6344 return mImpl->addLRN(input, window, alpha, beta, k);
6371 return mImpl->addScale(input, mode, shift, scale, power);
6384 return mImpl->addSoftMax(input);
6401 return mImpl->addConcatenation(inputs, nbInputs);
6425 return mImpl->addDeconvolution(input, nbOutputMaps, kernelSize, kernelWeights, biasWeights);
6452 return mImpl->addElementWise(input1, input2, op);
6474 return mImpl->addUnary(input, operation);
6491 return mImpl->addPadding(input, prePadding, postPadding);
6505 return mImpl->addShuffle(input);
6517 return mImpl->getNbLayers();
6531 return mImpl->getLayer(index);
6543 return mImpl->getNbInputs();
6559 return mImpl->getInput(index);
6573 return mImpl->getNbOutputs();
6589 return mImpl->getOutput(index);
6629 return mImpl->addReduce(input, operation, reduceAxes, keepDimensions);
6662 return mImpl->addTopK(input, op, k, reduceAxes);
6678 return mImpl->addGather(data, indices, axis);
6694 return mImpl->addGatherV2(data, indices, mode);
6712 return mImpl->addRaggedSoftMax(input, bounds);
6734 return mImpl->addMatrixMultiply(input0, op0, input1, op1);
6759 return mImpl->addConstant(dimensions, weights);
6827 ITensor& input, int32_t layerCount, int32_t hiddenSize, int32_t maxSeqLen,
RNNOperation op)
noexcept
6829 return mImpl->addRNNv2(input, layerCount, hiddenSize, maxSeqLen, op);
6843 return mImpl->addIdentity(input);
6858 mImpl->removeTensor(tensor);
6870 mImpl->unmarkOutput(tensor);
6889 return mImpl->addPluginV2(inputs, nbInputs, plugin);
6908 return mImpl->addSlice(input, start, size, stride);
6930 mImpl->setName(name);
6944 return mImpl->getName();
6962 return mImpl->addShape(input);
6981 return mImpl->hasImplicitBatchDimension();
6999 return mImpl->markOutputForShapes(tensor);
7011 return mImpl->unmarkOutputForShapes(tensor);
7029 return mImpl->addParametricReLU(input, slope);
7052 return mImpl->addConvolutionNd(input, nbOutputMaps, kernelSize, kernelWeights, biasWeights);
7071 return mImpl->addPoolingNd(input, type, windowSize);
7094 return mImpl->addDeconvolutionNd(input, nbOutputMaps, kernelSize, kernelWeights, biasWeights);
7130 return mImpl->addScaleNd(input, mode, shift, scale, power, channelAxis);
7145 return mImpl->addResize(input);
7162 return mImpl->hasExplicitPrecision();
7178 return mImpl->addLoop();
7218 return mImpl->addSelect(condition, thenInput, elseInput);
7235 return mImpl->addAssertion(condition, message);
7253 return mImpl->addFill(dimensions, op);
7270 return mImpl->addPaddingNd(input, prePadding, postPadding);
7290 return mImpl->setWeightsName(weights, name);
7309 mImpl->setErrorRecorder(recorder);
7324 return mImpl->getErrorRecorder();
7343 return mImpl->addDequantize(input, scale);
7363 return mImpl->addScatter(data, indices, updates, mode);
7382 return mImpl->addQuantize(input, scale);
7397 return mImpl->addIfConditional();
7411 return mImpl->addEinsum(inputs, nbInputs, equation);
7476 virtual
bool getBatch(
void* bindings[], const
char* names[], int32_t nbBindings) noexcept = 0;
7492 virtual const
void* readCalibrationCache(std::
size_t& length) noexcept = 0;
7502 virtual
void writeCalibrationCache(const
void* ptr, std::
size_t length) noexcept = 0;
7596 virtual
double getRegressionCutoff() const noexcept = 0;
7610 virtual const
void* readHistogramCache(std::
size_t& length) noexcept = 0;
7620 virtual
void writeHistogramCache(const
void* ptr, std::
size_t length) noexcept = 0;
7643 return mImpl->getTensorFormat();
7651 return mImpl->getDataType();
7659 return mImpl->getStrides();
7686 return mImpl->getImplementation();
7694 return mImpl->getTactic();
7719 return mImpl->getName();
7730 return mImpl->getDimensions(index, select);
7738 return mImpl->getNbInputs();
7746 return mImpl->getNbOutputs();
7778 return mImpl->getAlgorithmIOInfo(index);
7786 return mImpl->getAlgorithmVariant();
7794 return mImpl->getTimingMSec();
7802 return mImpl->getWorkspaceSize();
7815 return mImpl->getAlgorithmIOInfoByIndex(index);
7849 int32_t nbChoices, int32_t* selection)
noexcept
7862 int32_t nbAlgorithms)
noexcept
8008 return mImpl->serialize();
8032 return mImpl->combine(inputCache, ignoreMismatch);
8042 return mImpl->reset();
8123 virtual
void setMinTimingIterations(int32_t minTiming) noexcept
8125 mImpl->setMinTimingIterations(minTiming);
8137 return mImpl->getMinTimingIterations();
8150 mImpl->setAvgTimingIterations(avgTiming);
8162 return mImpl->getAvgTimingIterations();
8175 mImpl->setEngineCapability(capability);
8187 return mImpl->getEngineCapability();
8197 mImpl->setInt8Calibrator(calibrator);
8205 return mImpl->getInt8Calibrator();
8220 mImpl->setMaxWorkspaceSize(workspaceSize);
8237 return mImpl->getMaxWorkspaceSize();
8254 mImpl->setFlags(builderFlags);
8266 return mImpl->getFlags();
8278 mImpl->clearFlag(builderFlag);
8290 mImpl->setFlag(builderFlag);
8302 return mImpl->getFlag(builderFlag);
8317 mImpl->setDeviceType(layer, deviceType);
8326 return mImpl->getDeviceType(layer);
8336 return mImpl->isDeviceTypeSet(layer);
8346 mImpl->resetDeviceType(layer);
8355 return mImpl->canRunOnDLA(layer);
8370 mImpl->setDLACore(dlaCore);
8381 return mImpl->getDLACore();
8391 mImpl->setDefaultDeviceType(deviceType);
8401 return mImpl->getDefaultDeviceType();
8437 return mImpl->setProfileStream(stream);
8449 return mImpl->getProfileStream();
8465 return mImpl->addOptimizationProfile(profile);
8478 return mImpl->getNbOptimizationProfiles();
8490 mImpl->setProfilingVerbosity(verbosity);
8503 return mImpl->getProfilingVerbosity();
8512 mImpl->setAlgorithmSelector(selector);
8520 return mImpl->getAlgorithmSelector();
8535 return mImpl->setCalibrationProfile(profile);
8545 return mImpl->getCalibrationProfile();
8562 mImpl->setQuantizationFlags(flags);
8574 return mImpl->getQuantizationFlags();
8586 mImpl->clearQuantizationFlag(flag);
8598 mImpl->setQuantizationFlag(flag);
8610 return mImpl->getQuantizationFlag(flag);
8635 return mImpl->setTacticSources(tacticSources);
8650 return mImpl->getTacticSources();
8669 return mImpl->createTimingCache(blob, size);
8692 return mImpl->setTimingCache(cache, ignoreMismatch);
8702 return mImpl->getTimingCache();
8734 mImpl->setMemoryPoolLimit(pool, poolSize);
8753 return mImpl->getMemoryPoolLimit(pool);
8832 void setMaxBatchSize(int32_t batchSize) noexcept
8834 mImpl->setMaxBatchSize(batchSize);
8847 return mImpl->getMaxBatchSize();
8855 return mImpl->platformHasFastFp16();
8863 return mImpl->platformHasFastInt8();
8887 return mImpl->getMaxDLABatchSize();
8895 return mImpl->getNbDLACores();
8911 mImpl->setGpuAllocator(allocator);
8921 return mImpl->createBuilderConfig();
8937 return mImpl->buildEngineWithConfig(network, config);
8953 return mImpl->createNetworkV2(flags);
8967 return mImpl->createOptimizationProfile();
8986 mImpl->setErrorRecorder(recorder);
9001 return mImpl->getErrorRecorder();
9017 return mImpl->platformHasTf32();
9036 return mImpl->buildSerializedNetwork(network, config);
9060 return mImpl->isNetworkSupported(network, config);
9070 return mImpl->getLogger();
9084 return mImpl->setMaxThreads(maxThreads);
9098 return mImpl->getMaxThreads();
9111extern "C" TENSORRTAPI void* createInferBuilder_INTERNAL(
void* logger, int32_t version)
noexcept;
9125inline IBuilder* createInferBuilder(ILogger& logger)
noexcept
#define TENSORRTAPI
Definition: NvInferRuntimeCommon.h:91
#define NV_TENSORRT_VERSION
Definition: NvInferRuntimeCommon.h:110
#define TRT_DEPRECATED
Definition: NvInferRuntimeCommon.h:77
#define TRT_DEPRECATED_ENUM
Definition: NvInferRuntimeCommon.h:78
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:1589
void setBeta(float beta) noexcept
Set the beta parameter (must be finite).
Definition: NvInfer.h:1637
void setActivationType(ActivationType type) noexcept
Set the type of activation to be performed.
Definition: NvInfer.h:1598
ActivationType getActivationType() const noexcept
Get the type of activation to be performed.
Definition: NvInfer.h:1608
float getAlpha() const noexcept
Get the alpha parameter.
Definition: NvInfer.h:1646
virtual ~IActivationLayer() noexcept=default
float getBeta() const noexcept
Get the beta parameter.
Definition: NvInfer.h:1655
void setAlpha(float alpha) noexcept
Set the alpha parameter (must be finite).
Definition: NvInfer.h:1623
Describes the context and requirements, that could be fulfilled by one or more instances of IAlgorith...
Definition: NvInfer.h:7711
const char * getName() const noexcept
Return name of the algorithm node. This is a unique identifier for the IAlgorithmContext.
Definition: NvInfer.h:7717
int32_t getNbOutputs() const noexcept
Return number of outputs of the algorithm.
Definition: NvInfer.h:7744
int32_t getNbInputs() const noexcept
Return number of inputs of the algorithm.
Definition: NvInfer.h:7736
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:7728
Describes a variation of execution of a layer. An algorithm is represented by IAlgorithmVariant and t...
Definition: NvInfer.h:7764
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:7800
float getTimingMSec() const noexcept
The time in milliseconds to execute the algorithm.
Definition: NvInfer.h:7792
virtual ~IAlgorithm() noexcept=default
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:7776
const IAlgorithmVariant & getAlgorithmVariant() const noexcept
Returns the algorithm variant.
Definition: NvInfer.h:7784
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:7813
Carries information about input or output of the algorithm. IAlgorithmIOInfo for all the input and ou...
Definition: NvInfer.h:7636
virtual ~IAlgorithmIOInfo() noexcept=default
Dims getStrides() const noexcept
Return strides of the input/output tensor of algorithm.
Definition: NvInfer.h:7657
DataType getDataType() const noexcept
Return DataType of the input/output of algorithm.
Definition: NvInfer.h:7649
TensorFormat getTensorFormat() const noexcept
Return TensorFormat of the input/output of algorithm.
Definition: NvInfer.h:7641
Interface implemented by application for selecting and reporting algorithms of a layer provided by th...
Definition: NvInfer.h:7832
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.
virtual ~IAlgorithmSelector() noexcept=default
provides a unique 128-bit identifier, which along with the input and output information denotes the v...
Definition: NvInfer.h:7679
virtual ~IAlgorithmVariant() noexcept=default
int64_t getTactic() const noexcept
Return tactic of the algorithm.
Definition: NvInfer.h:7692
int64_t getImplementation() const noexcept
Return implementation of the algorithm.
Definition: NvInfer.h:7684
An assertion layer in a network.
Definition: NvInfer.h:5567
void setMessage(const char *message) noexcept
Set the message to print if the assertion fails.
Definition: NvInfer.h:5577
virtual ~IAssertionLayer() noexcept=default
const char * getMessage() const noexcept
Return the assertion message.
Definition: NvInfer.h:5587
Holds properties for configuring a builder to produce an engine.
Definition: NvInfer.h:8109
DeviceType getDeviceType(const ILayer *layer) const noexcept
Get the device that this layer executes on.
Definition: NvInfer.h:8324
void setMemoryPoolLimit(MemoryPoolType pool, std::size_t poolSize) noexcept
Set the memory size for the memory pool.
Definition: NvInfer.h:8732
void setQuantizationFlag(QuantizationFlag flag) noexcept
Set a single quantization flag.
Definition: NvInfer.h:8596
const IOptimizationProfile * getCalibrationProfile() noexcept
Get the current calibration profile.
Definition: NvInfer.h:8543
void setInt8Calibrator(IInt8Calibrator *calibrator) noexcept
Set Int8 Calibration interface.
Definition: NvInfer.h:8195
void clearQuantizationFlag(QuantizationFlag flag) noexcept
clear a quantization flag.
Definition: NvInfer.h:8584
void setDeviceType(const ILayer *layer, DeviceType deviceType) noexcept
Set the device that this layer must execute on.
Definition: NvInfer.h:8315
bool setTacticSources(TacticSources tacticSources) noexcept
Set tactic sources.
Definition: NvInfer.h:8633
bool getQuantizationFlag(QuantizationFlag flag) const noexcept
Returns true if the quantization flag is set.
Definition: NvInfer.h:8608
std::size_t getMemoryPoolLimit(MemoryPoolType pool) const noexcept
Get the memory size limit of the memory pool.
Definition: NvInfer.h:8751
int32_t getDLACore() const noexcept
Get the DLA core that the engine executes on.
Definition: NvInfer.h:8379
void setEngineCapability(EngineCapability capability) noexcept
Configure the builder to target specified EngineCapability flow.
Definition: NvInfer.h:8173
bool getFlag(BuilderFlag builderFlag) const noexcept
Returns true if the build mode flag is set.
Definition: NvInfer.h:8300
void setQuantizationFlags(QuantizationFlags flags) noexcept
Set the quantization flags.
Definition: NvInfer.h:8560
virtual void setAvgTimingIterations(int32_t avgTiming) noexcept
Set the number of averaging iterations used when timing layers.
Definition: NvInfer.h:8148
void setProfilingVerbosity(ProfilingVerbosity verbosity) noexcept
Set verbosity level of layer information exposed in NVTX annotations and IEngineInspector.
Definition: NvInfer.h:8488
nvinfer1::ITimingCache * createTimingCache(const void *blob, std::size_t size) const noexcept
Create timing cache.
Definition: NvInfer.h:8667
IAlgorithmSelector * getAlgorithmSelector() const noexcept
Get Algorithm Selector.
Definition: NvInfer.h:8518
int32_t getNbOptimizationProfiles() const noexcept
Get number of optimization profiles.
Definition: NvInfer.h:8476
QuantizationFlags getQuantizationFlags() const noexcept
Get the quantization flags.
Definition: NvInfer.h:8572
bool setCalibrationProfile(const IOptimizationProfile *profile) noexcept
Add a calibration profile.
Definition: NvInfer.h:8533
virtual int32_t getMinTimingIterations() const noexcept
Query the number of minimization iterations.
Definition: NvInfer.h:8135
void reset() noexcept
Resets the builder configuration to defaults.
Definition: NvInfer.h:8409
TRT_DEPRECATED void destroy() noexcept
Delete this IBuilderConfig.
Definition: NvInfer.h:8423
EngineCapability getEngineCapability() const noexcept
Query EngineCapability flow configured for the builder.
Definition: NvInfer.h:8185
void setAlgorithmSelector(IAlgorithmSelector *selector) noexcept
Set Algorithm Selector.
Definition: NvInfer.h:8510
TRT_DEPRECATED void setMaxWorkspaceSize(std::size_t workspaceSize) noexcept
Set the maximum workspace size.
Definition: NvInfer.h:8218
TRT_DEPRECATED std::size_t getMaxWorkspaceSize() const noexcept
Get the maximum workspace size.
Definition: NvInfer.h:8235
DeviceType getDefaultDeviceType() const noexcept
Get the default DeviceType which was set by setDefaultDeviceType.
Definition: NvInfer.h:8399
BuilderFlags getFlags() const noexcept
Get the build mode flags for this builder config. Defaults to 0.
Definition: NvInfer.h:8264
void setFlags(BuilderFlags builderFlags) noexcept
Set the build mode flags to turn on builder options for this network.
Definition: NvInfer.h:8252
TacticSources getTacticSources() const noexcept
Get tactic sources.
Definition: NvInfer.h:8648
bool setTimingCache(const ITimingCache &cache, bool ignoreMismatch) noexcept
Attach a timing cache to IBuilderConfig.
Definition: NvInfer.h:8690
bool canRunOnDLA(const ILayer *layer) const noexcept
Checks if a layer can run on DLA.
Definition: NvInfer.h:8353
void setDLACore(int32_t dlaCore) noexcept
Sets the DLA core used by the network.
Definition: NvInfer.h:8368
int32_t addOptimizationProfile(const IOptimizationProfile *profile) noexcept
Add an optimization profile.
Definition: NvInfer.h:8463
const nvinfer1::ITimingCache * getTimingCache() const noexcept
Get the pointer to the timing cache from current IBuilderConfig.
Definition: NvInfer.h:8700
bool isDeviceTypeSet(const ILayer *layer) const noexcept
whether the DeviceType has been explicitly set for this layer
Definition: NvInfer.h:8334
void clearFlag(BuilderFlag builderFlag) noexcept
clear a single build mode flag.
Definition: NvInfer.h:8276
apiv::VBuilderConfig * mImpl
Definition: NvInfer.h:8757
int32_t getAvgTimingIterations() const noexcept
Query the number of averaging iterations.
Definition: NvInfer.h:8160
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:8389
void setFlag(BuilderFlag builderFlag) noexcept
Set a single build mode flag.
Definition: NvInfer.h:8288
virtual ~IBuilderConfig() noexcept=default
cudaStream_t getProfileStream() const noexcept
Get the cuda stream that is used to profile this network.
Definition: NvInfer.h:8447
IInt8Calibrator * getInt8Calibrator() const noexcept
Get Int8 Calibration interface.
Definition: NvInfer.h:8203
ProfilingVerbosity getProfilingVerbosity() const noexcept
Get verbosity level of layer information exposed in NVTX annotations and IEngineInspector.
Definition: NvInfer.h:8501
void resetDeviceType(const ILayer *layer) noexcept
reset the DeviceType for this layer
Definition: NvInfer.h:8344
void setProfileStream(const cudaStream_t stream) noexcept
Set the cuda stream that is used to profile this network.
Definition: NvInfer.h:8435
Builds an engine from a network definition.
Definition: NvInfer.h:8820
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:8885
int32_t getNbDLACores() const noexcept
Return the number of DLA engines available to this builder.
Definition: NvInfer.h:8893
IErrorRecorder * getErrorRecorder() const noexcept
get the ErrorRecorder assigned to this interface.
Definition: NvInfer.h:8999
apiv::VBuilder * mImpl
Definition: NvInfer.h:9102
ILogger * getLogger() const noexcept
get the logger with which the builder was created
Definition: NvInfer.h:9068
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:9058
bool platformHasTf32() const noexcept
Determine whether the platform has TF32 support.
Definition: NvInfer.h:9015
int32_t getMaxThreads() const noexcept
get the maximum number of threads that can be used by the builder.
Definition: NvInfer.h:9096
TRT_DEPRECATED void destroy() noexcept
Destroy this object.
Definition: NvInfer.h:8873
int32_t getMaxBatchSize() const noexcept
Get the maximum batch size.
Definition: NvInfer.h:8845
nvinfer1::IOptimizationProfile * createOptimizationProfile() noexcept
Create a new optimization profile.
Definition: NvInfer.h:8965
bool platformHasFastFp16() const noexcept
Determine whether the platform has fast native fp16.
Definition: NvInfer.h:8853
void setGpuAllocator(IGpuAllocator *allocator) noexcept
Set the GPU allocator.
Definition: NvInfer.h:8909
nvinfer1::INetworkDefinition * createNetworkV2(NetworkDefinitionCreationFlags flags) noexcept
Create a network definition object.
Definition: NvInfer.h:8951
nvinfer1::IBuilderConfig * createBuilderConfig() noexcept
Create a builder configuration object.
Definition: NvInfer.h:8919
void reset() noexcept
Resets the builder state to default values.
Definition: NvInfer.h:9007
bool setMaxThreads(int32_t maxThreads) noexcept
Set the maximum number of threads.
Definition: NvInfer.h:9082
void setErrorRecorder(IErrorRecorder *recorder) noexcept
Set the ErrorRecorder for this interface.
Definition: NvInfer.h:8984
bool platformHasFastInt8() const noexcept
Determine whether the platform has fast native int8.
Definition: NvInfer.h:8861
nvinfer1::IHostMemory * buildSerializedNetwork(INetworkDefinition &network, IBuilderConfig &config) noexcept
Builds and serializes a network for the given INetworkDefinition and IBuilderConfig.
Definition: NvInfer.h:9034
virtual ~IBuilder() noexcept=default
TRT_DEPRECATED nvinfer1::ICudaEngine * buildEngineWithConfig(INetworkDefinition &network, IBuilderConfig &config) noexcept
Builds an engine for the given INetworkDefinition and given IBuilderConfig.
Definition: NvInfer.h:8934
A concatenation layer in a network definition.
Definition: NvInfer.h:2392
void setAxis(int32_t axis) noexcept
Set the axis along which concatenation occurs.
Definition: NvInfer.h:2404
int32_t getAxis() const noexcept
Get the axis along which concatenation occurs.
Definition: NvInfer.h:2414
virtual ~IConcatenationLayer() noexcept=default
Definition: NvInfer.h:5171
virtual ~IConditionLayer() noexcept=default
Layer that represents a constant value.
Definition: NvInfer.h:4603
void setWeights(Weights weights) noexcept
Set the weights for the layer.
Definition: NvInfer.h:4614
Weights getWeights() const noexcept
Get the weights for the layer.
Definition: NvInfer.h:4624
apiv::VConstantLayer * mImpl
Definition: NvInfer.h:4654
void setDimensions(Dims dimensions) noexcept
Set the dimensions for the layer.
Definition: NvInfer.h:4636
virtual ~IConstantLayer() noexcept=default
Dims getDimensions() const noexcept
Get the dimensions for the layer.
Definition: NvInfer.h:4648
A convolution layer in a network definition.
Definition: NvInfer.h:1024
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:1127
TRT_DEPRECATED DimsHW getStride() const noexcept
Get the stride of the convolution.
Definition: NvInfer.h:1095
TRT_DEPRECATED DimsHW getDilation() const noexcept
Get the dilation for a convolution.
Definition: NvInfer.h:1234
void setStrideNd(Dims stride) noexcept
Set the multi-dimension stride of the convolution.
Definition: NvInfer.h:1352
void setKernelSizeNd(Dims kernelSize) noexcept
Set the multi-dimension kernel size of the convolution.
Definition: NvInfer.h:1327
Dims getPrePadding() const noexcept
Get the pre-padding.
Definition: NvInfer.h:1261
Weights getBiasWeights() const noexcept
Get the bias weights for the convolution.
Definition: NvInfer.h:1206
int32_t getNbGroups() const noexcept
Get the number of groups of the convolution.
Definition: NvInfer.h:1157
void setPaddingMode(PaddingMode paddingMode) noexcept
Set the padding mode.
Definition: NvInfer.h:1302
TRT_DEPRECATED void setStride(DimsHW stride) noexcept
Get the stride of the convolution.
Definition: NvInfer.h:1085
Dims getPaddingNd() const noexcept
Get the multi-dimension padding of the convolution.
Definition: NvInfer.h:1392
Dims getStrideNd() const noexcept
Get the multi-dimension stride of the convolution.
Definition: NvInfer.h:1362
TRT_DEPRECATED void setPadding(DimsHW padding) noexcept
Set the padding of the convolution.
Definition: NvInfer.h:1115
Weights getKernelWeights() const noexcept
Get the kernel weights of the convolution.
Definition: NvInfer.h:1181
void setPaddingNd(Dims padding) noexcept
Set the multi-dimension padding of the convolution.
Definition: NvInfer.h:1380
Dims getDilationNd() const noexcept
Get the multi-dimension dilation of the convolution.
Definition: NvInfer.h:1416
void setKernelWeights(Weights weights) noexcept
Set the kernel weights for the convolution.
Definition: NvInfer.h:1171
void setDilationNd(Dims dilation) noexcept
Set the multi-dimension dilation of the convolution.
Definition: NvInfer.h:1406
Dims getPostPadding() const noexcept
Get the post-padding.
Definition: NvInfer.h:1288
TRT_DEPRECATED void setDilation(DimsHW dilation) noexcept
Set the dilation for a convolution.
Definition: NvInfer.h:1222
void setNbGroups(int32_t nbGroups) noexcept
Set the number of groups for a convolution.
Definition: NvInfer.h:1147
PaddingMode getPaddingMode() const noexcept
Get the padding mode.
Definition: NvInfer.h:1314
int32_t getNbOutputMaps() const noexcept
Get the number of output maps for the convolution.
Definition: NvInfer.h:1069
void setPrePadding(Dims padding) noexcept
Set the multi-dimension pre-padding of the convolution.
Definition: NvInfer.h:1251
virtual ~IConvolutionLayer() noexcept=default
void setBiasWeights(Weights weights) noexcept
Set the bias weights for the convolution.
Definition: NvInfer.h:1196
Dims getKernelSizeNd() const noexcept
Get the multi-dimension kernel size of the convolution.
Definition: NvInfer.h:1337
void setNbOutputMaps(int32_t nbOutputMaps) noexcept
Set the number of output maps for the convolution.
Definition: NvInfer.h:1059
TRT_DEPRECATED void setKernelSize(DimsHW kernelSize) noexcept
Set the HW kernel size of the convolution.
Definition: NvInfer.h:1035
void setPostPadding(Dims padding) noexcept
Set the multi-dimension post-padding of the convolution.
Definition: NvInfer.h:1278
TRT_DEPRECATED DimsHW getKernelSize() const noexcept
Get the HW kernel size of the convolution.
Definition: NvInfer.h:1047
An engine for executing inference on a built network, with functionally unsafe features.
Definition: NvInferRuntime.h:1379
A deconvolution layer in a network definition.
Definition: NvInfer.h:2432
void setBiasWeights(Weights weights) noexcept
Set the bias weights for the deconvolution.
Definition: NvInfer.h:2610
TRT_DEPRECATED void setStride(DimsHW stride) noexcept
Set the stride of the deconvolution.
Definition: NvInfer.h:2495
void setNbOutputMaps(int32_t nbOutputMaps) noexcept
Set the number of output feature maps for the deconvolution.
Definition: NvInfer.h:2469
Weights getKernelWeights() const noexcept
Get the kernel weights for the deconvolution.
Definition: NvInfer.h:2595
void setPrePadding(Dims padding) noexcept
Set the multi-dimension pre-padding of the deconvolution.
Definition: NvInfer.h:2638
TRT_DEPRECATED DimsHW getPadding() const noexcept
Get the padding of the deconvolution.
Definition: NvInfer.h:2541
Dims getStrideNd() const noexcept
Get the multi-dimension stride of the deconvolution.
Definition: NvInfer.h:2754
int32_t getNbGroups() const noexcept
Get the number of groups for a deconvolution.
Definition: NvInfer.h:2571
Dims getDilationNd() const noexcept
Get the multi-dimension dilation of the deconvolution.
Definition: NvInfer.h:2828
Weights getBiasWeights() const noexcept
Get the bias weights for the deconvolution.
Definition: NvInfer.h:2620
void setKernelWeights(Weights weights) noexcept
Set the kernel weights for the deconvolution.
Definition: NvInfer.h:2585
void setDilationNd(Dims dilation) noexcept
Set the multi-dimension dilation of the deconvolution.
Definition: NvInfer.h:2818
TRT_DEPRECATED DimsHW getStride() const noexcept
Get the stride of the deconvolution.
Definition: NvInfer.h:2507
TRT_DEPRECATED void setPadding(DimsHW padding) noexcept
Set the padding of the deconvolution.
Definition: NvInfer.h:2527
Dims getPostPadding() const noexcept
Get the padding.
Definition: NvInfer.h:2676
Dims getKernelSizeNd() const noexcept
Get the multi-dimension kernel size of the deconvolution.
Definition: NvInfer.h:2727
void setKernelSizeNd(Dims kernelSize) noexcept
Set the multi-dimension kernel size of the deconvolution.
Definition: NvInfer.h:2717
virtual ~IDeconvolutionLayer() noexcept=default
TRT_DEPRECATED void setKernelSize(DimsHW kernelSize) noexcept
Set the HW kernel size of the convolution.
Definition: NvInfer.h:2445
void setStrideNd(Dims stride) noexcept
Set the multi-dimension stride of the deconvolution.
Definition: NvInfer.h:2744
Dims getPaddingNd() const noexcept
Get the multi-dimension padding of the deconvolution.
Definition: NvInfer.h:2784
void setNbGroups(int32_t nbGroups) noexcept
Set the number of groups for a deconvolution.
Definition: NvInfer.h:2561
void setPaddingMode(PaddingMode paddingMode) noexcept
Set the padding mode.
Definition: NvInfer.h:2690
void setPaddingNd(Dims padding) noexcept
Set the multi-dimension padding of the deconvolution.
Definition: NvInfer.h:2772
TRT_DEPRECATED DimsHW getKernelSize() const noexcept
Get the HW kernel size of the deconvolution.
Definition: NvInfer.h:2457
void setPostPadding(Dims padding) noexcept
Set the multi-dimension post-padding of the deconvolution.
Definition: NvInfer.h:2666
int32_t getNbOutputMaps() const noexcept
Get the number of output feature maps for the deconvolution.
Definition: NvInfer.h:2479
Dims getPrePadding() const noexcept
Get the pre-padding.
Definition: NvInfer.h:2648
PaddingMode getPaddingMode() const noexcept
Get the padding mode.
Definition: NvInfer.h:2702
A Dequantize layer in a network definition.
Definition: NvInfer.h:5943
virtual ~IDequantizeLayer() noexcept=default
int32_t getAxis() const noexcept
Get the quantization axis.
Definition: NvInfer.h:5953
void setAxis(int32_t axis) noexcept
Set the quantization axis.
Definition: NvInfer.h:5964
An Einsum layer in a network.
Definition: NvInfer.h:6010
virtual ~IEinsumLayer() noexcept=default
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:6021
const char * getEquation() const noexcept
Return the equation.
Definition: NvInfer.h:6031
A elementwise layer in a network definition.
Definition: NvInfer.h:2903
virtual ~IElementWiseLayer() noexcept=default
apiv::VElementWiseLayer * mImpl
Definition: NvInfer.h:2932
ElementWiseOperation getOperation() const noexcept
Get the binary operation for the layer.
Definition: NvInfer.h:2926
void setOperation(ElementWiseOperation op) noexcept
Set the binary operation for the layer.
Definition: NvInfer.h:2914
Reference counted application-implemented error reporting interface for TensorRT objects.
Definition: NvInferRuntimeCommon.h:1699
Generate an output tensor with specified mode.
Definition: NvInfer.h:5648
FillOperation getOperation() const noexcept
Get the fill operation for the layer.
Definition: NvInfer.h:5694
void setOperation(FillOperation op) noexcept
Set the fill operation for the layer.
Definition: NvInfer.h:5684
void setDimensions(Dims dimensions) noexcept
Set the output tensor's dimensions.
Definition: NvInfer.h:5659
void setBeta(double beta) noexcept
Set the beta parameter.
Definition: NvInfer.h:5745
double getAlpha() const noexcept
Get the value of alpha parameter.
Definition: NvInfer.h:5727
void setAlpha(double alpha) noexcept
Set the alpha parameter.
Definition: NvInfer.h:5712
Dims getDimensions() const noexcept
Get the output tensor's dimensions.
Definition: NvInfer.h:5674
double getBeta() const noexcept
Get the value of beta parameter.
Definition: NvInfer.h:5760
virtual ~IFillLayer() noexcept=default
A fully connected layer in a network definition. This layer expects an input tensor of three or more ...
Definition: NvInfer.h:1481
virtual ~IFullyConnectedLayer() noexcept=default
void setNbOutputChannels(int32_t nbOutputs) noexcept
Set the number of output channels K from the fully connected layer.
Definition: NvInfer.h:1490
Weights getKernelWeights() const noexcept
Get the kernel weights.
Definition: NvInfer.h:1520
void setBiasWeights(Weights weights) noexcept
Set the bias weights.
Definition: NvInfer.h:1532
void setKernelWeights(Weights weights) noexcept
Set the kernel weights, given as a KxC matrix in row-major order.
Definition: NvInfer.h:1510
int32_t getNbOutputChannels() const noexcept
Get the number of output channels K from the fully connected layer.
Definition: NvInfer.h:1500
Weights getBiasWeights() const noexcept
Get the bias weights.
Definition: NvInfer.h:1542
A Gather layer in a network definition. Supports several kinds of gathering.
Definition: NvInfer.h:3038
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:3049
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:3081
apiv::VGatherLayer * mImpl
Definition: NvInfer.h:3117
int32_t getNbElementWiseDims() const noexcept
Get the number of leading dimensions of indices tensor to be handled elementwise.
Definition: NvInfer.h:3091
void setMode(GatherMode mode) noexcept
Set the gather mode.
Definition: NvInfer.h:3101
int32_t getGatherAxis() const noexcept
Get the axis to gather on.
Definition: NvInfer.h:3060
GatherMode getMode() const noexcept
Get the gather mode.
Definition: NvInfer.h:3111
virtual ~IGatherLayer() noexcept=default
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:181
A layer that represents the identity function.
Definition: NvInfer.h:4588
apiv::VIdentityLayer * mImpl
Definition: NvInfer.h:4590
virtual ~IIdentityLayer() noexcept=default
Definition: NvInfer.h:5154
IIfConditional * getConditional() const noexcept
Return pointer to the IIfConditional associated with this boundary layer.
Definition: NvInfer.h:5157
virtual ~IIfConditionalBoundaryLayer() noexcept=default
Definition: NvInfer.h:5224
IIfConditionalInputLayer * addInput(ITensor &input) noexcept
Add an If-conditional input.
Definition: NvInfer.h:5263
virtual ~IIfConditional() noexcept=default
void setName(const char *name) noexcept
Set the name of the conditional.
Definition: NvInfer.h:5276
IConditionLayer * setCondition(ITensor &condition) noexcept
Set the condition tensor for this If-Conditional construct.
Definition: NvInfer.h:5235
IIfConditionalOutputLayer * addOutput(ITensor &trueSubgraphOutput, ITensor &falseSubgraphOutput) noexcept
Add an If-conditional output.
Definition: NvInfer.h:5251
const char * getName() const noexcept
Return the name of the conditional.
Definition: NvInfer.h:5286
Definition: NvInfer.h:5184
virtual ~IIfConditionalOutputLayer() noexcept=default
Application-implemented interface for calibration.
Definition: NvInfer.h:7454
virtual int32_t getBatchSize() const noexcept=0
Get the batch size used for calibration batches.
Definition: NvInfer.h:7537
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:7542
virtual ~IInt8EntropyCalibrator2() noexcept=default
Definition: NvInfer.h:7519
virtual ~IInt8EntropyCalibrator() noexcept=default
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:7524
Definition: NvInfer.h:7572
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:7577
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:7554
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:7559
virtual ~IInt8MinMaxCalibrator() noexcept=default
Definition: NvInfer.h:5414
virtual ~IIteratorLayer() noexcept=default
void setReverse(bool reverse) noexcept
Definition: NvInfer.h:5433
bool getReverse() const noexcept
True if and only if reversing input.
Definition: NvInfer.h:5439
int32_t getAxis() const noexcept
Get axis being iterated over.
Definition: NvInfer.h:5423
void setAxis(int32_t axis) noexcept
Set axis to iterate over.
Definition: NvInfer.h:5417
A LRN layer in a network definition.
Definition: NvInfer.h:2041
int32_t getWindowSize() const noexcept
Get the LRN window size.
Definition: NvInfer.h:2062
float getAlpha() const noexcept
Get the LRN alpha value.
Definition: NvInfer.h:2083
void setWindowSize(int32_t windowSize) noexcept
Set the LRN window size.
Definition: NvInfer.h:2052
void setK(float k) noexcept
Set the LRN K value.
Definition: NvInfer.h:2115
void setAlpha(float alpha) noexcept
Set the LRN alpha value.
Definition: NvInfer.h:2073
void setBeta(float beta) noexcept
Set the LRN beta value.
Definition: NvInfer.h:2094
virtual ~ILRNLayer() noexcept=default
float getBeta() const noexcept
Get the LRN beta value.
Definition: NvInfer.h:2104
float getK() const noexcept
Get the LRN K value.
Definition: NvInfer.h:2125
Base class for all layer classes in a network definition.
Definition: NvInfer.h:529
bool precisionIsSet() const noexcept
whether the computational precision has been set for this layer
Definition: NvInfer.h:668
void setPrecision(DataType dataType) noexcept
Set the computational precision of this layer.
Definition: NvInfer.h:644
void resetPrecision() noexcept
reset the computational precision for this layer
Definition: NvInfer.h:678
int32_t getNbInputs() const noexcept
Get the number of inputs of a layer.
Definition: NvInfer.h:567
DataType getOutputType(int32_t index) const noexcept
get the output type of this layer
Definition: NvInfer.h:730
DataType getPrecision() const noexcept
get the computational precision of this layer
Definition: NvInfer.h:656
int32_t getNbOutputs() const noexcept
Get the number of outputs of a layer.
Definition: NvInfer.h:588
void setName(const char *name) noexcept
Set the name of a layer.
Definition: NvInfer.h:548
bool outputTypeIsSet(int32_t index) const noexcept
whether the output type has been set for this layer
Definition: NvInfer.h:743
ITensor * getOutput(int32_t index) const noexcept
Get the layer output corresponding to the given index.
Definition: NvInfer.h:599
void setInput(int32_t index, ITensor &tensor) noexcept
Replace an input of this layer with a specific tensor.
Definition: NvInfer.h:616
void resetOutputType(int32_t index) noexcept
reset the output type for this layer
Definition: NvInfer.h:755
ITensor * getInput(int32_t index) const noexcept
Get the layer input corresponding to the given index.
Definition: NvInfer.h:580
void setOutputType(int32_t index, DataType dataType) noexcept
Set the output type of this layer.
Definition: NvInfer.h:716
LayerType getType() const noexcept
Return the type of a layer.
Definition: NvInfer.h:536
const char * getName() const noexcept
Return the name of a layer.
Definition: NvInfer.h:559
virtual ~ILayer() noexcept=default
Application-implemented logging interface for the builder, refitter and runtime.
Definition: NvInferRuntimeCommon.h:1510
Definition: NvInfer.h:5135
virtual ~ILoopBoundaryLayer() noexcept=default
ILoop * getLoop() const noexcept
Return pointer to ILoop associated with this boundary layer.
Definition: NvInfer.h:5138
Definition: NvInfer.h:5455
ITripLimitLayer * addTripLimit(ITensor &tensor, TripLimit limit) noexcept
Add a trip-count limiter, based on the given tensor.
Definition: NvInfer.h:5484
IIteratorLayer * addIterator(ITensor &tensor, int32_t axis=0, bool reverse=false) noexcept
Return layer that subscripts tensor by loop iteration.
Definition: NvInfer.h:5497
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:5509
virtual ~ILoop() noexcept=default
const char * getName() const noexcept
Return the name of the loop.
Definition: NvInfer.h:5532
void setName(const char *name) noexcept
Set the name of the loop.
Definition: NvInfer.h:5522
IRecurrenceLayer * addRecurrence(ITensor &initialValue) noexcept
Create a recurrence layer for this loop with initialValue as its first input.
Definition: NvInfer.h:5463
Definition: NvInfer.h:5343
virtual ~ILoopOutputLayer() noexcept=default
int32_t getAxis() const noexcept
Get axis being concatenated over.
Definition: NvInfer.h:5368
LoopOutput getLoopOutput() const noexcept
Definition: NvInfer.h:5345
void setAxis(int32_t axis) noexcept
Set where to insert the contenation axis. Ignored if getLoopOutput() is kLAST_VALUE.
Definition: NvInfer.h:5362
Layer that represents a Matrix Multiplication.
Definition: NvInfer.h:4513
apiv::VMatrixMultiplyLayer * mImpl
Definition: NvInfer.h:4539
virtual ~IMatrixMultiplyLayer() noexcept=default
MatrixOperation getOperation(int32_t index) const noexcept
Get the operation for an input tensor.
Definition: NvInfer.h:4533
void setOperation(int32_t index, MatrixOperation op) noexcept
Set the operation for an input tensor.
Definition: NvInfer.h:4521
A network definition for input to the builder.
Definition: NvInfer.h:6185
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:6887
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:7049
IDequantizeLayer * addDequantize(ITensor &input, ITensor &scale) noexcept
Add a dequantization layer to the network.
Definition: NvInfer.h:7341
IConcatenationLayer * addConcatenation(ITensor *const *inputs, int32_t nbInputs) noexcept
Add a concatenation layer to the network.
Definition: NvInfer.h:6399
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:6422
IGatherLayer * addGatherV2(ITensor &data, ITensor &indices, GatherMode mode)
Add gather with specified mode, axis=0 and nbElementWiseDims=0.
Definition: NvInfer.h:6692
IShuffleLayer * addShuffle(ITensor &input) noexcept
Add a shuffle layer to the network.
Definition: NvInfer.h:6503
ILRNLayer * addLRN(ITensor &input, int32_t window, float alpha, float beta, float k) noexcept
Add a LRN layer to the network.
Definition: NvInfer.h:6342
ITopKLayer * addTopK(ITensor &input, TopKOperation op, int32_t k, uint32_t reduceAxes) noexcept
Add a TopK layer to the network.
Definition: NvInfer.h:6660
IAssertionLayer * addAssertion(ITensor &condition, const char *message) noexcept
Add an assertion layer to the network.
Definition: NvInfer.h:7233
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:7127
TRT_DEPRECATED bool hasExplicitPrecision() const noexcept
True if network is an explicit precision network.
Definition: NvInfer.h:7160
IParametricReLULayer * addParametricReLU(ITensor &input, ITensor &slope) noexcept
Add a parametric ReLU layer to the network.
Definition: NvInfer.h:7027
ITensor * getOutput(int32_t index) const noexcept
Get the output tensor specified by the given index.
Definition: NvInfer.h:6587
ITensor * getInput(int32_t index) const noexcept
Get the input tensor specified by the given index.
Definition: NvInfer.h:6557
bool unmarkOutputForShapes(ITensor &tensor) noexcept
Undo markOutputForShapes.
Definition: NvInfer.h:7009
IFillLayer * addFill(Dims dimensions, FillOperation op) noexcept
Add a fill layer to the network.
Definition: NvInfer.h:7251
ILoop * addLoop() noexcept
Add a loop to the network.
Definition: NvInfer.h:7176
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:7091
IActivationLayer * addActivation(ITensor &input, ActivationType type) noexcept
Add an activation layer to the network.
Definition: NvInfer.h:6304
virtual ~INetworkDefinition() noexcept=default
void setName(const char *name) noexcept
Sets the name of the network.
Definition: NvInfer.h:6928
ILayer * getLayer(int32_t index) const noexcept
Get the layer specified by the given index.
Definition: NvInfer.h:6529
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:6826
TRT_DEPRECATED IFullyConnectedLayer * addFullyConnected(ITensor &input, int32_t nbOutputs, Weights kernelWeights, Weights biasWeights) noexcept
Add a fully connected layer to the network.
Definition: NvInfer.h:6284
IIfConditional * addIfConditional() noexcept
Add an If-conditional layer to the network.
Definition: NvInfer.h:7395
IErrorRecorder * getErrorRecorder() const noexcept
get the ErrorRecorder assigned to this interface.
Definition: NvInfer.h:7322
int32_t getNbInputs() const noexcept
Get the number of inputs in the network.
Definition: NvInfer.h:6541
bool hasImplicitBatchDimension() const noexcept
Query whether the network was created with an implicit batch dimension.
Definition: NvInfer.h:6979
IReduceLayer * addReduce(ITensor &input, ReduceOperation operation, uint32_t reduceAxes, bool keepDimensions) noexcept
Add a reduce layer to the network.
Definition: NvInfer.h:6626
IUnaryLayer * addUnary(ITensor &input, UnaryOperation operation) noexcept
Add a unary layer to the network.
Definition: NvInfer.h:6472
const char * getName() const noexcept
Returns the name associated with the network.
Definition: NvInfer.h:6942
void removeTensor(ITensor &tensor) noexcept
remove a tensor from the network definition.
Definition: NvInfer.h:6856
ISelectLayer * addSelect(ITensor &condition, ITensor &thenInput, ITensor &elseInput) noexcept
Add a select layer to the network.
Definition: NvInfer.h:7216
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:7361
int32_t getNbLayers() const noexcept
Get the number of layers in the network.
Definition: NvInfer.h:6515
apiv::VNetworkDefinition * mImpl
Definition: NvInfer.h:7415
IPoolingLayer * addPoolingNd(ITensor &input, PoolingType type, Dims windowSize) noexcept
Add a multi-dimension pooling layer to the network.
Definition: NvInfer.h:7069
bool markOutputForShapes(ITensor &tensor) noexcept
Enable tensor's value to be computed by IExecutionContext::getShapeBinding.
Definition: NvInfer.h:6997
TRT_DEPRECATED IPaddingLayer * addPadding(ITensor &input, DimsHW prePadding, DimsHW postPadding) noexcept
Add a padding layer to the network.
Definition: NvInfer.h:6489
IScaleLayer * addScale(ITensor &input, ScaleMode mode, Weights shift, Weights scale, Weights power) noexcept
Add a Scale layer to the network.
Definition: NvInfer.h:6369
void unmarkOutput(ITensor &tensor) noexcept
unmark a tensor as a network output.
Definition: NvInfer.h:6868
IIdentityLayer * addIdentity(ITensor &input) noexcept
Add an identity layer.
Definition: NvInfer.h:6841
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:6261
IQuantizeLayer * addQuantize(ITensor &input, ITensor &scale) noexcept
Add a quantization layer to the network.
Definition: NvInfer.h:7380
IElementWiseLayer * addElementWise(ITensor &input1, ITensor &input2, ElementWiseOperation op) noexcept
Add an elementwise layer to the network.
Definition: NvInfer.h:6450
void setErrorRecorder(IErrorRecorder *recorder) noexcept
Set the ErrorRecorder for this interface.
Definition: NvInfer.h:7307
ISliceLayer * addSlice(ITensor &input, Dims start, Dims size, Dims stride) noexcept
Add a slice layer to the network.
Definition: NvInfer.h:6906
IConstantLayer * addConstant(Dims dimensions, Weights weights) noexcept
Add a constant layer to the network.
Definition: NvInfer.h:6757
IRaggedSoftMaxLayer * addRaggedSoftMax(ITensor &input, ITensor &bounds) noexcept
Add a RaggedSoftMax layer to the network.
Definition: NvInfer.h:6710
IShapeLayer * addShape(ITensor &input) noexcept
Add a shape layer to the network.
Definition: NvInfer.h:6960
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:6676
IEinsumLayer * addEinsum(ITensor *const *inputs, int32_t nbInputs, const char *equation) noexcept
Add an Einsum layer to the network.
Definition: NvInfer.h:7409
TRT_DEPRECATED IPoolingLayer * addPooling(ITensor &input, PoolingType type, DimsHW windowSize) noexcept
Add a pooling layer to the network.
Definition: NvInfer.h:6323
TRT_DEPRECATED void destroy() noexcept
Destroy this INetworkDefinition object.
Definition: NvInfer.h:6599
IResizeLayer * addResize(ITensor &input) noexcept
Add a resize layer to the network.
Definition: NvInfer.h:7143
IMatrixMultiplyLayer * addMatrixMultiply(ITensor &input0, MatrixOperation op0, ITensor &input1, MatrixOperation op1) noexcept
Add a MatrixMultiply layer to the network.
Definition: NvInfer.h:6731
ISoftMaxLayer * addSoftMax(ITensor &input) noexcept
Add a SoftMax layer to the network.
Definition: NvInfer.h:6382
void markOutput(ITensor &tensor) noexcept
Mark a tensor as a network output.
Definition: NvInfer.h:6238
bool setWeightsName(Weights weights, const char *name) noexcept
Associate a name with all current uses of the given weights.
Definition: NvInfer.h:7288
int32_t getNbOutputs() const noexcept
Get the number of outputs in the network.
Definition: NvInfer.h:6571
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:7268
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:1160
Layer that represents a padding operation.
Definition: NvInfer.h:3835
Dims getPostPaddingNd() const noexcept
Get the padding that is applied at the end of the tensor.
Definition: NvInfer.h:3936
TRT_DEPRECATED DimsHW getPrePadding() const noexcept
Get the padding that is applied at the start of the tensor.
Definition: NvInfer.h:3858
TRT_DEPRECATED void setPostPadding(DimsHW padding) noexcept
Set the padding that is applied at the end of the tensor.
Definition: NvInfer.h:3872
virtual ~IPaddingLayer() noexcept=default
TRT_DEPRECATED void setPrePadding(DimsHW padding) noexcept
Set the padding that is applied at the start of the tensor.
Definition: NvInfer.h:3846
Dims getPrePaddingNd() const noexcept
Get the padding that is applied at the start of the tensor.
Definition: NvInfer.h:3910
TRT_DEPRECATED DimsHW getPostPadding() const noexcept
Get the padding that is applied at the end of the tensor.
Definition: NvInfer.h:3884
void setPostPaddingNd(Dims padding) noexcept
Set the padding that is applied at the end of the tensor.
Definition: NvInfer.h:3924
void setPrePaddingNd(Dims padding) noexcept
Set the padding that is applied at the start of the tensor.
Definition: NvInfer.h:3898
apiv::VPaddingLayer * mImpl
Definition: NvInfer.h:3942
Layer that represents a parametric ReLU operation.
Definition: NvInfer.h:4668
apiv::VParametricReLULayer * mImpl
Definition: NvInfer.h:4670
virtual ~IParametricReLULayer() noexcept=default
Single registration point for all plugins in an application. It is used to find plugin implementation...
Definition: NvInferRuntimeCommon.h:1244
Plugin class for user-implemented layers.
Definition: NvInferRuntimeCommon.h:411
Layer type for pluginV2.
Definition: NvInfer.h:3606
virtual ~IPluginV2Layer() noexcept=default
apiv::VPluginV2Layer * mImpl
Definition: NvInfer.h:3619
IPluginV2 & getPlugin() noexcept
Get the plugin for the layer.
Definition: NvInfer.h:3613
A Pooling layer in a network definition.
Definition: NvInfer.h:1703
TRT_DEPRECATED DimsHW getStride() const noexcept
Get the stride for pooling.
Definition: NvInfer.h:1776
PoolingType getPoolingType() const noexcept
Get the type of activation to be performed.
Definition: NvInfer.h:1722
void setWindowSizeNd(Dims windowSize) noexcept
Set the multi-dimension window size for pooling.
Definition: NvInfer.h:1955
PaddingMode getPaddingMode() const noexcept
Get the padding mode.
Definition: NvInfer.h:1942
Dims getPostPadding() const noexcept
Get the padding.
Definition: NvInfer.h:1918
TRT_DEPRECATED void setStride(DimsHW stride) noexcept
Set the stride for pooling.
Definition: NvInfer.h:1764
bool getAverageCountExcludesPadding() const noexcept
Get whether average pooling uses as a denominator the overlap area between the window and the unpadde...
Definition: NvInfer.h:1862
Dims getPrePadding() const noexcept
Get the pre-padding.
Definition: NvInfer.h:1890
void setPoolingType(PoolingType type) noexcept
Set the type of activation to be performed.
Definition: NvInfer.h:1712
void setPaddingMode(PaddingMode paddingMode) noexcept
Set the padding mode.
Definition: NvInfer.h:1931
TRT_DEPRECATED void setWindowSize(DimsHW windowSize) noexcept
Set the window size for pooling.
Definition: NvInfer.h:1736
Dims getWindowSizeNd() const noexcept
Get the multi-dimension window size for pooling.
Definition: NvInfer.h:1965
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:1851
void setPaddingNd(Dims padding) noexcept
Set the multi-dimension padding for pooling.
Definition: NvInfer.h:2009
void setPrePadding(Dims padding) noexcept
Set the multi-dimension pre-padding for pooling.
Definition: NvInfer.h:1880
float getBlendFactor() const noexcept
Get the blending factor for the max_average_blend mode: max_average_blendPool = (1-blendFactor)*maxPo...
Definition: NvInfer.h:1834
Dims getStrideNd() const noexcept
Get the multi-dimension stride for pooling.
Definition: NvInfer.h:1990
virtual ~IPoolingLayer() noexcept=default
Dims getPaddingNd() const noexcept
Get the multi-dimension padding for pooling.
Definition: NvInfer.h:2021
TRT_DEPRECATED DimsHW getWindowSize() const noexcept
Get the window size for pooling.
Definition: NvInfer.h:1748
TRT_DEPRECATED DimsHW getPadding() const noexcept
Get the padding for pooling.
Definition: NvInfer.h:1806
void setStrideNd(Dims stride) noexcept
Set the multi-dimension stride for pooling.
Definition: NvInfer.h:1980
void setPostPadding(Dims padding) noexcept
Set the multi-dimension post-padding for pooling.
Definition: NvInfer.h:1908
TRT_DEPRECATED void setPadding(DimsHW padding) noexcept
Set the padding for pooling.
Definition: NvInfer.h:1792
void setBlendFactor(float blendFactor) noexcept
Set the blending factor for the max_average_blend mode: max_average_blendPool = (1-blendFactor)*maxPo...
Definition: NvInfer.h:1821
A Quantize layer in a network definition.
Definition: NvInfer.h:5856
void setAxis(int32_t axis) noexcept
Set the quantization axis.
Definition: NvInfer.h:5877
int32_t getAxis() const noexcept
Get the quantization axis.
Definition: NvInfer.h:5866
virtual ~IQuantizeLayer() noexcept=default
An RNN layer in a network definition, version 2.
Definition: NvInfer.h:3316
void setDirection(RNNDirection op) noexcept
Set the direction of the RNN layer.
Definition: NvInfer.h:3416
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:3520
void setCellState(ITensor &cell) noexcept
Set the initial cell state of the LSTM with the provided cell ITensor.
Definition: NvInfer.h:3576
int32_t getDataLength() const noexcept
Get the maximum data length of the RNN.
Definition: NvInfer.h:3330
void setInputMode(RNNInputMode op) noexcept
Set the input mode of the RNN layer.
Definition: NvInfer.h:3391
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:3530
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:3485
void setOperation(RNNOperation op) noexcept
Set the operation of the RNN layer.
Definition: NvInfer.h:3371
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:3495
RNNDirection getDirection() const noexcept
Get the direction of the RNN layer.
Definition: NvInfer.h:3426
void setSequenceLengths(ITensor &seqLengths) noexcept
Specify individual sequence lengths in the batch with the ITensor pointed to by seqLengths.
Definition: NvInfer.h:3349
RNNInputMode getInputMode() const noexcept
Get the input mode of the RNN layer.
Definition: NvInfer.h:3401
ITensor * getHiddenState() const noexcept
Get the initial hidden state of the RNN.
Definition: NvInfer.h:3557
ITensor * getCellState() const noexcept
Get the initial cell state of the RNN.
Definition: NvInfer.h:3586
int32_t getMaxSeqLength() const noexcept
Get the maximum sequence length of the RNN.
Definition: NvInfer.h:3326
int32_t getLayerCount() const noexcept
Get the layer count of the RNN.
Definition: NvInfer.h:3318
apiv::VRNNv2Layer * mImpl
Definition: NvInfer.h:3592
ITensor * getSequenceLengths() const noexcept
Get the sequence lengths specified for the RNN.
Definition: NvInfer.h:3361
RNNOperation getOperation() const noexcept
Get the operation of the RNN layer.
Definition: NvInfer.h:3381
virtual ~IRNNv2Layer() noexcept=default
int32_t getHiddenSize() const noexcept
Get the hidden size of the RNN.
Definition: NvInfer.h:3322
void setHiddenState(ITensor &hidden) noexcept
Set the initial hidden state of the RNN with the provided hidden ITensor.
Definition: NvInfer.h:3547
A RaggedSoftmax layer in a network definition.
Definition: NvInfer.h:4558
apiv::VRaggedSoftMaxLayer * mImpl
Definition: NvInfer.h:4560
virtual ~IRaggedSoftMaxLayer() noexcept=default
Definition: NvInfer.h:5298
virtual ~IRecurrenceLayer() noexcept=default
Layer that represents a reduction across a non-bool tensor.
Definition: NvInfer.h:3757
void setKeepDimensions(bool keepDimensions) noexcept
Set the boolean that specifies whether or not to keep the reduced dimensions for the layer.
Definition: NvInfer.h:3804
void setOperation(ReduceOperation op) noexcept
Set the reduce operation for the layer.
Definition: NvInfer.h:3764
ReduceOperation getOperation() const noexcept
Get the reduce operation for the layer.
Definition: NvInfer.h:3774
virtual ~IReduceLayer() noexcept=default
uint32_t getReduceAxes() const noexcept
Get the axes over which to reduce for the layer.
Definition: NvInfer.h:3794
void setReduceAxes(uint32_t reduceAxes) noexcept
Set the axes over which to reduce.
Definition: NvInfer.h:3784
apiv::VReduceLayer * mImpl
Definition: NvInfer.h:3820
bool getKeepDimensions() const noexcept
Get the boolean that specifies whether or not to keep the reduced dimensions for the layer.
Definition: NvInfer.h:3814
A resize layer in a network definition.
Definition: NvInfer.h:4856
void setSelectorForSinglePixel(ResizeSelector selector) noexcept
Set coordinate selector function when resized to single pixel.
Definition: NvInfer.h:5045
void setOutputDimensions(Dims dimensions) noexcept
Set the output dimensions.
Definition: NvInfer.h:4876
void setNearestRounding(ResizeRoundMode value) noexcept
Set rounding mode for nearest neighbor resize.
Definition: NvInfer.h:5069
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:4935
ResizeMode getResizeMode() const noexcept
Get resize mode for an input tensor.
Definition: NvInfer.h:4957
void setResizeMode(ResizeMode resizeMode) noexcept
Set resize mode for an input tensor.
Definition: NvInfer.h:4947
ResizeSelector getSelectorForSinglePixel() const noexcept
Get the coordinate selector function when resized to single pixel.
Definition: NvInfer.h:5055
TRT_DEPRECATED bool getAlignCorners() const noexcept
True if align corners has been set.
Definition: NvInfer.h:4985
void setCoordinateTransformation(ResizeCoordinateTransformation coordTransform) noexcept
Set coordinate transformation function.
Definition: NvInfer.h:5020
void setScales(const float *scales, int32_t nbScales) noexcept
Set the resize scales.
Definition: NvInfer.h:4916
Dims getOutputDimensions() const noexcept
Get the output dimensions.
Definition: NvInfer.h:4886
ResizeRoundMode getNearestRounding() const noexcept
Get rounding mode for nearest neighbor resize.
Definition: NvInfer.h:5079
TRT_DEPRECATED void setAlignCorners(bool alignCorners) noexcept
Set whether to align corners while resizing.
Definition: NvInfer.h:4973
ResizeCoordinateTransformation getCoordinateTransformation() const noexcept
Get coordinate transformation function.
Definition: NvInfer.h:5030
A Scale layer in a network definition.
Definition: NvInfer.h:2185
Weights getScale() const noexcept
Get the scale value.
Definition: NvInfer.h:2242
Weights getPower() const noexcept
Get the power value.
Definition: NvInfer.h:2262
void setScale(Weights scale) noexcept
Set the scale value.
Definition: NvInfer.h:2232
void setPower(Weights power) noexcept
Set the power value.
Definition: NvInfer.h:2252
ScaleMode getMode() const noexcept
Get the scale mode.
Definition: NvInfer.h:2202
void setShift(Weights shift) noexcept
Set the shift value.
Definition: NvInfer.h:2212
void setChannelAxis(int32_t channelAxis) noexcept
Set the channel axis.
Definition: NvInfer.h:2298
Weights getShift() const noexcept
Get the shift value.
Definition: NvInfer.h:2222
virtual ~IScaleLayer() noexcept=default
void setMode(ScaleMode mode) noexcept
Set the scale mode.
Definition: NvInfer.h:2192
int32_t getChannelAxis() const noexcept
Get the channel axis.
Definition: NvInfer.h:2277
A scatter layer in a network definition. Supports several kinds of scattering.
Definition: NvInfer.h:6120
void setMode(ScatterMode mode) noexcept
Set the scatter mode.
Definition: NvInfer.h:6127
apiv::VScatterLayer * mImpl
Definition: NvInfer.h:6161
void setAxis(int32_t axis) noexcept
Set the axis used by ScatterMode::kELEMENTS.
Definition: NvInfer.h:6147
int32_t getAxis() const noexcept
Get the axis.
Definition: NvInfer.h:6155
ScatterMode getMode() const noexcept
Get the scatter mode.
Definition: NvInfer.h:6137
virtual ~IScatterLayer() noexcept=default
Definition: NvInfer.h:5546
virtual ~ISelectLayer() noexcept=default
Layer type for getting shape of a tensor.
Definition: NvInfer.h:4342
virtual ~IShapeLayer() noexcept=default
apiv::VShapeLayer * mImpl
Definition: NvInfer.h:4344
Layer type for shuffling data.
Definition: NvInfer.h:3970
apiv::VShuffleLayer * mImpl
Definition: NvInfer.h:4125
void setReshapeDimensions(Dims dimensions) noexcept
Set the reshaped dimensions.
Definition: NvInfer.h:4018
void setFirstTranspose(Permutation permutation) noexcept
Set the permutation applied by the first transpose operation.
Definition: NvInfer.h:3981
void setSecondTranspose(Permutation permutation) noexcept
Set the permutation applied by the second transpose operation.
Definition: NvInfer.h:4078
Dims getReshapeDimensions() const noexcept
Get the reshaped dimensions.
Definition: NvInfer.h:4031
Permutation getFirstTranspose() const noexcept
Get the permutation applied by the first transpose operation.
Definition: NvInfer.h:3993
virtual ~IShuffleLayer() noexcept=default
Permutation getSecondTranspose() const noexcept
Get the permutation applied by the second transpose operation.
Definition: NvInfer.h:4090
bool getZeroIsPlaceholder() const noexcept
Get meaning of 0 in reshape dimensions.
Definition: NvInfer.h:4119
void setZeroIsPlaceholder(bool zeroIsPlaceholder) noexcept
Set meaning of 0 in reshape dimensions.
Definition: NvInfer.h:4106
Slices an input tensor into an output tensor based on the offset and strides.
Definition: NvInfer.h:4192
void setMode(SliceMode mode) noexcept
Set the slice mode.
Definition: NvInfer.h:4286
apiv::VSliceLayer * mImpl
Definition: NvInfer.h:4325
virtual ~ISliceLayer() noexcept=default
void setStride(Dims stride) noexcept
Set the stride for computing the output slice data.
Definition: NvInfer.h:4261
void setStart(Dims start) noexcept
Set the start offset that the slice layer uses to create the output slice.
Definition: NvInfer.h:4203
Dims getStart() const noexcept
Get the start offset for the slice layer.
Definition: NvInfer.h:4218
void setSize(Dims size) noexcept
Set the dimensions of the output slice.
Definition: NvInfer.h:4232
Dims getSize() const noexcept
Get dimensions of the output slice.
Definition: NvInfer.h:4247
SliceMode getMode() const noexcept
Get the slice mode.
Definition: NvInfer.h:4296
Dims getStride() const noexcept
Get the stride for the output slice.
Definition: NvInfer.h:4276
A Softmax layer in a network definition.
Definition: NvInfer.h:2325
void setAxes(uint32_t axes) noexcept
Set the axis along which softmax is computed. Currently, only one axis can be set.
Definition: NvInfer.h:2359
uint32_t getAxes() const noexcept
Get the axis along which softmax occurs.
Definition: NvInfer.h:2369
virtual ~ISoftMaxLayer() noexcept=default
A tensor in a network definition.
Definition: NvInfer.h:204
bool setDynamicRange(float min, float max) noexcept
Set dynamic range for the tensor.
Definition: NvInfer.h:304
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:440
TensorLocation getLocation() const noexcept
Get the storage location of a tensor.
Definition: NvInfer.h:368
void setName(const char *name) noexcept
Set the tensor name.
Definition: NvInfer.h:218
void resetDynamicRange() noexcept
Undo effect of setDynamicRange.
Definition: NvInfer.h:401
bool isExecutionTensor() const noexcept
Whether the tensor is an execution tensor.
Definition: NvInfer.h:511
void setType(DataType type) noexcept
Set the data type of a tensor.
Definition: NvInfer.h:277
bool dynamicRangeIsSet() const noexcept
Query whether dynamic range is set.
Definition: NvInfer.h:393
void setLocation(TensorLocation location) noexcept
Set the storage location of a tensor.
Definition: NvInfer.h:383
bool isShapeTensor() const noexcept
Whether the tensor is a shape tensor.
Definition: NvInfer.h:488
float getDynamicRangeMax() const noexcept
Get maximum of dynamic range.
Definition: NvInfer.h:421
bool isNetworkInput() const noexcept
Whether the tensor is a network input.
Definition: NvInfer.h:312
bool isNetworkOutput() const noexcept
Whether the tensor is a network output.
Definition: NvInfer.h:320
bool getBroadcastAcrossBatch() const noexcept
Check if tensor is broadcast across the batch.
Definition: NvInfer.h:358
void setBroadcastAcrossBatch(bool broadcastAcrossBatch) noexcept
Set whether to enable broadcast of tensor across the batch.
Definition: NvInfer.h:342
DataType getType() const noexcept
Get the data type of a tensor.
Definition: NvInfer.h:289
apiv::VTensor * mImpl
Definition: NvInfer.h:517
float getDynamicRangeMin() const noexcept
Get minimum of dynamic range.
Definition: NvInfer.h:411
virtual ~ITensor() noexcept=default
void setDimensions(Dims dimensions) noexcept
Set the dimensions of a tensor.
Definition: NvInfer.h:249
const char * getName() const noexcept
Get the tensor name.
Definition: NvInfer.h:230
Dims getDimensions() const noexcept
Get the dimensions of a tensor.
Definition: NvInfer.h:262
TensorFormats getAllowedFormats() const noexcept
Get a bitmask of TensorFormat values that the tensor supports. For a shape tensor,...
Definition: NvInfer.h:453
Class to handle tactic timing info collected from builder.
Definition: NvInfer.h:7993
bool combine(const ITimingCache &inputCache, bool ignoreMismatch) noexcept
Combine input timing cache into local instance.
Definition: NvInfer.h:8030
virtual ~ITimingCache() noexcept=default
apiv::VTimingCache * mImpl
Definition: NvInfer.h:8046
bool reset() noexcept
Empty the timing cache.
Definition: NvInfer.h:8040
Layer that represents a TopK reduction.
Definition: NvInfer.h:4378
void setK(int32_t k) noexcept
Set the k value for the layer.
Definition: NvInfer.h:4407
void setReduceAxes(uint32_t reduceAxes) noexcept
Set which axes to reduce for the layer.
Definition: NvInfer.h:4427
TopKOperation getOperation() const noexcept
Get the operation for the layer.
Definition: NvInfer.h:4395
apiv::VTopKLayer * mImpl
Definition: NvInfer.h:4443
void setOperation(TopKOperation op) noexcept
Set the operation for the layer.
Definition: NvInfer.h:4385
int32_t getK() const noexcept
Get the k value for the layer.
Definition: NvInfer.h:4417
uint32_t getReduceAxes() const noexcept
Get the axes to reduce for the layer.
Definition: NvInfer.h:4437
virtual ~ITopKLayer() noexcept=default
Definition: NvInfer.h:5401
TripLimit getTripLimit() const noexcept
Definition: NvInfer.h:5403
virtual ~ITripLimitLayer() noexcept=default
Layer that represents an unary operation.
Definition: NvInfer.h:3682
void setOperation(UnaryOperation op) noexcept
Set the unary operation for the layer.
Definition: NvInfer.h:3691
apiv::VUnaryLayer * mImpl
Definition: NvInfer.h:3707
UnaryOperation getOperation() const noexcept
Get the unary operation for the layer.
Definition: NvInfer.h:3701
virtual ~IUnaryLayer() noexcept=default
An array of weights used as a layer parameter.
Definition: NvInferRuntime.h:163
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:1341
ResizeSelector
The coordinate selector when resize to single pixel output.
Definition: NvInfer.h:4761
@ 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
nvinfer1::IPluginRegistry * getBuilderPluginRegistry(nvinfer1::EngineCapability capability) noexcept
Return the plugin registry for the given capability or nullptr if no registry exists.
MemoryPoolType
The type for memory pools used by TensorRT.
Definition: NvInfer.h:8057
ScaleMode
Controls how shift, scale and power are applied in a Scale layer.
Definition: NvInfer.h:2141
@ 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:7874
constexpr int32_t EnumMax< RNNDirection >() noexcept
Definition: NvInfer.h:3238
constexpr int32_t EnumMax< BuilderFlag >() noexcept
Definition: NvInfer.h:7977
constexpr int32_t EnumMax< LayerType >() noexcept
Definition: NvInfer.h:140
constexpr int32_t EnumMax< RNNGateType >() noexcept
Definition: NvInfer.h:3299
constexpr int32_t EnumMax< CalibrationAlgoType >() noexcept
Definition: NvInfer.h:7437
UnaryOperation
Enumerates the unary operations that may be performed by a Unary layer.
Definition: NvInfer.h:3637
@ 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
Definition: NvInfer.h:3744
constexpr int32_t EnumMax< TripLimit >() noexcept
Definition: NvInfer.h:5127
constexpr int32_t EnumMax< RNNInputMode >() noexcept
Definition: NvInfer.h:3270
RNNInputMode
Enumerates the RNN input modes that may occur with an RNN layer.
Definition: NvInfer.h:3259
@ 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:159
@ 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:5606
@ 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:4791
@ kHALF_DOWN
Round half down.
RNNGateType
Identifies an individual gate within an RNN cell.
Definition: NvInfer.h:3283
@ 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:988
@ 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:5115
@ 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:8766
constexpr int32_t EnumMax< GatherMode >() noexcept
Definition: NvInfer.h:2954
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:7908
DeviceType
The device that this layer/network will execute on.
Definition: NvInferRuntime.h:639
constexpr int32_t EnumMax< ScaleMode >() noexcept
Definition: NvInfer.h:2153
CalibrationAlgoType
Version of calibration algorithm to use.
Definition: NvInfer.h:7424
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:4135
@ kCLAMP
Out of bounds indices are clamped to bounds.
@ kWRAP
Coordinates wrap around periodically.
constexpr int32_t EnumMax< QuantizationFlag >() noexcept
Definition: NvInfer.h:7897
GatherMode
Control form of IGatherLayer.
Definition: NvInfer.h:2942
@ 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:151
ProfilingVerbosity
List of verbosity levels of layer information exposed in NVTX annotations and in IEngineInspector.
Definition: NvInferRuntime.h:1353
ResizeMode
Enumerates various modes of resize in the resize layer. Resize mode set using setResizeMode().
Definition: NvInfer.h:4680
@ kNEAREST
ND (0 < N <= 8) nearest neighbor resizing.
constexpr int32_t EnumMax< SliceMode >() noexcept
Definition: NvInfer.h:4151
NetworkDefinitionCreationFlag
List of immutable network properties expressed at network creation time. NetworkDefinitionCreationFla...
Definition: NvInfer.h:8778
ElementWiseOperation
Enumerates the binary operations that may be performed by an ElementWise layer.
Definition: NvInfer.h:2852
@ 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:7884
@ kCALIBRATE_BEFORE_FUSION
RNNDirection
Enumerates the RNN direction that may be performed by an RNN layer.
Definition: NvInfer.h:3227
@ 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:7918
@ 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
Definition: NvInfer.h:4365
TensorFormat
Format of the input/output tensors.
Definition: NvInferRuntimeCommon.h:221
constexpr int32_t EnumMax< MemoryPoolType >() noexcept
Definition: NvInfer.h:8096
TopKOperation
Enumerates the operations that may be performed by a TopK layer.
Definition: NvInfer.h:4354
ReduceOperation
Enumerates the reduce operations that may be performed by a Reduce layer.
Definition: NvInfer.h:3730
constexpr int32_t EnumMax< LoopOutput >() noexcept
Definition: NvInfer.h:5108
RNNOperation
Enumerates the RNN operations that may be performed by an RNN layer.
Definition: NvInfer.h:3201
@ kGRU
Three-gate network consisting of Gated Recurrent Units.
@ kLSTM
Four-gate LSTM network w/o peephole connections.
constexpr int32_t EnumMax< NetworkDefinitionCreationFlag >() noexcept
Definition: NvInfer.h:8807
ScatterMode
Control form of IScatterLayer.
Definition: NvInfer.h:6047
MatrixOperation
Enumerates the operations that may be performed on a tensor by IMatrixMultiplyLayer before multiplica...
Definition: NvInfer.h:4454
@ kTRANSPOSE
Like kNONE, but transpose the matrix dimensions.
ResizeCoordinateTransformation
The resize coordinate transformation function.
Definition: NvInfer.h:4707
constexpr int32_t EnumMax< UnaryOperation >() noexcept
Definition: NvInfer.h:3669
LoopOutput
Enum that describes kinds of loop outputs.
Definition: NvInfer.h:5091
@ 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
Definition: NvInfer.h:4482
constexpr int32_t EnumMax< RNNOperation >() noexcept
Definition: NvInfer.h:3214
PoolingType
The type of pooling to perform in a pooling layer.
Definition: NvInfer.h:1671
constexpr int32_t EnumMax< FillOperation >() noexcept
Definition: NvInfer.h:5617
TensorLocation
The location for tensor data storage, device or host.
Definition: NvInferRuntime.h:253
OptProfileSelector
When setting or querying optimization profile parameters (such as shape tensor inputs or dynamic dime...
Definition: NvInferRuntime.h:1120
constexpr int32_t EnumMax< ScatterMode >() noexcept
Definition: NvInfer.h:6058
Definition: NvInfer.h:3947
Declaration of EnumMaxImpl struct to store maximum number of elements in an enumeration type.
Definition: NvInferRuntimeCommon.h:136