151 static constexpr int32_t kVALUE = 12;
190 mImpl->setName(name);
202 return mImpl->getName();
221 mImpl->setDimensions(dimensions);
234 return mImpl->getDimensions();
249 mImpl->setType(type);
261 return mImpl->getType();
276 return mImpl->setDynamicRange(min, max);
284 return mImpl->isNetworkInput();
292 return mImpl->isNetworkOutput();
314 mImpl->setBroadcastAcrossBatch(broadcastAcrossBatch);
330 return mImpl->getBroadcastAcrossBatch();
340 return mImpl->getLocation();
355 mImpl->setLocation(location);
365 return mImpl->dynamicRangeIsSet();
373 mImpl->resetDynamicRange();
383 return mImpl->getDynamicRangeMin();
393 return mImpl->getDynamicRangeMax();
412 mImpl->setAllowedFormats(formats);
425 return mImpl->getAllowedFormats();
460 return mImpl->isShapeTensor();
483 return mImpl->isExecutionTensor();
509 mImpl->setDimensionName(index, name);
524 return mImpl->getDimensionName(index);
549 return mLayer->getType();
563 mLayer->setName(name);
573 return mLayer->getName();
581 return mLayer->getNbInputs();
594 return mLayer->getInput(index);
602 return mLayer->getNbOutputs();
613 return mLayer->getOutput(index);
630 return mLayer->setInput(index, tensor);
658 mLayer->setPrecision(dataType);
670 return mLayer->getPrecision();
682 return mLayer->precisionIsSet();
692 mLayer->resetPrecision();
730 mLayer->setOutputType(index, dataType);
744 return mLayer->getOutputType(index);
757 return mLayer->outputTypeIsSet(index);
769 return mLayer->resetOutputType(index);
774 apiv::VLayer* mLayer;
1019 static constexpr int32_t kVALUE = 6;
1049 mImpl->setKernelSize(kernelSize);
1061 return mImpl->getKernelSize();
1073 mImpl->setNbOutputMaps(nbOutputMaps);
1083 return mImpl->getNbOutputMaps();
1099 mImpl->setStride(stride);
1109 return mImpl->getStride();
1129 return mImpl->setPadding(padding);
1141 return mImpl->getPadding();
1161 mImpl->setNbGroups(nbGroups);
1171 return mImpl->getNbGroups();
1185 mImpl->setKernelWeights(weights);
1195 return mImpl->getKernelWeights();
1210 mImpl->setBiasWeights(weights);
1220 return mImpl->getBiasWeights();
1236 return mImpl->setDilation(dilation);
1248 return mImpl->getDilation();
1265 mImpl->setPrePadding(padding);
1275 return mImpl->getPrePadding();
1292 mImpl->setPostPadding(padding);
1302 return mImpl->getPostPadding();
1316 mImpl->setPaddingMode(paddingMode);
1328 return mImpl->getPaddingMode();
1341 mImpl->setKernelSizeNd(kernelSize);
1351 return mImpl->getKernelSizeNd();
1366 mImpl->setStrideNd(stride);
1376 return mImpl->getStrideNd();
1394 mImpl->setPaddingNd(padding);
1406 return mImpl->getPaddingNd();
1420 mImpl->setDilationNd(dilation);
1430 return mImpl->getDilationNd();
1496 mImpl->setNbOutputChannels(nbOutputs);
1506 return mImpl->getNbOutputChannels();
1516 mImpl->setKernelWeights(weights);
1526 return mImpl->getKernelWeights();
1538 mImpl->setBiasWeights(weights);
1548 return mImpl->getBiasWeights();
1604 mImpl->setActivationType(type);
1614 return mImpl->getActivationType();
1629 mImpl->setAlpha(alpha);
1643 mImpl->setBeta(beta);
1652 return mImpl->getAlpha();
1661 return mImpl->getBeta();
1691 static constexpr int32_t kVALUE = 3;
1718 mImpl->setPoolingType(type);
1728 return mImpl->getPoolingType();
1742 mImpl->setWindowSize(windowSize);
1754 return mImpl->getWindowSize();
1770 mImpl->setStride(stride);
1782 return mImpl->getStride();
1798 mImpl->setPadding(padding);
1812 return mImpl->getPadding();
1827 mImpl->setBlendFactor(blendFactor);
1840 return mImpl->getBlendFactor();
1857 mImpl->setAverageCountExcludesPadding(exclusive);
1868 return mImpl->getAverageCountExcludesPadding();
1886 mImpl->setPrePadding(padding);
1896 return mImpl->getPrePadding();
1914 mImpl->setPostPadding(padding);
1924 return mImpl->getPostPadding();
1937 mImpl->setPaddingMode(paddingMode);
1948 return mImpl->getPaddingMode();
1961 mImpl->setWindowSizeNd(windowSize);
1971 return mImpl->getWindowSizeNd();
1986 mImpl->setStrideNd(stride);
1996 return mImpl->getStrideNd();
2015 mImpl->setPaddingNd(padding);
2027 return mImpl->getPaddingNd();
2058 mImpl->setWindowSize(windowSize);
2068 return mImpl->getWindowSize();
2079 mImpl->setAlpha(alpha);
2089 return mImpl->getAlpha();
2100 mImpl->setBeta(beta);
2110 return mImpl->getBeta();
2131 return mImpl->getK();
2198 mImpl->setMode(mode);
2208 return mImpl->getMode();
2218 mImpl->setShift(shift);
2228 return mImpl->getShift();
2238 mImpl->setScale(scale);
2248 return mImpl->getScale();
2258 mImpl->setPower(power);
2268 return mImpl->getPower();
2283 return mImpl->getChannelAxis();
2304 mImpl->setChannelAxis(channelAxis);
2368 mImpl->setAxes(axes);
2378 return mImpl->getAxes();
2415 mImpl->setAxis(axis);
2425 return mImpl->getAxis();
2456 mImpl->setKernelSize(kernelSize);
2468 return mImpl->getKernelSize();
2480 mImpl->setNbOutputMaps(nbOutputMaps);
2490 return mImpl->getNbOutputMaps();
2506 mImpl->setStride(stride);
2518 return mImpl->getStride();
2538 mImpl->setPadding(padding);
2552 return mImpl->getPadding();
2572 mImpl->setNbGroups(nbGroups);
2582 return mImpl->getNbGroups();
2596 mImpl->setKernelWeights(weights);
2606 return mImpl->getKernelWeights();
2621 mImpl->setBiasWeights(weights);
2631 return mImpl->getBiasWeights();
2649 mImpl->setPrePadding(padding);
2659 return mImpl->getPrePadding();
2677 mImpl->setPostPadding(padding);
2687 return mImpl->getPostPadding();
2701 mImpl->setPaddingMode(paddingMode);
2713 return mImpl->getPaddingMode();
2728 mImpl->setKernelSizeNd(kernelSize);
2738 return mImpl->getKernelSizeNd();
2755 mImpl->setStrideNd(stride);
2765 return mImpl->getStrideNd();
2783 mImpl->setPaddingNd(padding);
2795 return mImpl->getPaddingNd();
2819 mImpl->setDilationNd(dilation);
2829 return mImpl->getDilationNd();
2878 static constexpr int32_t kVALUE = 14;
2915 return mImpl->setOperation(op);
2927 return mImpl->getOperation();
3050 mImpl->setGatherAxis(axis);
3061 return mImpl->getGatherAxis();
3082 mImpl->setNbElementWiseDims(elementWiseDims);
3092 return mImpl->getNbElementWiseDims();
3102 mImpl->setMode(mode);
3112 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();
4216 mImpl->setStart(start);
4231 return mImpl->getStart();
4245 return mImpl->setSize(size);
4260 return mImpl->getSize();
4274 mImpl->setStride(stride);
4289 return mImpl->getStride();
4299 mImpl->setMode(mode);
4309 return mImpl->getMode();
4398 mImpl->setOperation(op);
4408 return mImpl->getOperation();
4430 return mImpl->getK();
4440 mImpl->setReduceAxes(reduceAxes);
4450 return mImpl->getReduceAxes();
4534 mImpl->setOperation(index, op);
4546 return mImpl->getOperation(index);
4659 mImpl->setWeights(weights);
4669 return mImpl->getWeights();
4681 mImpl->setDimensions(dimensions);
4693 return mImpl->getDimensions();
4742 static constexpr int32_t kVALUE = 3;
4796 static constexpr int32_t kVALUE = 3;
4826 static constexpr int32_t kVALUE = 2;
4862 static constexpr int32_t kVALUE = 4;
4925 return mImpl->setOutputDimensions(dimensions);
4935 return mImpl->getOutputDimensions();
4963 void setScales(
float const* scales, int32_t nbScales)
noexcept
4965 mImpl->setScales(scales, nbScales);
4982 int32_t
getScales(int32_t size,
float* scales)
const noexcept
4984 return mImpl->getScales(size, scales);
4996 mImpl->setResizeMode(resizeMode);
5006 return mImpl->getResizeMode();
5022 mImpl->setAlignCorners(alignCorners);
5034 return mImpl->getAlignCorners();
5069 mImpl->setCoordinateTransformation(coordTransform);
5079 return mImpl->getCoordinateTransformation();
5094 mImpl->setSelectorForSinglePixel(selector);
5104 return mImpl->getSelectorForSinglePixel();
5118 mImpl->setNearestRounding(value);
5128 return mImpl->getNearestRounding();
5150 mImpl->setCubicCoeff(A);
5160 return mImpl->getCubicCoeff();
5173 mImpl->setExcludeOutside(excludeFlag);
5183 return mImpl->getExcludeOutside();
5242 return mBoundary->getLoop();
5247 apiv::VLoopBoundaryLayer* mBoundary;
5261 return mBoundary->getConditional();
5266 apiv::VConditionalBoundaryLayer* mBoundary;
5339 return mImpl->setCondition(condition);
5355 return mImpl->addOutput(trueSubgraphOutput, falseSubgraphOutput);
5367 return mImpl->addInput(input);
5382 mImpl->setName(name);
5392 return mImpl->getName();
5451 return mImpl->getLoopOutput();
5468 mImpl->setAxis(axis);
5474 return mImpl->getAxis();
5509 return mImpl->getTripLimit();
5523 mImpl->setAxis(axis);
5529 return mImpl->getAxis();
5539 mImpl->setReverse(reverse);
5545 return mImpl->getReverse();
5569 return mImpl->addRecurrence(initialValue);
5590 return mImpl->addTripLimit(tensor, limit);
5603 return mImpl->addIterator(tensor, axis, reverse);
5615 return mImpl->addLoopOutput(tensor, outputKind, axis);
5630 mImpl->setName(name);
5640 return mImpl->getName();
5685 mImpl->setMessage(message);
5695 return mImpl->getMessage();
5768 mImpl->setDimensions(dimensions);
5783 return mImpl->getDimensions();
5793 mImpl->setOperation(op);
5803 return mImpl->getOperation();
5822 mImpl->setAlpha(alpha);
5837 return mImpl->getAlpha();
5856 mImpl->setBeta(beta);
5871 return mImpl->getBeta();
5983 return mImpl->getAxis();
5994 mImpl->setAxis(axis);
6069 return mImpl->getAxis();
6080 mImpl->setAxis(axis);
6137 return mImpl->setEquation(equation);
6147 return mImpl->getEquation();
6243 mImpl->setMode(mode);
6253 return mImpl->getMode();
6263 mImpl->setAxis(axis);
6271 return mImpl->getAxis();
6316 mImpl->setAxis(axis);
6324 return mImpl->getAxis();
6351 mImpl->setInterpolationMode(mode);
6363 return mImpl->getInterpolationMode();
6373 mImpl->setAlignCorners(alignCorners);
6385 return mImpl->getAlignCorners();
6397 return mImpl->setSampleMode(mode);
6409 return mImpl->getSampleMode();
6498 mImpl->setBoundingBoxFormat(fmt);
6510 return mImpl->getBoundingBoxFormat();
6524 mImpl->setTopKBoxLimit(limit);
6534 return mImpl->getTopKBoxLimit();
6625 return mImpl->addInput(name, type, dimensions);
6639 mImpl->markOutput(tensor);
6663 return mImpl->addConvolution(input, nbOutputMaps, kernelSize, kernelWeights, biasWeights);
6686 return mImpl->addFullyConnected(input, nbOutputs, kernelWeights, biasWeights);
6705 return mImpl->addActivation(input, type);
6724 return mImpl->addPooling(input, type, windowSize);
6743 return mImpl->addLRN(input, window, alpha, beta, k);
6770 return mImpl->addScale(input, mode, shift, scale, power);
6783 return mImpl->addSoftMax(input);
6800 return mImpl->addConcatenation(inputs, nbInputs);
6824 return mImpl->addDeconvolution(input, nbOutputMaps, kernelSize, kernelWeights, biasWeights);
6851 return mImpl->addElementWise(input1, input2, op);
6873 return mImpl->addUnary(input, operation);
6890 return mImpl->addPadding(input, prePadding, postPadding);
6904 return mImpl->addShuffle(input);
6921 return mImpl->addOneHot(indices, values, depth, axis);
6933 return mImpl->getNbLayers();
6947 return mImpl->getLayer(index);
6959 return mImpl->getNbInputs();
6975 return mImpl->getInput(index);
6989 return mImpl->getNbOutputs();
7005 return mImpl->getOutput(index);
7045 return mImpl->addReduce(input, operation, reduceAxes, keepDimensions);
7078 return mImpl->addTopK(input, op, k, reduceAxes);
7094 return mImpl->addGather(data, indices, axis);
7110 return mImpl->addGatherV2(data, indices, mode);
7128 return mImpl->addRaggedSoftMax(input, bounds);
7150 return mImpl->addMatrixMultiply(input0, op0, input1, op1);
7164 return mImpl->addNonZero(input);
7191 return mImpl->addConstant(dimensions, weights);
7259 ITensor& input, int32_t layerCount, int32_t hiddenSize, int32_t maxSeqLen,
RNNOperation op)
noexcept
7261 return mImpl->addRNNv2(input, layerCount, hiddenSize, maxSeqLen, op);
7275 return mImpl->addIdentity(input);
7290 mImpl->removeTensor(tensor);
7302 mImpl->unmarkOutput(tensor);
7321 return mImpl->addPluginV2(inputs, nbInputs, plugin);
7340 return mImpl->addSlice(input, start, size, stride);
7364 mImpl->setName(name);
7378 return mImpl->getName();
7396 return mImpl->addShape(input);
7414 return mImpl->hasImplicitBatchDimension();
7432 return mImpl->markOutputForShapes(tensor);
7444 return mImpl->unmarkOutputForShapes(tensor);
7462 return mImpl->addParametricReLU(input, slope);
7485 return mImpl->addConvolutionNd(input, nbOutputMaps, kernelSize, kernelWeights, biasWeights);
7504 return mImpl->addPoolingNd(input, type, windowSize);
7527 return mImpl->addDeconvolutionNd(input, nbOutputMaps, kernelSize, kernelWeights, biasWeights);
7563 return mImpl->addScaleNd(input, mode, shift, scale, power, channelAxis);
7578 return mImpl->addResize(input);
7592 return mImpl->hasExplicitPrecision();
7608 return mImpl->addLoop();
7648 return mImpl->addSelect(condition, thenInput, elseInput);
7665 return mImpl->addAssertion(condition, message);
7688 return mImpl->addFill(dimensions, op);
7705 return mImpl->addPaddingNd(input, prePadding, postPadding);
7728 return mImpl->setWeightsName(weights, name);
7747 mImpl->setErrorRecorder(recorder);
7762 return mImpl->getErrorRecorder();
7781 return mImpl->addDequantize(input, scale);
7801 return mImpl->addScatter(data, indices, updates, mode);
7820 return mImpl->addQuantize(input, scale);
7835 return mImpl->addIfConditional();
7849 return mImpl->addEinsum(inputs, nbInputs, equation);
7865 return mImpl->addGridSample(input, grid);
7883 return mImpl->addNMS(boxes, scores, maxOutputBoxesPerClass);
7948 virtual
bool getBatch(
void* bindings[],
char const* names[], int32_t nbBindings) noexcept = 0;
7964 virtual
void const* readCalibrationCache(std::
size_t& length) noexcept = 0;
7974 virtual
void writeCalibrationCache(
void const* ptr, std::
size_t length) noexcept = 0;
8068 virtual
double getRegressionCutoff() const noexcept = 0;
8082 virtual
void const* readHistogramCache(std::
size_t& length) noexcept = 0;
8092 virtual
void writeHistogramCache(
void const* ptr, std::
size_t length) noexcept = 0;
8115 return mImpl->getTensorFormat();
8123 return mImpl->getDataType();
8131 return mImpl->getStrides();
8158 return mImpl->getImplementation();
8166 return mImpl->getTactic();
8191 return mImpl->getName();
8202 return mImpl->getDimensions(index, select);
8210 return mImpl->getNbInputs();
8218 return mImpl->getNbOutputs();
8250 return mImpl->getAlgorithmIOInfo(index);
8258 return mImpl->getAlgorithmVariant();
8266 return mImpl->getTimingMSec();
8274 return mImpl->getWorkspaceSize();
8287 return mImpl->getAlgorithmIOInfoByIndex(index);
8321 int32_t nbChoices, int32_t* selection)
noexcept = 0;
8333 int32_t nbAlgorithms)
noexcept = 0;
8485 return mImpl->serialize();
8509 return mImpl->combine(inputCache, ignoreMismatch);
8519 return mImpl->reset();
8622 static constexpr int32_t kVALUE = 2;
8652 mImpl->setMinTimingIterations(minTiming);
8666 return mImpl->getMinTimingIterations();
8679 mImpl->setAvgTimingIterations(avgTiming);
8691 return mImpl->getAvgTimingIterations();
8704 mImpl->setEngineCapability(capability);
8716 return mImpl->getEngineCapability();
8726 mImpl->setInt8Calibrator(calibrator);
8734 return mImpl->getInt8Calibrator();
8749 mImpl->setMaxWorkspaceSize(workspaceSize);
8766 return mImpl->getMaxWorkspaceSize();
8783 mImpl->setFlags(builderFlags);
8795 return mImpl->getFlags();
8807 mImpl->clearFlag(builderFlag);
8819 mImpl->setFlag(builderFlag);
8831 return mImpl->getFlag(builderFlag);
8846 mImpl->setDeviceType(layer, deviceType);
8855 return mImpl->getDeviceType(layer);
8865 return mImpl->isDeviceTypeSet(layer);
8875 mImpl->resetDeviceType(layer);
8884 return mImpl->canRunOnDLA(layer);
8899 mImpl->setDLACore(dlaCore);
8908 return mImpl->getDLACore();
8918 mImpl->setDefaultDeviceType(deviceType);
8928 return mImpl->getDefaultDeviceType();
8964 return mImpl->setProfileStream(stream);
8976 return mImpl->getProfileStream();
8992 return mImpl->addOptimizationProfile(profile);
9005 return mImpl->getNbOptimizationProfiles();
9017 mImpl->setProfilingVerbosity(verbosity);
9030 return mImpl->getProfilingVerbosity();
9039 mImpl->setAlgorithmSelector(selector);
9047 return mImpl->getAlgorithmSelector();
9062 return mImpl->setCalibrationProfile(profile);
9072 return mImpl->getCalibrationProfile();
9089 mImpl->setQuantizationFlags(flags);
9101 return mImpl->getQuantizationFlags();
9113 mImpl->clearQuantizationFlag(flag);
9125 mImpl->setQuantizationFlag(flag);
9137 return mImpl->getQuantizationFlag(flag);
9159 return mImpl->setTacticSources(tacticSources);
9174 return mImpl->getTacticSources();
9193 return mImpl->createTimingCache(blob, size);
9216 return mImpl->setTimingCache(cache, ignoreMismatch);
9226 return mImpl->getTimingCache();
9258 mImpl->setMemoryPoolLimit(pool, poolSize);
9277 return mImpl->getMemoryPoolLimit(pool);
9295 mImpl->setPreviewFeature(feature, enable);
9309 return mImpl->getPreviewFeature(feature);
9381 mImpl->setMaxBatchSize(batchSize);
9396 return mImpl->getMaxBatchSize();
9404 return mImpl->platformHasFastFp16();
9412 return mImpl->platformHasFastInt8();
9436 return mImpl->getMaxDLABatchSize();
9444 return mImpl->getNbDLACores();
9460 mImpl->setGpuAllocator(allocator);
9470 return mImpl->createBuilderConfig();
9486 return mImpl->buildEngineWithConfig(network, config);
9503 return mImpl->createNetworkV2(flags);
9517 return mImpl->createOptimizationProfile();
9536 mImpl->setErrorRecorder(recorder);
9551 return mImpl->getErrorRecorder();
9567 return mImpl->platformHasTf32();
9586 return mImpl->buildSerializedNetwork(network, config);
9610 return mImpl->isNetworkSupported(network, config);
9620 return mImpl->getLogger();
9634 return mImpl->setMaxThreads(maxThreads);
9648 return mImpl->getMaxThreads();
9661extern "C" TENSORRTAPI void* createInferBuilder_INTERNAL(
void* logger, int32_t version)
noexcept;
9675inline IBuilder* createInferBuilder(ILogger& logger)
noexcept
#define TENSORRTAPI
Definition: NvInferRuntimeCommon.h:54
#define NV_TENSORRT_VERSION
Definition: NvInferRuntimeCommon.h:73
#define TRT_DEPRECATED
Definition: NvInferRuntimeCommon.h:40
#define TRT_DEPRECATED_ENUM
Definition: NvInferRuntimeCommon.h:41
Definition: NvInferRuntimeCommon.h:171
static constexpr int32_t MAX_DIMS
The maximum rank (number of dimensions) supported for a tensor.
Definition: NvInferRuntimeCommon.h:174
Descriptor for two-dimensional spatial data.
Definition: NvInferLegacyDims.h:70
An Activation layer in a network definition.
Definition: NvInfer.h:1593
void setBeta(float beta) noexcept
Set the beta parameter (must be finite).
Definition: NvInfer.h:1641
void setActivationType(ActivationType type) noexcept
Set the type of activation to be performed.
Definition: NvInfer.h:1602
ActivationType getActivationType() const noexcept
Get the type of activation to be performed.
Definition: NvInfer.h:1612
float getAlpha() const noexcept
Get the alpha parameter.
Definition: NvInfer.h:1650
virtual ~IActivationLayer() noexcept=default
float getBeta() const noexcept
Get the beta parameter.
Definition: NvInfer.h:1659
void setAlpha(float alpha) noexcept
Set the alpha parameter (must be finite).
Definition: NvInfer.h:1627
Describes the context and requirements, that could be fulfilled by one or more instances of IAlgorith...
Definition: NvInfer.h:8183
int32_t getNbOutputs() const noexcept
Return number of outputs of the algorithm.
Definition: NvInfer.h:8216
int32_t getNbInputs() const noexcept
Return number of inputs of the algorithm.
Definition: NvInfer.h:8208
char const * getName() const noexcept
Return name of the algorithm node. This is a unique identifier for the IAlgorithmContext.
Definition: NvInfer.h:8189
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:8200
Describes a variation of execution of a layer. An algorithm is represented by IAlgorithmVariant and t...
Definition: NvInfer.h:8236
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:8272
float getTimingMSec() const noexcept
The time in milliseconds to execute the algorithm.
Definition: NvInfer.h:8264
IAlgorithmIOInfo const * getAlgorithmIOInfoByIndex(int32_t index) const noexcept
Returns the format of an Algorithm input or output. Algorithm inputs are incrementally numbered first...
Definition: NvInfer.h:8285
virtual ~IAlgorithm() noexcept=default
TRT_DEPRECATED IAlgorithmIOInfo const & getAlgorithmIOInfo(int32_t index) const noexcept
Returns the format of an Algorithm input or output. Algorithm inputs are incrementally numbered first...
Definition: NvInfer.h:8248
IAlgorithmVariant const & getAlgorithmVariant() const noexcept
Returns the algorithm variant.
Definition: NvInfer.h:8256
Carries information about input or output of the algorithm. IAlgorithmIOInfo for all the input and ou...
Definition: NvInfer.h:8108
virtual ~IAlgorithmIOInfo() noexcept=default
Dims getStrides() const noexcept
Return strides of the input/output tensor of algorithm.
Definition: NvInfer.h:8129
DataType getDataType() const noexcept
Return DataType of the input/output of algorithm.
Definition: NvInfer.h:8121
TensorFormat getTensorFormat() const noexcept
Return TensorFormat of the input/output of algorithm.
Definition: NvInfer.h:8113
Interface implemented by application for selecting and reporting algorithms of a layer provided by th...
Definition: NvInfer.h:8304
virtual void reportAlgorithms(IAlgorithmContext const *const *algoContexts, IAlgorithm const *const *algoChoices, int32_t nbAlgorithms) noexcept=0
Called by TensorRT to report choices it made.
virtual int32_t selectAlgorithms(IAlgorithmContext const &context, IAlgorithm const *const *choices, int32_t nbChoices, int32_t *selection) noexcept=0
Select Algorithms for a layer from the given list of algorithm choices.
virtual ~IAlgorithmSelector() noexcept=default
provides a unique 128-bit identifier, which along with the input and output information denotes the v...
Definition: NvInfer.h:8151
virtual ~IAlgorithmVariant() noexcept=default
int64_t getTactic() const noexcept
Return tactic of the algorithm.
Definition: NvInfer.h:8164
int64_t getImplementation() const noexcept
Return implementation of the algorithm.
Definition: NvInfer.h:8156
An assertion layer in a network.
Definition: NvInfer.h:5673
void setMessage(char const *message) noexcept
Set the message to print if the assertion fails.
Definition: NvInfer.h:5683
char const * getMessage() const noexcept
Return the assertion message.
Definition: NvInfer.h:5693
virtual ~IAssertionLayer() noexcept=default
Holds properties for configuring a builder to produce an engine.
Definition: NvInfer.h:8634
virtual TRT_DEPRECATED int32_t getMinTimingIterations() const noexcept
Query the number of minimization iterations.
Definition: NvInfer.h:8664
IOptimizationProfile const * getCalibrationProfile() noexcept
Get the current calibration profile.
Definition: NvInfer.h:9070
void setMemoryPoolLimit(MemoryPoolType pool, std::size_t poolSize) noexcept
Set the memory size for the memory pool.
Definition: NvInfer.h:9256
void setQuantizationFlag(QuantizationFlag flag) noexcept
Set a single quantization flag.
Definition: NvInfer.h:9123
nvinfer1::ITimingCache * createTimingCache(void const *blob, std::size_t size) const noexcept
Create timing cache.
Definition: NvInfer.h:9191
void setPreviewFeature(PreviewFeature feature, bool enable) noexcept
Enable or disable a specific preview feature.
Definition: NvInfer.h:9293
void setInt8Calibrator(IInt8Calibrator *calibrator) noexcept
Set Int8 Calibration interface.
Definition: NvInfer.h:8724
bool getPreviewFeature(PreviewFeature feature) const noexcept
Get status of preview feature.
Definition: NvInfer.h:9307
void clearQuantizationFlag(QuantizationFlag flag) noexcept
clear a quantization flag.
Definition: NvInfer.h:9111
bool setTacticSources(TacticSources tacticSources) noexcept
Set tactic sources.
Definition: NvInfer.h:9157
bool getQuantizationFlag(QuantizationFlag flag) const noexcept
Returns true if the quantization flag is set.
Definition: NvInfer.h:9135
std::size_t getMemoryPoolLimit(MemoryPoolType pool) const noexcept
Get the memory size limit of the memory pool.
Definition: NvInfer.h:9275
int32_t getDLACore() const noexcept
Get the DLA core that the engine executes on.
Definition: NvInfer.h:8906
void setDeviceType(ILayer const *layer, DeviceType deviceType) noexcept
Set the device that this layer must execute on.
Definition: NvInfer.h:8844
void setEngineCapability(EngineCapability capability) noexcept
Configure the builder to target specified EngineCapability flow.
Definition: NvInfer.h:8702
bool getFlag(BuilderFlag builderFlag) const noexcept
Returns true if the build mode flag is set.
Definition: NvInfer.h:8829
void setQuantizationFlags(QuantizationFlags flags) noexcept
Set the quantization flags.
Definition: NvInfer.h:9087
bool setCalibrationProfile(IOptimizationProfile const *profile) noexcept
Add a calibration profile.
Definition: NvInfer.h:9060
virtual void setAvgTimingIterations(int32_t avgTiming) noexcept
Set the number of averaging iterations used when timing layers.
Definition: NvInfer.h:8677
void setProfilingVerbosity(ProfilingVerbosity verbosity) noexcept
Set verbosity level of layer information exposed in NVTX annotations and IEngineInspector.
Definition: NvInfer.h:9015
IAlgorithmSelector * getAlgorithmSelector() const noexcept
Get Algorithm Selector.
Definition: NvInfer.h:9045
int32_t getNbOptimizationProfiles() const noexcept
Get number of optimization profiles.
Definition: NvInfer.h:9003
QuantizationFlags getQuantizationFlags() const noexcept
Get the quantization flags.
Definition: NvInfer.h:9099
nvinfer1::ITimingCache const * getTimingCache() const noexcept
Get the pointer to the timing cache from current IBuilderConfig.
Definition: NvInfer.h:9224
void reset() noexcept
Resets the builder configuration to defaults.
Definition: NvInfer.h:8936
bool setTimingCache(ITimingCache const &cache, bool ignoreMismatch) noexcept
Attach a timing cache to IBuilderConfig.
Definition: NvInfer.h:9214
TRT_DEPRECATED void destroy() noexcept
Delete this IBuilderConfig.
Definition: NvInfer.h:8950
EngineCapability getEngineCapability() const noexcept
Query EngineCapability flow configured for the builder.
Definition: NvInfer.h:8714
void setAlgorithmSelector(IAlgorithmSelector *selector) noexcept
Set Algorithm Selector.
Definition: NvInfer.h:9037
TRT_DEPRECATED void setMaxWorkspaceSize(std::size_t workspaceSize) noexcept
Set the maximum workspace size.
Definition: NvInfer.h:8747
TRT_DEPRECATED std::size_t getMaxWorkspaceSize() const noexcept
Get the maximum workspace size.
Definition: NvInfer.h:8764
DeviceType getDefaultDeviceType() const noexcept
Get the default DeviceType which was set by setDefaultDeviceType.
Definition: NvInfer.h:8926
BuilderFlags getFlags() const noexcept
Get the build mode flags for this builder config. Defaults to 0.
Definition: NvInfer.h:8793
void setFlags(BuilderFlags builderFlags) noexcept
Set the build mode flags to turn on builder options for this network.
Definition: NvInfer.h:8781
TacticSources getTacticSources() const noexcept
Get tactic sources.
Definition: NvInfer.h:9172
void resetDeviceType(ILayer const *layer) noexcept
reset the DeviceType for this layer
Definition: NvInfer.h:8873
void setDLACore(int32_t dlaCore) noexcept
Sets the DLA core used by the network. Defaults to -1.
Definition: NvInfer.h:8897
void clearFlag(BuilderFlag builderFlag) noexcept
clear a single build mode flag.
Definition: NvInfer.h:8805
int32_t addOptimizationProfile(IOptimizationProfile const *profile) noexcept
Add an optimization profile.
Definition: NvInfer.h:8990
apiv::VBuilderConfig * mImpl
Definition: NvInfer.h:9313
int32_t getAvgTimingIterations() const noexcept
Query the number of averaging iterations.
Definition: NvInfer.h:8689
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:8916
void setFlag(BuilderFlag builderFlag) noexcept
Set a single build mode flag.
Definition: NvInfer.h:8817
virtual ~IBuilderConfig() noexcept=default
DeviceType getDeviceType(ILayer const *layer) const noexcept
Get the device that this layer executes on.
Definition: NvInfer.h:8853
bool canRunOnDLA(ILayer const *layer) const noexcept
Checks if a layer can run on DLA.
Definition: NvInfer.h:8882
cudaStream_t getProfileStream() const noexcept
Get the cuda stream that is used to profile this network.
Definition: NvInfer.h:8974
IInt8Calibrator * getInt8Calibrator() const noexcept
Get Int8 Calibration interface.
Definition: NvInfer.h:8732
ProfilingVerbosity getProfilingVerbosity() const noexcept
Get verbosity level of layer information exposed in NVTX annotations and IEngineInspector.
Definition: NvInfer.h:9028
bool isDeviceTypeSet(ILayer const *layer) const noexcept
whether the DeviceType has been explicitly set for this layer
Definition: NvInfer.h:8863
void setProfileStream(const cudaStream_t stream) noexcept
Set the cuda stream that is used to profile this network.
Definition: NvInfer.h:8962
Builds an engine from a network definition.
Definition: NvInfer.h:9365
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:9434
int32_t getNbDLACores() const noexcept
Return the number of DLA engines available to this builder.
Definition: NvInfer.h:9442
IErrorRecorder * getErrorRecorder() const noexcept
get the ErrorRecorder assigned to this interface.
Definition: NvInfer.h:9549
apiv::VBuilder * mImpl
Definition: NvInfer.h:9652
ILogger * getLogger() const noexcept
get the logger with which the builder was created
Definition: NvInfer.h:9618
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:9608
bool platformHasTf32() const noexcept
Determine whether the platform has TF32 support.
Definition: NvInfer.h:9565
int32_t getMaxThreads() const noexcept
get the maximum number of threads that can be used by the builder.
Definition: NvInfer.h:9646
TRT_DEPRECATED void destroy() noexcept
Destroy this object.
Definition: NvInfer.h:9422
nvinfer1::IOptimizationProfile * createOptimizationProfile() noexcept
Create a new optimization profile.
Definition: NvInfer.h:9515
bool platformHasFastFp16() const noexcept
Determine whether the platform has fast native fp16.
Definition: NvInfer.h:9402
void setGpuAllocator(IGpuAllocator *allocator) noexcept
Set the GPU allocator.
Definition: NvInfer.h:9458
nvinfer1::INetworkDefinition * createNetworkV2(NetworkDefinitionCreationFlags flags) noexcept
Create a network definition object.
Definition: NvInfer.h:9501
nvinfer1::IBuilderConfig * createBuilderConfig() noexcept
Create a builder configuration object.
Definition: NvInfer.h:9468
void reset() noexcept
Resets the builder state to default values.
Definition: NvInfer.h:9557
bool setMaxThreads(int32_t maxThreads) noexcept
Set the maximum number of threads.
Definition: NvInfer.h:9632
void setErrorRecorder(IErrorRecorder *recorder) noexcept
Set the ErrorRecorder for this interface.
Definition: NvInfer.h:9534
bool platformHasFastInt8() const noexcept
Determine whether the platform has fast native int8.
Definition: NvInfer.h:9410
nvinfer1::IHostMemory * buildSerializedNetwork(INetworkDefinition &network, IBuilderConfig &config) noexcept
Builds and serializes a network for the given INetworkDefinition and IBuilderConfig.
Definition: NvInfer.h:9584
virtual ~IBuilder() noexcept=default
TRT_DEPRECATED int32_t getMaxBatchSize() const noexcept
Get the maximum batch size.
Definition: NvInfer.h:9394
TRT_DEPRECATED nvinfer1::ICudaEngine * buildEngineWithConfig(INetworkDefinition &network, IBuilderConfig &config) noexcept
Builds an engine for the given INetworkDefinition and given IBuilderConfig.
Definition: NvInfer.h:9483
A concatenation layer in a network definition.
Definition: NvInfer.h:2399
void setAxis(int32_t axis) noexcept
Set the axis along which concatenation occurs.
Definition: NvInfer.h:2413
int32_t getAxis() const noexcept
Get the axis along which concatenation occurs.
Definition: NvInfer.h:2423
virtual ~IConcatenationLayer() noexcept=default
Definition: NvInfer.h:5273
virtual ~IConditionLayer() noexcept=default
Layer that represents a constant value.
Definition: NvInfer.h:4646
void setWeights(Weights weights) noexcept
Set the weights for the layer.
Definition: NvInfer.h:4657
Weights getWeights() const noexcept
Get the weights for the layer.
Definition: NvInfer.h:4667
apiv::VConstantLayer * mImpl
Definition: NvInfer.h:4697
void setDimensions(Dims dimensions) noexcept
Set the dimensions for the layer.
Definition: NvInfer.h:4679
virtual ~IConstantLayer() noexcept=default
Dims getDimensions() const noexcept
Get the dimensions for the layer.
Definition: NvInfer.h:4691
A convolution layer in a network definition.
Definition: NvInfer.h:1036
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:1139
TRT_DEPRECATED DimsHW getStride() const noexcept
Get the stride of the convolution.
Definition: NvInfer.h:1107
TRT_DEPRECATED DimsHW getDilation() const noexcept
Get the dilation for a convolution.
Definition: NvInfer.h:1246
void setStrideNd(Dims stride) noexcept
Set the multi-dimension stride of the convolution.
Definition: NvInfer.h:1364
void setKernelSizeNd(Dims kernelSize) noexcept
Set the multi-dimension kernel size of the convolution.
Definition: NvInfer.h:1339
Dims getPrePadding() const noexcept
Get the pre-padding.
Definition: NvInfer.h:1273
Weights getBiasWeights() const noexcept
Get the bias weights for the convolution.
Definition: NvInfer.h:1218
int32_t getNbGroups() const noexcept
Get the number of groups of the convolution.
Definition: NvInfer.h:1169
void setPaddingMode(PaddingMode paddingMode) noexcept
Set the padding mode.
Definition: NvInfer.h:1314
TRT_DEPRECATED void setStride(DimsHW stride) noexcept
Get the stride of the convolution.
Definition: NvInfer.h:1097
Dims getPaddingNd() const noexcept
Get the multi-dimension padding of the convolution.
Definition: NvInfer.h:1404
Dims getStrideNd() const noexcept
Get the multi-dimension stride of the convolution.
Definition: NvInfer.h:1374
TRT_DEPRECATED void setPadding(DimsHW padding) noexcept
Set the padding of the convolution.
Definition: NvInfer.h:1127
Weights getKernelWeights() const noexcept
Get the kernel weights of the convolution.
Definition: NvInfer.h:1193
void setPaddingNd(Dims padding) noexcept
Set the multi-dimension padding of the convolution.
Definition: NvInfer.h:1392
Dims getDilationNd() const noexcept
Get the multi-dimension dilation of the convolution.
Definition: NvInfer.h:1428
void setKernelWeights(Weights weights) noexcept
Set the kernel weights for the convolution.
Definition: NvInfer.h:1183
void setDilationNd(Dims dilation) noexcept
Set the multi-dimension dilation of the convolution.
Definition: NvInfer.h:1418
Dims getPostPadding() const noexcept
Get the post-padding.
Definition: NvInfer.h:1300
TRT_DEPRECATED void setDilation(DimsHW dilation) noexcept
Set the dilation for a convolution.
Definition: NvInfer.h:1234
void setNbGroups(int32_t nbGroups) noexcept
Set the number of groups for a convolution.
Definition: NvInfer.h:1159
PaddingMode getPaddingMode() const noexcept
Get the padding mode.
Definition: NvInfer.h:1326
int32_t getNbOutputMaps() const noexcept
Get the number of output maps for the convolution.
Definition: NvInfer.h:1081
void setPrePadding(Dims padding) noexcept
Set the multi-dimension pre-padding of the convolution.
Definition: NvInfer.h:1263
virtual ~IConvolutionLayer() noexcept=default
void setBiasWeights(Weights weights) noexcept
Set the bias weights for the convolution.
Definition: NvInfer.h:1208
Dims getKernelSizeNd() const noexcept
Get the multi-dimension kernel size of the convolution.
Definition: NvInfer.h:1349
void setNbOutputMaps(int32_t nbOutputMaps) noexcept
Set the number of output maps for the convolution.
Definition: NvInfer.h:1071
TRT_DEPRECATED void setKernelSize(DimsHW kernelSize) noexcept
Set the HW kernel size of the convolution.
Definition: NvInfer.h:1047
void setPostPadding(Dims padding) noexcept
Set the multi-dimension post-padding of the convolution.
Definition: NvInfer.h:1290
TRT_DEPRECATED DimsHW getKernelSize() const noexcept
Get the HW kernel size of the convolution.
Definition: NvInfer.h:1059
An engine for executing inference on a built network, with functionally unsafe features.
Definition: NvInferRuntime.h:1383
A deconvolution layer in a network definition.
Definition: NvInfer.h:2441
void setBiasWeights(Weights weights) noexcept
Set the bias weights for the deconvolution.
Definition: NvInfer.h:2619
TRT_DEPRECATED void setStride(DimsHW stride) noexcept
Set the stride of the deconvolution.
Definition: NvInfer.h:2504
void setNbOutputMaps(int32_t nbOutputMaps) noexcept
Set the number of output feature maps for the deconvolution.
Definition: NvInfer.h:2478
Weights getKernelWeights() const noexcept
Get the kernel weights for the deconvolution.
Definition: NvInfer.h:2604
void setPrePadding(Dims padding) noexcept
Set the multi-dimension pre-padding of the deconvolution.
Definition: NvInfer.h:2647
TRT_DEPRECATED DimsHW getPadding() const noexcept
Get the padding of the deconvolution.
Definition: NvInfer.h:2550
Dims getStrideNd() const noexcept
Get the multi-dimension stride of the deconvolution.
Definition: NvInfer.h:2763
int32_t getNbGroups() const noexcept
Get the number of groups for a deconvolution.
Definition: NvInfer.h:2580
Dims getDilationNd() const noexcept
Get the multi-dimension dilation of the deconvolution.
Definition: NvInfer.h:2827
Weights getBiasWeights() const noexcept
Get the bias weights for the deconvolution.
Definition: NvInfer.h:2629
void setKernelWeights(Weights weights) noexcept
Set the kernel weights for the deconvolution.
Definition: NvInfer.h:2594
void setDilationNd(Dims dilation) noexcept
Set the multi-dimension dilation of the deconvolution.
Definition: NvInfer.h:2817
TRT_DEPRECATED DimsHW getStride() const noexcept
Get the stride of the deconvolution.
Definition: NvInfer.h:2516
TRT_DEPRECATED void setPadding(DimsHW padding) noexcept
Set the padding of the deconvolution.
Definition: NvInfer.h:2536
Dims getPostPadding() const noexcept
Get the padding.
Definition: NvInfer.h:2685
Dims getKernelSizeNd() const noexcept
Get the multi-dimension kernel size of the deconvolution.
Definition: NvInfer.h:2736
void setKernelSizeNd(Dims kernelSize) noexcept
Set the multi-dimension kernel size of the deconvolution.
Definition: NvInfer.h:2726
virtual ~IDeconvolutionLayer() noexcept=default
TRT_DEPRECATED void setKernelSize(DimsHW kernelSize) noexcept
Set the HW kernel size of the convolution.
Definition: NvInfer.h:2454
void setStrideNd(Dims stride) noexcept
Set the multi-dimension stride of the deconvolution.
Definition: NvInfer.h:2753
Dims getPaddingNd() const noexcept
Get the multi-dimension padding of the deconvolution.
Definition: NvInfer.h:2793
void setNbGroups(int32_t nbGroups) noexcept
Set the number of groups for a deconvolution.
Definition: NvInfer.h:2570
void setPaddingMode(PaddingMode paddingMode) noexcept
Set the padding mode.
Definition: NvInfer.h:2699
void setPaddingNd(Dims padding) noexcept
Set the multi-dimension padding of the deconvolution.
Definition: NvInfer.h:2781
TRT_DEPRECATED DimsHW getKernelSize() const noexcept
Get the HW kernel size of the deconvolution.
Definition: NvInfer.h:2466
void setPostPadding(Dims padding) noexcept
Set the multi-dimension post-padding of the deconvolution.
Definition: NvInfer.h:2675
int32_t getNbOutputMaps() const noexcept
Get the number of output feature maps for the deconvolution.
Definition: NvInfer.h:2488
Dims getPrePadding() const noexcept
Get the pre-padding.
Definition: NvInfer.h:2657
PaddingMode getPaddingMode() const noexcept
Get the padding mode.
Definition: NvInfer.h:2711
A Dequantize layer in a network definition.
Definition: NvInfer.h:6057
virtual ~IDequantizeLayer() noexcept=default
int32_t getAxis() const noexcept
Get the quantization axis.
Definition: NvInfer.h:6067
void setAxis(int32_t axis) noexcept
Set the quantization axis.
Definition: NvInfer.h:6078
An Einsum layer in a network.
Definition: NvInfer.h:6124
bool setEquation(char const *equation) noexcept
Set the equation. The equation is a comma-separated list of subscript labels, where each label refers...
Definition: NvInfer.h:6135
virtual ~IEinsumLayer() noexcept=default
char const * getEquation() const noexcept
Return the equation.
Definition: NvInfer.h:6145
A elementwise layer in a network definition.
Definition: NvInfer.h:2902
virtual ~IElementWiseLayer() noexcept=default
apiv::VElementWiseLayer * mImpl
Definition: NvInfer.h:2931
ElementWiseOperation getOperation() const noexcept
Get the binary operation for the layer.
Definition: NvInfer.h:2925
void setOperation(ElementWiseOperation op) noexcept
Set the binary operation for the layer.
Definition: NvInfer.h:2913
Reference counted application-implemented error reporting interface for TensorRT objects.
Definition: NvInferRuntimeCommon.h:1689
Generate an output tensor with specified mode.
Definition: NvInfer.h:5755
FillOperation getOperation() const noexcept
Get the fill operation for the layer.
Definition: NvInfer.h:5801
void setOperation(FillOperation op) noexcept
Set the fill operation for the layer.
Definition: NvInfer.h:5791
void setDimensions(Dims dimensions) noexcept
Set the output tensor's dimensions.
Definition: NvInfer.h:5766
void setBeta(double beta) noexcept
Set the beta parameter.
Definition: NvInfer.h:5854
double getAlpha() const noexcept
Get the value of alpha parameter.
Definition: NvInfer.h:5835
void setAlpha(double alpha) noexcept
Set the alpha parameter.
Definition: NvInfer.h:5820
Dims getDimensions() const noexcept
Get the output tensor's dimensions.
Definition: NvInfer.h:5781
double getBeta() const noexcept
Get the value of beta parameter.
Definition: NvInfer.h:5869
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:1485
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:1494
Weights getKernelWeights() const noexcept
Get the kernel weights.
Definition: NvInfer.h:1524
void setBiasWeights(Weights weights) noexcept
Set the bias weights.
Definition: NvInfer.h:1536
void setKernelWeights(Weights weights) noexcept
Set the kernel weights, given as a KxC matrix in row-major order.
Definition: NvInfer.h:1514
int32_t getNbOutputChannels() const noexcept
Get the number of output channels K from the fully connected layer.
Definition: NvInfer.h:1504
Weights getBiasWeights() const noexcept
Get the bias weights.
Definition: NvInfer.h:1546
A Gather layer in a network definition. Supports several kinds of gathering.
Definition: NvInfer.h:3037
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:3048
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:3080
apiv::VGatherLayer * mImpl
Definition: NvInfer.h:3116
int32_t getNbElementWiseDims() const noexcept
Get the number of leading dimensions of indices tensor to be handled elementwise.
Definition: NvInfer.h:3090
void setMode(GatherMode mode) noexcept
Set the gather mode.
Definition: NvInfer.h:3100
int32_t getGatherAxis() const noexcept
Get the axis to gather on.
Definition: NvInfer.h:3059
GatherMode getMode() const noexcept
Get the gather mode.
Definition: NvInfer.h:3110
virtual ~IGatherLayer() noexcept=default
Application-implemented class for controlling allocation on the GPU.
Definition: NvInferRuntimeCommon.h:1362
A GridSample layer in a network definition.
Definition: NvInfer.h:6342
void setInterpolationMode(InterpolationMode mode) noexcept
Set the grid sample interpolation mode.
Definition: NvInfer.h:6349
bool setSampleMode(SampleMode mode) noexcept
Set the sample mode.
Definition: NvInfer.h:6395
void setAlignCorners(bool alignCorners) noexcept
Set the align corners mode.
Definition: NvInfer.h:6371
apiv::VGridSampleLayer * mImpl
Definition: NvInfer.h:6413
SampleMode getSampleMode() const noexcept
Get the sample mode.
Definition: NvInfer.h:6407
InterpolationMode getInterpolationMode() const noexcept
Get the grid sample interpolation mode.
Definition: NvInfer.h:6361
bool getAlignCorners() const noexcept
Get the align corners mode.
Definition: NvInfer.h:6383
virtual ~IGridSampleLayer() noexcept=default
Class to handle library allocated memory that is accessible to the user.
Definition: NvInferRuntime.h:144
A layer that represents the identity function.
Definition: NvInfer.h:4631
apiv::VIdentityLayer * mImpl
Definition: NvInfer.h:4633
virtual ~IIdentityLayer() noexcept=default
Definition: NvInfer.h:5256
IIfConditional * getConditional() const noexcept
Return pointer to the IIfConditional associated with this boundary layer.
Definition: NvInfer.h:5259
virtual ~IIfConditionalBoundaryLayer() noexcept=default
Definition: NvInfer.h:5326
IIfConditionalInputLayer * addInput(ITensor &input) noexcept
Add an If-conditional input.
Definition: NvInfer.h:5365
char const * getName() const noexcept
Return the name of the conditional.
Definition: NvInfer.h:5390
virtual ~IIfConditional() noexcept=default
IConditionLayer * setCondition(ITensor &condition) noexcept
Set the condition tensor for this If-Conditional construct.
Definition: NvInfer.h:5337
IIfConditionalOutputLayer * addOutput(ITensor &trueSubgraphOutput, ITensor &falseSubgraphOutput) noexcept
Add an If-conditional output.
Definition: NvInfer.h:5353
void setName(char const *name) noexcept
Set the name of the conditional.
Definition: NvInfer.h:5380
Definition: NvInfer.h:5286
virtual ~IIfConditionalOutputLayer() noexcept=default
Application-implemented interface for calibration.
Definition: NvInfer.h:7926
virtual int32_t getBatchSize() const noexcept=0
Get the batch size used for calibration batches.
Definition: NvInfer.h:8009
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:8014
virtual ~IInt8EntropyCalibrator2() noexcept=default
Definition: NvInfer.h:7991
virtual ~IInt8EntropyCalibrator() noexcept=default
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:7996
Definition: NvInfer.h:8044
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:8049
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:8026
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:8031
virtual ~IInt8MinMaxCalibrator() noexcept=default
Definition: NvInfer.h:5518
virtual ~IIteratorLayer() noexcept=default
void setReverse(bool reverse) noexcept
Definition: NvInfer.h:5537
bool getReverse() const noexcept
True if and only if reversing input.
Definition: NvInfer.h:5543
int32_t getAxis() const noexcept
Get axis being iterated over.
Definition: NvInfer.h:5527
void setAxis(int32_t axis) noexcept
Set axis to iterate over.
Definition: NvInfer.h:5521
A LRN layer in a network definition.
Definition: NvInfer.h:2045
int32_t getWindowSize() const noexcept
Get the LRN window size.
Definition: NvInfer.h:2066
float getAlpha() const noexcept
Get the LRN alpha value.
Definition: NvInfer.h:2087
void setWindowSize(int32_t windowSize) noexcept
Set the LRN window size.
Definition: NvInfer.h:2056
void setK(float k) noexcept
Set the LRN K value.
Definition: NvInfer.h:2119
void setAlpha(float alpha) noexcept
Set the LRN alpha value.
Definition: NvInfer.h:2077
void setBeta(float beta) noexcept
Set the LRN beta value.
Definition: NvInfer.h:2098
virtual ~ILRNLayer() noexcept=default
float getBeta() const noexcept
Get the LRN beta value.
Definition: NvInfer.h:2108
float getK() const noexcept
Get the LRN K value.
Definition: NvInfer.h:2129
Base class for all layer classes in a network definition.
Definition: NvInfer.h:540
bool precisionIsSet() const noexcept
whether the computational precision has been set for this layer
Definition: NvInfer.h:680
void setPrecision(DataType dataType) noexcept
Set the computational precision of this layer.
Definition: NvInfer.h:656
void setName(char const *name) noexcept
Set the name of a layer.
Definition: NvInfer.h:561
void resetPrecision() noexcept
reset the computational precision for this layer
Definition: NvInfer.h:690
int32_t getNbInputs() const noexcept
Get the number of inputs of a layer.
Definition: NvInfer.h:579
DataType getOutputType(int32_t index) const noexcept
get the output type of this layer
Definition: NvInfer.h:742
DataType getPrecision() const noexcept
get the computational precision of this layer
Definition: NvInfer.h:668
char const * getName() const noexcept
Return the name of a layer.
Definition: NvInfer.h:571
int32_t getNbOutputs() const noexcept
Get the number of outputs of a layer.
Definition: NvInfer.h:600
bool outputTypeIsSet(int32_t index) const noexcept
whether the output type has been set for this layer
Definition: NvInfer.h:755
ITensor * getOutput(int32_t index) const noexcept
Get the layer output corresponding to the given index.
Definition: NvInfer.h:611
void setInput(int32_t index, ITensor &tensor) noexcept
Replace an input of this layer with a specific tensor.
Definition: NvInfer.h:628
void resetOutputType(int32_t index) noexcept
reset the output type for this layer
Definition: NvInfer.h:767
ITensor * getInput(int32_t index) const noexcept
Get the layer input corresponding to the given index.
Definition: NvInfer.h:592
void setOutputType(int32_t index, DataType dataType) noexcept
Set the output type of this layer.
Definition: NvInfer.h:728
LayerType getType() const noexcept
Return the type of a layer.
Definition: NvInfer.h:547
virtual ~ILayer() noexcept=default
Application-implemented logging interface for the builder, refitter and runtime.
Definition: NvInferRuntimeCommon.h:1500
Definition: NvInfer.h:5237
virtual ~ILoopBoundaryLayer() noexcept=default
ILoop * getLoop() const noexcept
Return pointer to ILoop associated with this boundary layer.
Definition: NvInfer.h:5240
Definition: NvInfer.h:5559
void setName(char const *name) noexcept
Set the name of the loop.
Definition: NvInfer.h:5628
ITripLimitLayer * addTripLimit(ITensor &tensor, TripLimit limit) noexcept
Add a trip-count limiter, based on the given tensor.
Definition: NvInfer.h:5588
IIteratorLayer * addIterator(ITensor &tensor, int32_t axis=0, bool reverse=false) noexcept
Return layer that subscripts tensor by loop iteration.
Definition: NvInfer.h:5601
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:5613
virtual ~ILoop() noexcept=default
char const * getName() const noexcept
Return the name of the loop.
Definition: NvInfer.h:5638
IRecurrenceLayer * addRecurrence(ITensor &initialValue) noexcept
Create a recurrence layer for this loop with initialValue as its first input.
Definition: NvInfer.h:5567
Definition: NvInfer.h:5447
virtual ~ILoopOutputLayer() noexcept=default
int32_t getAxis() const noexcept
Get axis being concatenated over.
Definition: NvInfer.h:5472
LoopOutput getLoopOutput() const noexcept
Definition: NvInfer.h:5449
void setAxis(int32_t axis) noexcept
Set where to insert the contenation axis. Ignored if getLoopOutput() is kLAST_VALUE.
Definition: NvInfer.h:5466
Layer that represents a Matrix Multiplication.
Definition: NvInfer.h:4524
apiv::VMatrixMultiplyLayer * mImpl
Definition: NvInfer.h:4550
virtual ~IMatrixMultiplyLayer() noexcept=default
MatrixOperation getOperation(int32_t index) const noexcept
Get the operation for an input tensor.
Definition: NvInfer.h:4544
void setOperation(int32_t index, MatrixOperation op) noexcept
Set the operation for an input tensor.
Definition: NvInfer.h:4532
A non-maximum suppression layer in a network definition.
Definition: NvInfer.h:6485
virtual ~INMSLayer() noexcept=default
void setTopKBoxLimit(int32_t limit) noexcept
Set the TopK box limit parameter for the layer.
Definition: NvInfer.h:6522
void setBoundingBoxFormat(BoundingBoxFormat fmt) noexcept
Set the bounding box format parameter for the layer.
Definition: NvInfer.h:6496
BoundingBoxFormat getBoundingBoxFormat() const noexcept
Get the bounding box format parameter for the layer.
Definition: NvInfer.h:6508
apiv::VNMSLayer * mImpl
Definition: NvInfer.h:6558
int32_t getTopKBoxLimit() const noexcept
Get the TopK box limit parameter for the layer.
Definition: NvInfer.h:6532
A network definition for input to the builder.
Definition: NvInfer.h:6582
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:7319
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:7482
IDequantizeLayer * addDequantize(ITensor &input, ITensor &scale) noexcept
Add a dequantization layer to the network.
Definition: NvInfer.h:7779
IConcatenationLayer * addConcatenation(ITensor *const *inputs, int32_t nbInputs) noexcept
Add a concatenation layer to the network.
Definition: NvInfer.h:6798
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:6821
IShuffleLayer * addShuffle(ITensor &input) noexcept
Add a shuffle layer to the network.
Definition: NvInfer.h:6902
void setName(char const *name) noexcept
Sets the name of the network.
Definition: NvInfer.h:7362
ILRNLayer * addLRN(ITensor &input, int32_t window, float alpha, float beta, float k) noexcept
Add a LRN layer to the network.
Definition: NvInfer.h:6741
ITopKLayer * addTopK(ITensor &input, TopKOperation op, int32_t k, uint32_t reduceAxes) noexcept
Add a TopK layer to the network.
Definition: NvInfer.h:7076
IAssertionLayer * addAssertion(ITensor &condition, char const *message) noexcept
Add an assertion layer to the network.
Definition: NvInfer.h:7663
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:7560
char const * getName() const noexcept
Returns the name associated with the network.
Definition: NvInfer.h:7376
TRT_DEPRECATED bool hasExplicitPrecision() const noexcept
True if network is an explicit precision network.
Definition: NvInfer.h:7590
IParametricReLULayer * addParametricReLU(ITensor &input, ITensor &slope) noexcept
Add a parametric ReLU layer to the network.
Definition: NvInfer.h:7460
ITensor * getOutput(int32_t index) const noexcept
Get the output tensor specified by the given index.
Definition: NvInfer.h:7003
ITensor * getInput(int32_t index) const noexcept
Get the input tensor specified by the given index.
Definition: NvInfer.h:6973
bool unmarkOutputForShapes(ITensor &tensor) noexcept
Undo markOutputForShapes.
Definition: NvInfer.h:7442
IFillLayer * addFill(Dims dimensions, FillOperation op) noexcept
Add a fill layer to the network.
Definition: NvInfer.h:7686
ILoop * addLoop() noexcept
Add a loop to the network.
Definition: NvInfer.h:7606
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:7524
IActivationLayer * addActivation(ITensor &input, ActivationType type) noexcept
Add an activation layer to the network.
Definition: NvInfer.h:6703
virtual ~INetworkDefinition() noexcept=default
INMSLayer * addNMS(ITensor &boxes, ITensor &scores, ITensor &maxOutputBoxesPerClass) noexcept
Add a non-maximum suppression layer to the network.
Definition: NvInfer.h:7881
ILayer * getLayer(int32_t index) const noexcept
Get the layer specified by the given index.
Definition: NvInfer.h:6945
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:7258
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:6683
IIfConditional * addIfConditional() noexcept
Add an If-conditional layer to the network.
Definition: NvInfer.h:7833
IErrorRecorder * getErrorRecorder() const noexcept
get the ErrorRecorder assigned to this interface.
Definition: NvInfer.h:7760
int32_t getNbInputs() const noexcept
Get the number of inputs in the network.
Definition: NvInfer.h:6957
bool hasImplicitBatchDimension() const noexcept
Query whether the network was created with an implicit batch dimension.
Definition: NvInfer.h:7412
IReduceLayer * addReduce(ITensor &input, ReduceOperation operation, uint32_t reduceAxes, bool keepDimensions) noexcept
Add a reduce layer to the network.
Definition: NvInfer.h:7042
IUnaryLayer * addUnary(ITensor &input, UnaryOperation operation) noexcept
Add a unary layer to the network.
Definition: NvInfer.h:6871
IGridSampleLayer * addGridSample(ITensor &input, ITensor &grid) noexcept
Add a GridSample layer to the network.
Definition: NvInfer.h:7863
void removeTensor(ITensor &tensor) noexcept
remove a tensor from the network definition.
Definition: NvInfer.h:7288
ISelectLayer * addSelect(ITensor &condition, ITensor &thenInput, ITensor &elseInput) noexcept
Add a select layer to the network.
Definition: NvInfer.h:7646
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:7799
int32_t getNbLayers() const noexcept
Get the number of layers in the network.
Definition: NvInfer.h:6931
apiv::VNetworkDefinition * mImpl
Definition: NvInfer.h:7887
IPoolingLayer * addPoolingNd(ITensor &input, PoolingType type, Dims windowSize) noexcept
Add a multi-dimension pooling layer to the network.
Definition: NvInfer.h:7502
bool markOutputForShapes(ITensor &tensor) noexcept
Enable tensor's value to be computed by IExecutionContext::getShapeBinding.
Definition: NvInfer.h:7430
IOneHotLayer * addOneHot(ITensor &indices, ITensor &values, ITensor &depth, int32_t axis) noexcept
Add a OneHot layer to the network.
Definition: NvInfer.h:6919
TRT_DEPRECATED IPaddingLayer * addPadding(ITensor &input, DimsHW prePadding, DimsHW postPadding) noexcept
Add a padding layer to the network.
Definition: NvInfer.h:6888
IScaleLayer * addScale(ITensor &input, ScaleMode mode, Weights shift, Weights scale, Weights power) noexcept
Add a Scale layer to the network.
Definition: NvInfer.h:6768
void unmarkOutput(ITensor &tensor) noexcept
unmark a tensor as a network output.
Definition: NvInfer.h:7300
IIdentityLayer * addIdentity(ITensor &input) noexcept
Add an identity layer.
Definition: NvInfer.h:7273
IGatherLayer * addGatherV2(ITensor &data, ITensor &indices, GatherMode mode) noexcept
Add gather with specified mode, axis=0 and nbElementWiseDims=0.
Definition: NvInfer.h:7108
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:6660
IQuantizeLayer * addQuantize(ITensor &input, ITensor &scale) noexcept
Add a quantization layer to the network.
Definition: NvInfer.h:7818
IElementWiseLayer * addElementWise(ITensor &input1, ITensor &input2, ElementWiseOperation op) noexcept
Add an elementwise layer to the network.
Definition: NvInfer.h:6849
void setErrorRecorder(IErrorRecorder *recorder) noexcept
Set the ErrorRecorder for this interface.
Definition: NvInfer.h:7745
ISliceLayer * addSlice(ITensor &input, Dims start, Dims size, Dims stride) noexcept
Add a slice layer to the network.
Definition: NvInfer.h:7338
IConstantLayer * addConstant(Dims dimensions, Weights weights) noexcept
Add a constant layer to the network.
Definition: NvInfer.h:7189
IRaggedSoftMaxLayer * addRaggedSoftMax(ITensor &input, ITensor &bounds) noexcept
Add a RaggedSoftMax layer to the network.
Definition: NvInfer.h:7126
IShapeLayer * addShape(ITensor &input) noexcept
Add a shape layer to the network.
Definition: NvInfer.h:7394
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:7092
TRT_DEPRECATED IPoolingLayer * addPooling(ITensor &input, PoolingType type, DimsHW windowSize) noexcept
Add a pooling layer to the network.
Definition: NvInfer.h:6722
TRT_DEPRECATED void destroy() noexcept
Destroy this INetworkDefinition object.
Definition: NvInfer.h:7015
IResizeLayer * addResize(ITensor &input) noexcept
Add a resize layer to the network.
Definition: NvInfer.h:7576
IMatrixMultiplyLayer * addMatrixMultiply(ITensor &input0, MatrixOperation op0, ITensor &input1, MatrixOperation op1) noexcept
Add a MatrixMultiply layer to the network.
Definition: NvInfer.h:7147
ISoftMaxLayer * addSoftMax(ITensor &input) noexcept
Add a SoftMax layer to the network.
Definition: NvInfer.h:6781
IEinsumLayer * addEinsum(ITensor *const *inputs, int32_t nbInputs, char const *equation) noexcept
Add an Einsum layer to the network.
Definition: NvInfer.h:7847
void markOutput(ITensor &tensor) noexcept
Mark a tensor as a network output.
Definition: NvInfer.h:6637
INonZeroLayer * addNonZero(ITensor &input) noexcept
Add a nonzero layer to the network.
Definition: NvInfer.h:7162
int32_t getNbOutputs() const noexcept
Get the number of outputs in the network.
Definition: NvInfer.h:6987
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:7703
bool setWeightsName(Weights weights, char const *name) noexcept
Associate a name with all current uses of the given weights.
Definition: NvInfer.h:7726
Forward declaration of IEngineInspector for use by other interfaces.
Definition: NvInferRuntime.h:43
Definition: NvInfer.h:4576
virtual ~INonZeroLayer() noexcept=default
A OneHot layer in a network definition.
Definition: NvInfer.h:6307
apiv::VOneHotLayer * mImpl
Definition: NvInfer.h:6328
void setAxis(int32_t axis) noexcept
Set the axis parameter.
Definition: NvInfer.h:6314
int32_t getAxis() const noexcept
Get the value of the axis parameter.
Definition: NvInfer.h:6322
Optimization profile for dynamic input dimensions and shape tensors.
Definition: NvInferRuntime.h:1133
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:4711
apiv::VParametricReLULayer * mImpl
Definition: NvInfer.h:4713
virtual ~IParametricReLULayer() noexcept=default
Single registration point for all plugins in an application. It is used to find plugin implementation...
Definition: NvInferRuntimeCommon.h:1234
Plugin class for user-implemented layers.
Definition: NvInferRuntimeCommon.h:394
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:1707
TRT_DEPRECATED DimsHW getStride() const noexcept
Get the stride for pooling.
Definition: NvInfer.h:1780
PoolingType getPoolingType() const noexcept
Get the type of activation to be performed.
Definition: NvInfer.h:1726
void setWindowSizeNd(Dims windowSize) noexcept
Set the multi-dimension window size for pooling.
Definition: NvInfer.h:1959
PaddingMode getPaddingMode() const noexcept
Get the padding mode.
Definition: NvInfer.h:1946
Dims getPostPadding() const noexcept
Get the padding.
Definition: NvInfer.h:1922
TRT_DEPRECATED void setStride(DimsHW stride) noexcept
Set the stride for pooling.
Definition: NvInfer.h:1768
bool getAverageCountExcludesPadding() const noexcept
Get whether average pooling uses as a denominator the overlap area between the window and the unpadde...
Definition: NvInfer.h:1866
Dims getPrePadding() const noexcept
Get the pre-padding.
Definition: NvInfer.h:1894
void setPoolingType(PoolingType type) noexcept
Set the type of activation to be performed.
Definition: NvInfer.h:1716
void setPaddingMode(PaddingMode paddingMode) noexcept
Set the padding mode.
Definition: NvInfer.h:1935
TRT_DEPRECATED void setWindowSize(DimsHW windowSize) noexcept
Set the window size for pooling.
Definition: NvInfer.h:1740
Dims getWindowSizeNd() const noexcept
Get the multi-dimension window size for pooling.
Definition: NvInfer.h:1969
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:1855
void setPaddingNd(Dims padding) noexcept
Set the multi-dimension padding for pooling.
Definition: NvInfer.h:2013
void setPrePadding(Dims padding) noexcept
Set the multi-dimension pre-padding for pooling.
Definition: NvInfer.h:1884
float getBlendFactor() const noexcept
Get the blending factor for the max_average_blend mode: max_average_blendPool = (1-blendFactor)*maxPo...
Definition: NvInfer.h:1838
Dims getStrideNd() const noexcept
Get the multi-dimension stride for pooling.
Definition: NvInfer.h:1994
virtual ~IPoolingLayer() noexcept=default
Dims getPaddingNd() const noexcept
Get the multi-dimension padding for pooling.
Definition: NvInfer.h:2025
TRT_DEPRECATED DimsHW getWindowSize() const noexcept
Get the window size for pooling.
Definition: NvInfer.h:1752
TRT_DEPRECATED DimsHW getPadding() const noexcept
Get the padding for pooling.
Definition: NvInfer.h:1810
void setStrideNd(Dims stride) noexcept
Set the multi-dimension stride for pooling.
Definition: NvInfer.h:1984
void setPostPadding(Dims padding) noexcept
Set the multi-dimension post-padding for pooling.
Definition: NvInfer.h:1912
TRT_DEPRECATED void setPadding(DimsHW padding) noexcept
Set the padding for pooling.
Definition: NvInfer.h:1796
void setBlendFactor(float blendFactor) noexcept
Set the blending factor for the max_average_blend mode: max_average_blendPool = (1-blendFactor)*maxPo...
Definition: NvInfer.h:1825
A Quantize layer in a network definition.
Definition: NvInfer.h:5971
void setAxis(int32_t axis) noexcept
Set the quantization axis.
Definition: NvInfer.h:5992
int32_t getAxis() const noexcept
Get the quantization axis.
Definition: NvInfer.h:5981
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:4597
apiv::VRaggedSoftMaxLayer * mImpl
Definition: NvInfer.h:4599
virtual ~IRaggedSoftMaxLayer() noexcept=default
Definition: NvInfer.h:5402
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:4903
void setSelectorForSinglePixel(ResizeSelector selector) noexcept
Set coordinate selector function when resized to single pixel.
Definition: NvInfer.h:5092
void setOutputDimensions(Dims dimensions) noexcept
Set the output dimensions.
Definition: NvInfer.h:4923
void setNearestRounding(ResizeRoundMode value) noexcept
Set rounding mode for nearest neighbor resize.
Definition: NvInfer.h:5116
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:4982
void setCubicCoeff(float A) noexcept
Set the coefficient 'A' used in cubic interpolation.
Definition: NvInfer.h:5148
ResizeMode getResizeMode() const noexcept
Get resize mode for an input tensor.
Definition: NvInfer.h:5004
void setScales(float const *scales, int32_t nbScales) noexcept
Set the resize scales.
Definition: NvInfer.h:4963
void setResizeMode(ResizeMode resizeMode) noexcept
Set resize mode for an input tensor.
Definition: NvInfer.h:4994
float getCubicCoeff() const noexcept
Get the coefficient 'A' used in cubic interpolation.
Definition: NvInfer.h:5158
ResizeSelector getSelectorForSinglePixel() const noexcept
Get the coordinate selector function when resized to single pixel.
Definition: NvInfer.h:5102
TRT_DEPRECATED bool getAlignCorners() const noexcept
True if align corners has been set.
Definition: NvInfer.h:5032
void setCoordinateTransformation(ResizeCoordinateTransformation coordTransform) noexcept
Set coordinate transformation function.
Definition: NvInfer.h:5067
void setExcludeOutside(bool excludeFlag) noexcept
Set the state for excluding outside pixels.
Definition: NvInfer.h:5171
Dims getOutputDimensions() const noexcept
Get the output dimensions.
Definition: NvInfer.h:4933
ResizeRoundMode getNearestRounding() const noexcept
Get rounding mode for nearest neighbor resize.
Definition: NvInfer.h:5126
bool getExcludeOutside() const noexcept
Get the state for excluding outside pixels.
Definition: NvInfer.h:5181
TRT_DEPRECATED void setAlignCorners(bool alignCorners) noexcept
Set whether to align corners while resizing.
Definition: NvInfer.h:5020
ResizeCoordinateTransformation getCoordinateTransformation() const noexcept
Get coordinate transformation function.
Definition: NvInfer.h:5077
A Scale layer in a network definition.
Definition: NvInfer.h:2189
Weights getScale() const noexcept
Get the scale value.
Definition: NvInfer.h:2246
Weights getPower() const noexcept
Get the power value.
Definition: NvInfer.h:2266
void setScale(Weights scale) noexcept
Set the scale value.
Definition: NvInfer.h:2236
void setPower(Weights power) noexcept
Set the power value.
Definition: NvInfer.h:2256
ScaleMode getMode() const noexcept
Get the scale mode.
Definition: NvInfer.h:2206
void setShift(Weights shift) noexcept
Set the shift value.
Definition: NvInfer.h:2216
void setChannelAxis(int32_t channelAxis) noexcept
Set the channel axis.
Definition: NvInfer.h:2302
Weights getShift() const noexcept
Get the shift value.
Definition: NvInfer.h:2226
virtual ~IScaleLayer() noexcept=default
void setMode(ScaleMode mode) noexcept
Set the scale mode.
Definition: NvInfer.h:2196
int32_t getChannelAxis() const noexcept
Get the channel axis.
Definition: NvInfer.h:2281
A scatter layer in a network definition. Supports several kinds of scattering.
Definition: NvInfer.h:6234
void setMode(ScatterMode mode) noexcept
Set the scatter mode.
Definition: NvInfer.h:6241
apiv::VScatterLayer * mImpl
Definition: NvInfer.h:6275
void setAxis(int32_t axis) noexcept
Set the axis used by ScatterMode::kELEMENTS.
Definition: NvInfer.h:6261
int32_t getAxis() const noexcept
Get the axis.
Definition: NvInfer.h:6269
ScatterMode getMode() const noexcept
Get the scatter mode.
Definition: NvInfer.h:6251
virtual ~IScatterLayer() noexcept=default
Definition: NvInfer.h:5652
virtual ~ISelectLayer() noexcept=default
Layer type for getting shape of a tensor.
Definition: NvInfer.h:4353
virtual ~IShapeLayer() noexcept=default
apiv::VShapeLayer * mImpl
Definition: NvInfer.h:4355
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:4203
void setMode(SliceMode mode) noexcept
Set the slice mode.
Definition: NvInfer.h:4297
apiv::VSliceLayer * mImpl
Definition: NvInfer.h:4336
virtual ~ISliceLayer() noexcept=default
void setStride(Dims stride) noexcept
Set the stride for computing the output slice data.
Definition: NvInfer.h:4272
void setStart(Dims start) noexcept
Set the start offset that the slice layer uses to create the output slice.
Definition: NvInfer.h:4214
Dims getStart() const noexcept
Get the start offset for the slice layer.
Definition: NvInfer.h:4229
void setSize(Dims size) noexcept
Set the dimensions of the output slice.
Definition: NvInfer.h:4243
Dims getSize() const noexcept
Get dimensions of the output slice.
Definition: NvInfer.h:4258
SliceMode getMode() const noexcept
Get the slice mode.
Definition: NvInfer.h:4307
Dims getStride() const noexcept
Get the stride for the output slice.
Definition: NvInfer.h:4287
A Softmax layer in a network definition.
Definition: NvInfer.h:2334
void setAxes(uint32_t axes) noexcept
Set the axis along which softmax is computed. Currently, only one axis can be set.
Definition: NvInfer.h:2366
uint32_t getAxes() const noexcept
Get the axis along which softmax occurs.
Definition: NvInfer.h:2376
virtual ~ISoftMaxLayer() noexcept=default
A tensor in a network definition.
Definition: NvInfer.h:172
bool setDynamicRange(float min, float max) noexcept
Set dynamic range for the tensor.
Definition: NvInfer.h:274
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:410
TensorLocation getLocation() const noexcept
Get the storage location of a tensor.
Definition: NvInfer.h:338
void resetDynamicRange() noexcept
Undo effect of setDynamicRange.
Definition: NvInfer.h:371
void setName(char const *name) noexcept
Set the tensor name.
Definition: NvInfer.h:188
bool isExecutionTensor() const noexcept
Whether the tensor is an execution tensor.
Definition: NvInfer.h:481
void setType(DataType type) noexcept
Set the data type of a tensor.
Definition: NvInfer.h:247
bool dynamicRangeIsSet() const noexcept
Query whether dynamic range is set.
Definition: NvInfer.h:363
void setLocation(TensorLocation location) noexcept
Set the storage location of a tensor.
Definition: NvInfer.h:353
char const * getName() const noexcept
Get the tensor name.
Definition: NvInfer.h:200
bool isShapeTensor() const noexcept
Whether the tensor is a shape tensor.
Definition: NvInfer.h:458
float getDynamicRangeMax() const noexcept
Get maximum of dynamic range.
Definition: NvInfer.h:391
bool isNetworkInput() const noexcept
Whether the tensor is a network input.
Definition: NvInfer.h:282
bool isNetworkOutput() const noexcept
Whether the tensor is a network output.
Definition: NvInfer.h:290
bool getBroadcastAcrossBatch() const noexcept
Check if tensor is broadcast across the batch.
Definition: NvInfer.h:328
void setBroadcastAcrossBatch(bool broadcastAcrossBatch) noexcept
Set whether to enable broadcast of tensor across the batch.
Definition: NvInfer.h:312
DataType getType() const noexcept
Get the data type of a tensor.
Definition: NvInfer.h:259
apiv::VTensor * mImpl
Definition: NvInfer.h:528
float getDynamicRangeMin() const noexcept
Get minimum of dynamic range.
Definition: NvInfer.h:381
virtual ~ITensor() noexcept=default
void setDimensionName(int32_t index, char const *name) noexcept
Name a dimension of an input tensor.
Definition: NvInfer.h:507
char const * getDimensionName(int32_t index) const noexcept
Get the name of an input dimension.
Definition: NvInfer.h:522
void setDimensions(Dims dimensions) noexcept
Set the dimensions of a tensor.
Definition: NvInfer.h:219
Dims getDimensions() const noexcept
Get the dimensions of a tensor.
Definition: NvInfer.h:232
TensorFormats getAllowedFormats() const noexcept
Get a bitmask of TensorFormat values that the tensor supports. For a shape tensor,...
Definition: NvInfer.h:423
Class to handle tactic timing info collected from builder.
Definition: NvInfer.h:8470
bool combine(ITimingCache const &inputCache, bool ignoreMismatch) noexcept
Combine input timing cache into local instance.
Definition: NvInfer.h:8507
virtual ~ITimingCache() noexcept=default
apiv::VTimingCache * mImpl
Definition: NvInfer.h:8523
bool reset() noexcept
Empty the timing cache.
Definition: NvInfer.h:8517
Layer that represents a TopK reduction.
Definition: NvInfer.h:4389
void setK(int32_t k) noexcept
Set the k value for the layer.
Definition: NvInfer.h:4418
void setReduceAxes(uint32_t reduceAxes) noexcept
Set which axes to reduce for the layer.
Definition: NvInfer.h:4438
TopKOperation getOperation() const noexcept
Get the operation for the layer.
Definition: NvInfer.h:4406
apiv::VTopKLayer * mImpl
Definition: NvInfer.h:4454
void setOperation(TopKOperation op) noexcept
Set the operation for the layer.
Definition: NvInfer.h:4396
int32_t getK() const noexcept
Get the k value for the layer.
Definition: NvInfer.h:4428
uint32_t getReduceAxes() const noexcept
Get the axes to reduce for the layer.
Definition: NvInfer.h:4448
virtual ~ITopKLayer() noexcept=default
Definition: NvInfer.h:5505
TripLimit getTripLimit() const noexcept
Definition: NvInfer.h:5507
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:126
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:1345
ResizeSelector
The coordinate selector when resize to single pixel output.
Definition: NvInfer.h:4808
@ kFORMULA
Use formula to map the original index.
@ kUPPER
Select the upper left pixel.
EngineCapability
List of supported engine capability flows.
Definition: NvInferRuntime.h:69
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:8534
ScaleMode
Controls how shift, scale and power are applied in a Scale layer.
Definition: NvInfer.h:2145
@ 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:8344
constexpr int32_t EnumMax< RNNDirection >() noexcept
Definition: NvInfer.h:3237
BoundingBoxFormat
Representation of bounding box data used for the Boxes input tensor in INMSLayer.
Definition: NvInfer.h:6423
@ kCENTER_SIZES
(x_center, y_center, width, height) where (x_center, y_center) is the center point of the box
@ kCORNER_PAIRS
(x1, y1, x2, y2) where (x1, y1) and (x2, y2) are any pair of diagonal corners
constexpr int32_t EnumMax< BuilderFlag >() noexcept
Definition: NvInfer.h:8454
constexpr int32_t EnumMax< LayerType >() noexcept
Definition: NvInfer.h:107
constexpr int32_t EnumMax< RNNGateType >() noexcept
Definition: NvInfer.h:3298
constexpr int32_t EnumMax< CalibrationAlgoType >() noexcept
Definition: NvInfer.h:7909
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:5229
constexpr int32_t EnumMax< RNNInputMode >() noexcept
Definition: NvInfer.h:3269
RNNInputMode
Enumerates the RNN input modes that may occur with an RNN layer.
Definition: NvInfer.h:3258
@ 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:126
@ 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:5712
@ kLINSPACE
Generate evenly spaced numbers over a specified interval.
@ kRANDOM_UNIFORM
Generate a tensor with random values drawn from a uniform distribution.
@ kRANDOM_NORMAL
Generate a tensor with random values drawn from a normal distribution.
ResizeRoundMode
The rounding mode for nearest neighbor resize.
Definition: NvInfer.h:4838
@ kHALF_DOWN
Round half down.
RNNGateType
Identifies an individual gate within an RNN cell.
Definition: NvInfer.h:3282
@ 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:1000
@ 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:5217
@ 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:9322
PreviewFeature
Define preview features.
Definition: NvInfer.h:8587
@ kDISABLE_EXTERNAL_TACTIC_SOURCES_FOR_CORE_0805
@ kFASTER_DYNAMIC_SHAPES_0805
constexpr int32_t EnumMax< GatherMode >() noexcept
Definition: NvInfer.h:2953
DataType
The type of weights and tensors.
Definition: NvInferRuntimeCommon.h:117
uint32_t BuilderFlags
Represents one or more QuantizationFlag values using binary OR operations, e.g., 1U << BuilderFlag::k...
Definition: NvInfer.h:8378
DeviceType
The device that this layer/network will execute on.
Definition: NvInferRuntime.h:598
constexpr int32_t EnumMax< ScaleMode >() noexcept
Definition: NvInfer.h:2157
CalibrationAlgoType
Version of calibration algorithm to use.
Definition: NvInfer.h:7896
LayerType
The type values of layer classes.
Definition: NvInfer.h:53
@ kGRID_SAMPLE
Grid sample layer.
@ 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.
@ kNON_ZERO
NonZero layer.
constexpr int32_t EnumMax< QuantizationFlag >() noexcept
Definition: NvInfer.h:8367
SampleMode
Controls how ISliceLayer and IGridSample handle out-of-bounds coordinates.
Definition: NvInfer.h:4135
@ kCLAMP
Out of bounds indices are clamped to bounds.
@ kSTRICT_BOUNDS
Fail with error when the coordinates are out of bounds.
@ kWRAP
Coordinates wrap around periodically.
GatherMode
Control form of IGatherLayer.
Definition: NvInfer.h:2941
@ 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:118
ProfilingVerbosity
List of verbosity levels of layer information exposed in NVTX annotations and in IEngineInspector.
Definition: NvInferRuntime.h:1357
NetworkDefinitionCreationFlag
List of immutable network properties expressed at network creation time. NetworkDefinitionCreationFla...
Definition: NvInfer.h:9333
ElementWiseOperation
Enumerates the binary operations that may be performed by an ElementWise layer.
Definition: NvInfer.h:2851
@ 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:8354
@ kCALIBRATE_BEFORE_FUSION
constexpr int32_t EnumMax< SampleMode >() noexcept
Definition: NvInfer.h:4155
RNNDirection
Enumerates the RNN direction that may be performed by an RNN layer.
Definition: NvInfer.h:3226
@ kBIDIRECTION
Network iterates from first to last and vice versa and outputs concatenated.
@ kUNIDIRECTION
Network iterations from first input to last input.
InterpolationMode
Enumerates various modes of interpolation.
Definition: NvInfer.h:4723
@ kNEAREST
ND (0 < N <= 8) nearest neighbor resizing.
@ kCUBIC
Supports bicubic (2D) interpolation.
BuilderFlag
List of valid modes that the builder can enable when creating an engine from a network definition.
Definition: NvInfer.h:8388
@ kENABLE_TACTIC_HEURISTIC
@ 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:4376
TensorFormat
Format of the input/output tensors.
Definition: NvInferRuntimeCommon.h:201
constexpr int32_t EnumMax< MemoryPoolType >() noexcept
Definition: NvInfer.h:8573
TopKOperation
Enumerates the operations that may be performed by a TopK layer.
Definition: NvInfer.h:4365
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:5210
RNNOperation
Enumerates the RNN operations that may be performed by an RNN layer.
Definition: NvInfer.h:3200
@ 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:9352
ScatterMode
Control form of IScatterLayer.
Definition: NvInfer.h:6161
MatrixOperation
Enumerates the operations that may be performed on a tensor by IMatrixMultiplyLayer before multiplica...
Definition: NvInfer.h:4465
@ kTRANSPOSE
Like kNONE, but transpose the matrix dimensions.
ResizeCoordinateTransformation
The resize coordinate transformation function.
Definition: NvInfer.h:4754
constexpr int32_t EnumMax< UnaryOperation >() noexcept
Definition: NvInfer.h:3669
LoopOutput
Enum that describes kinds of loop outputs.
Definition: NvInfer.h:5193
@ kLAST_VALUE
Output value is value of tensor for last iteration.
@ kCONCATENATE
Output value is concatenation of values of tensor for each iteration, in forward order.
@ kREVERSE
Output value is concatenation of values of tensor for each iteration, in reverse order.
constexpr int32_t EnumMax< BoundingBoxFormat >() noexcept
Definition: NvInfer.h:6436
constexpr int32_t EnumMax< MatrixOperation >() noexcept
Definition: NvInfer.h:4493
constexpr int32_t EnumMax< RNNOperation >() noexcept
Definition: NvInfer.h:3213
PoolingType
The type of pooling to perform in a pooling layer.
Definition: NvInfer.h:1675
constexpr int32_t EnumMax< FillOperation >() noexcept
Definition: NvInfer.h:5724
TensorLocation
The location for tensor data storage, device or host.
Definition: NvInferRuntime.h:216
OptProfileSelector
When setting or querying optimization profile parameters (such as shape tensor inputs or dynamic dime...
Definition: NvInferRuntime.h:1093
constexpr int32_t EnumMax< ScatterMode >() noexcept
Definition: NvInfer.h:6172
Definition: NvInfer.h:3947
Declaration of EnumMaxImpl struct to store maximum number of elements in an enumeration type.
Definition: NvInferRuntimeCommon.h:102