Go to the documentation of this file.
170 static constexpr int32_t kVALUE = 12;
204 mImpl->setName(name);
216 return mImpl->getName();
235 mImpl->setDimensions(dimensions);
248 return mImpl->getDimensions();
263 mImpl->setType(type);
275 return mImpl->getType();
290 return mImpl->setDynamicRange(min, max);
298 return mImpl->isNetworkInput();
306 return mImpl->isNetworkOutput();
328 mImpl->setBroadcastAcrossBatch(broadcastAcrossBatch);
344 return mImpl->getBroadcastAcrossBatch();
354 return mImpl->getLocation();
369 mImpl->setLocation(location);
379 return mImpl->dynamicRangeIsSet();
387 mImpl->resetDynamicRange();
397 return mImpl->getDynamicRangeMin();
407 return mImpl->getDynamicRangeMax();
426 mImpl->setAllowedFormats(formats);
439 return mImpl->getAllowedFormats();
473 return mImpl->isShapeTensor();
496 return mImpl->isExecutionTensor();
501 virtual ~
ITensor() noexcept = default;
521 return mLayer->getType();
533 mLayer->setName(name);
544 return mLayer->getName();
552 return mLayer->getNbInputs();
565 return mLayer->getInput(index);
573 return mLayer->getNbOutputs();
584 return mLayer->getOutput(index);
601 return mLayer->setInput(index, tensor);
621 mLayer->setPrecision(dataType);
633 return mLayer->getPrecision();
645 return mLayer->precisionIsSet();
655 mLayer->resetPrecision();
686 mLayer->setOutputType(index, dataType);
700 return mLayer->getOutputType(index);
713 return mLayer->outputTypeIsSet(index);
725 return mLayer->resetOutputType(index);
729 virtual ~
ILayer() noexcept = default;
730 apiv::VLayer* mLayer;
971 static constexpr int32_t kVALUE = 6;
1001 mImpl->setKernelSize(kernelSize);
1013 return mImpl->getKernelSize();
1025 mImpl->setNbOutputMaps(nbOutputMaps);
1035 return mImpl->getNbOutputMaps();
1051 mImpl->setStride(stride);
1061 return mImpl->getStride();
1081 return mImpl->setPadding(padding);
1093 return mImpl->getPadding();
1113 mImpl->setNbGroups(nbGroups);
1123 return mImpl->getNbGroups();
1137 mImpl->setKernelWeights(weights);
1147 return mImpl->getKernelWeights();
1162 mImpl->setBiasWeights(weights);
1172 return mImpl->getBiasWeights();
1188 return mImpl->setDilation(dilation);
1200 return mImpl->getDilation();
1217 mImpl->setPrePadding(padding);
1227 return mImpl->getPrePadding();
1244 mImpl->setPostPadding(padding);
1254 return mImpl->getPostPadding();
1268 mImpl->setPaddingMode(paddingMode);
1280 return mImpl->getPaddingMode();
1293 mImpl->setKernelSizeNd(kernelSize);
1303 return mImpl->getKernelSizeNd();
1318 mImpl->setStrideNd(stride);
1328 return mImpl->getStrideNd();
1346 mImpl->setPaddingNd(padding);
1358 return mImpl->getPaddingNd();
1372 mImpl->setDilationNd(dilation);
1382 return mImpl->getDilationNd();
1407 apiv::VConvolutionLayer* mImpl;
1451 mImpl->setNbOutputChannels(nbOutputs);
1461 return mImpl->getNbOutputChannels();
1471 mImpl->setKernelWeights(weights);
1481 return mImpl->getKernelWeights();
1493 mImpl->setBiasWeights(weights);
1503 return mImpl->getBiasWeights();
1526 apiv::VFullyConnectedLayer* mImpl;
1552 mImpl->setActivationType(type);
1562 return mImpl->getActivationType();
1577 mImpl->setAlpha(alpha);
1591 mImpl->setBeta(beta);
1600 return mImpl->getAlpha();
1609 return mImpl->getBeta();
1614 apiv::VActivationLayer* mImpl;
1626 kMAX_AVERAGE_BLEND = 2
1635 static constexpr int32_t kVALUE = 3;
1662 mImpl->setPoolingType(type);
1672 return mImpl->getPoolingType();
1686 mImpl->setWindowSize(windowSize);
1698 return mImpl->getWindowSize();
1714 mImpl->setStride(stride);
1726 return mImpl->getStride();
1742 mImpl->setPadding(padding);
1756 return mImpl->getPadding();
1771 mImpl->setBlendFactor(blendFactor);
1784 return mImpl->getBlendFactor();
1801 mImpl->setAverageCountExcludesPadding(exclusive);
1812 return mImpl->getAverageCountExcludesPadding();
1830 mImpl->setPrePadding(padding);
1840 return mImpl->getPrePadding();
1858 mImpl->setPostPadding(padding);
1868 return mImpl->getPostPadding();
1881 mImpl->setPaddingMode(paddingMode);
1892 return mImpl->getPaddingMode();
1905 mImpl->setWindowSizeNd(windowSize);
1915 return mImpl->getWindowSizeNd();
1930 mImpl->setStrideNd(stride);
1940 return mImpl->getStrideNd();
1959 mImpl->setPaddingNd(padding);
1971 return mImpl->getPaddingNd();
1976 apiv::VPoolingLayer* mImpl;
2002 mImpl->setWindowSize(windowSize);
2012 return mImpl->getWindowSize();
2023 mImpl->setAlpha(alpha);
2033 return mImpl->getAlpha();
2044 mImpl->setBeta(beta);
2054 return mImpl->getBeta();
2075 return mImpl->getK();
2079 virtual ~
ILRNLayer() noexcept = default;
2080 apiv::VLRNLayer* mImpl;
2138 mImpl->setMode(mode);
2148 return mImpl->getMode();
2158 mImpl->setShift(shift);
2168 return mImpl->getShift();
2178 mImpl->setScale(scale);
2188 return mImpl->getScale();
2198 mImpl->setPower(power);
2208 return mImpl->getPower();
2223 return mImpl->getChannelAxis();
2244 mImpl->setChannelAxis(channelAxis);
2249 apiv::VScaleLayer* mImpl;
2298 mImpl->setAxes(axes);
2308 return mImpl->getAxes();
2313 apiv::VSoftMaxLayer* mImpl;
2343 mImpl->setAxis(axis);
2353 return mImpl->getAxis();
2358 apiv::VConcatenationLayer* mImpl;
2384 mImpl->setKernelSize(kernelSize);
2396 return mImpl->getKernelSize();
2408 mImpl->setNbOutputMaps(nbOutputMaps);
2418 return mImpl->getNbOutputMaps();
2434 mImpl->setStride(stride);
2446 return mImpl->getStride();
2466 mImpl->setPadding(padding);
2480 return mImpl->getPadding();
2500 mImpl->setNbGroups(nbGroups);
2510 return mImpl->getNbGroups();
2524 mImpl->setKernelWeights(weights);
2534 return mImpl->getKernelWeights();
2549 mImpl->setBiasWeights(weights);
2559 return mImpl->getBiasWeights();
2577 mImpl->setPrePadding(padding);
2587 return mImpl->getPrePadding();
2605 mImpl->setPostPadding(padding);
2615 return mImpl->getPostPadding();
2629 mImpl->setPaddingMode(paddingMode);
2641 return mImpl->getPaddingMode();
2654 mImpl->setKernelSizeNd(kernelSize);
2664 return mImpl->getKernelSizeNd();
2679 mImpl->setStrideNd(stride);
2689 return mImpl->getStrideNd();
2707 mImpl->setPaddingNd(padding);
2719 return mImpl->getPaddingNd();
2749 mImpl->setDilationNd(dilation);
2759 return mImpl->getDilationNd();
2764 apiv::VDeconvolutionLayer* mImpl;
2798 static constexpr int32_t kVALUE = 14;
2835 return mImpl->setOperation(op);
2847 return mImpl->getOperation();
2869 mImpl->setGatherAxis(axis);
2879 return mImpl->getGatherAxis();
2891 mImpl->setNbElementWiseDims(k);
2901 return mImpl->getNbElementWiseDims();
3091 return mImpl->getLayerCount();
3095 return mImpl->getHiddenSize();
3099 return mImpl->getMaxSeqLength();
3103 return mImpl->getDataLength();
3122 return mImpl->setSequenceLengths(seqLengths);
3134 return mImpl->getSequenceLengths();
3143 mImpl->setOperation(op);
3152 return mImpl->getOperation();
3161 mImpl->setInputMode(op);
3170 return mImpl->getInputMode();
3185 mImpl->setDirection(op);
3194 return mImpl->getDirection();
3253 mImpl->setWeightsForGate(layerIndex, gate, isW, weights);
3262 return mImpl->getWeightsForGate(layerIndex, gate, isW);
3287 mImpl->setBiasForGate(layerIndex, gate, isW, bias);
3296 return mImpl->getBiasForGate(layerIndex, gate, isW);
3313 mImpl->setHiddenState(hidden);
3322 return mImpl->getHiddenState();
3341 mImpl->setCellState(cell);
3350 return mImpl->getCellState();
3377 return mImpl->getPlugin();
3441 mImpl->setOperation(op);
3451 return mImpl->getOperation();
3510 mImpl->setOperation(op);
3520 return mImpl->getOperation();
3530 mImpl->setReduceAxes(reduceAxes);
3540 return mImpl->getReduceAxes();
3550 mImpl->setKeepDimensions(keepDimensions);
3560 return mImpl->getKeepDimensions();
3592 mImpl->setPrePadding(padding);
3604 return mImpl->getPrePadding();
3618 mImpl->setPostPadding(padding);
3630 return mImpl->getPostPadding();
3644 mImpl->setPrePaddingNd(padding);
3656 return mImpl->getPrePaddingNd();
3670 mImpl->setPostPaddingNd(padding);
3682 return mImpl->getPostPaddingNd();
3727 mImpl->setFirstTranspose(permutation);
3739 return mImpl->getFirstTranspose();
3764 mImpl->setReshapeDimensions(dimensions);
3777 return mImpl->getReshapeDimensions();
3824 mImpl->setSecondTranspose(permutation);
3836 return mImpl->getSecondTranspose();
3852 return mImpl->setZeroIsPlaceholder(zeroIsPlaceholder);
3865 return mImpl->getZeroIsPlaceholder();
3935 mImpl->setStart(start);
3950 return mImpl->getStart();
3964 return mImpl->setSize(size);
3979 return mImpl->getSize();
3993 mImpl->setStride(stride);
4008 return mImpl->getStride();
4018 mImpl->setMode(mode);
4028 return mImpl->getMode();
4110 mImpl->setOperation(op);
4120 return mImpl->getOperation();
4142 return mImpl->getK();
4152 mImpl->setReduceAxes(reduceAxes);
4162 return mImpl->getReduceAxes();
4241 mImpl->setOperation(index, op);
4251 return mImpl->getOperation(index);
4319 mImpl->setWeights(weights);
4329 return mImpl->getWeights();
4341 mImpl->setDimensions(dimensions);
4353 return mImpl->getDimensions();
4392 static constexpr int32_t kVALUE = 2;
4442 static constexpr int32_t kVALUE = 3;
4468 static constexpr int32_t kVALUE = 2;
4501 static constexpr int32_t kVALUE = 4;
4547 return mImpl->setOutputDimensions(dimensions);
4557 return mImpl->getOutputDimensions();
4578 void setScales(
const float* scales, int32_t nbScales) noexcept
4580 mImpl->setScales(scales, nbScales);
4597 int32_t
getScales(int32_t size,
float* scales)
const noexcept
4599 return mImpl->getScales(size, scales);
4611 mImpl->setResizeMode(resizeMode);
4621 return mImpl->getResizeMode();
4638 mImpl->setAlignCorners(alignCorners);
4651 return mImpl->getAlignCorners();
4686 mImpl->setCoordinateTransformation(coordTransform);
4696 return mImpl->getCoordinateTransformation();
4711 mImpl->setSelectorForSinglePixel(selector);
4721 return mImpl->getSelectorForSinglePixel();
4735 mImpl->setNearestRounding(value);
4745 return mImpl->getNearestRounding();
4750 apiv::VResizeLayer* mImpl;
4796 return mBoundary->getLoop();
4801 apiv::VLoopBoundaryLayer* mBoundary;
4854 return mImpl->getLoopOutput();
4871 mImpl->setAxis(axis);
4877 return mImpl->getAxis();
4904 apiv::VLoopOutputLayer* mImpl;
4912 return mImpl->getTripLimit();
4926 mImpl->setAxis(axis);
4932 return mImpl->getAxis();
4942 mImpl->setReverse(reverse);
4948 return mImpl->getReverse();
4953 apiv::VIteratorLayer* mImpl;
4972 return mImpl->addRecurrence(initialValue);
4993 return mImpl->addTripLimit(tensor, limit);
5006 return mImpl->addIterator(tensor, axis, reverse);
5018 return mImpl->addLoopOutput(tensor, outputKind, axis);
5031 mImpl->setName(name);
5041 return mImpl->getName();
5045 virtual ~
ILoop() noexcept = default;
5118 mImpl->setDimensions(dimensions);
5133 return mImpl->getDimensions();
5143 mImpl->setOperation(op);
5153 return mImpl->getOperation();
5171 mImpl->setAlpha(alpha);
5186 return mImpl->getAlpha();
5204 mImpl->setBeta(beta);
5219 return mImpl->getBeta();
5252 apiv::VFillLayer* mImpl;
5325 return mImpl->getAxis();
5336 mImpl->setAxis(axis);
5341 apiv::VQuantizeLayer* mImpl;
5412 return mImpl->getAxis();
5423 mImpl->setAxis(axis);
5428 apiv::VDequantizeLayer* mImpl;
5492 return mImpl->addInput(name, type, dimensions);
5504 mImpl->markOutput(tensor);
5528 return mImpl->addConvolution(input, nbOutputMaps, kernelSize, kernelWeights, biasWeights);
5549 return mImpl->addFullyConnected(input, nbOutputs, kernelWeights, biasWeights);
5568 return mImpl->addActivation(input, type);
5587 return mImpl->addPooling(input, type, windowSize);
5606 return mImpl->addLRN(input, window, alpha, beta, k);
5633 return mImpl->addScale(input, mode, shift, scale, power);
5646 return mImpl->addSoftMax(input);
5663 return mImpl->addConcatenation(inputs, nbInputs);
5687 return mImpl->addDeconvolution(input, nbOutputMaps, kernelSize, kernelWeights, biasWeights);
5712 return mImpl->addElementWise(input1, input2, op);
5731 return mImpl->addUnary(input, operation);
5748 return mImpl->addPadding(input, prePadding, postPadding);
5762 return mImpl->addShuffle(input);
5774 return mImpl->getNbLayers();
5788 return mImpl->getLayer(index);
5800 return mImpl->getNbInputs();
5816 return mImpl->getInput(index);
5830 return mImpl->getNbOutputs();
5846 return mImpl->getOutput(index);
5887 return mImpl->addReduce(input, operation, reduceAxes, keepDimensions);
5920 return mImpl->addTopK(input, op, k, reduceAxes);
5936 return mImpl->addGather(data, indices, axis);
5954 return mImpl->addRaggedSoftMax(input, bounds);
5974 return mImpl->addMatrixMultiply(input0, op0, input1, op1);
5999 return mImpl->addConstant(dimensions, weights);
6066 ITensor& input, int32_t layerCount, int32_t hiddenSize, int32_t maxSeqLen,
RNNOperation op) noexcept
6068 return mImpl->addRNNv2(input, layerCount, hiddenSize, maxSeqLen, op);
6084 return mImpl->addIdentity(input);
6099 mImpl->removeTensor(tensor);
6111 mImpl->unmarkOutput(tensor);
6130 return mImpl->addPluginV2(inputs, nbInputs, plugin);
6149 return mImpl->addSlice(input, start, size, stride);
6171 mImpl->setName(name);
6185 return mImpl->getName();
6203 return mImpl->addShape(input);
6222 return mImpl->hasImplicitBatchDimension();
6240 return mImpl->markOutputForShapes(tensor);
6252 return mImpl->unmarkOutputForShapes(tensor);
6270 return mImpl->addParametricReLU(input, slope);
6293 return mImpl->addConvolutionNd(input, nbOutputMaps, kernelSize, kernelWeights, biasWeights);
6312 return mImpl->addPoolingNd(input, type, windowSize);
6335 return mImpl->addDeconvolutionNd(input, nbOutputMaps, kernelSize, kernelWeights, biasWeights);
6371 return mImpl->addScaleNd(input, mode, shift, scale, power, channelAxis);
6386 return mImpl->addResize(input);
6403 return mImpl->hasExplicitPrecision();
6419 return mImpl->addLoop();
6457 return mImpl->addSelect(condition, thenInput, elseInput);
6475 return mImpl->addFill(dimensions, op);
6490 return mImpl->addPaddingNd(input, prePadding, postPadding);
6510 return mImpl->setWeightsName(weights, name);
6529 mImpl->setErrorRecorder(recorder);
6544 return mImpl->getErrorRecorder();
6563 return mImpl->addDequantize(input, scale);
6582 return mImpl->addQuantize(input, scale);
6596 kLEGACY_CALIBRATION = 0,
6597 kENTROPY_CALIBRATION = 1,
6598 kENTROPY_CALIBRATION_2 = 2,
6599 kMINMAX_CALIBRATION = 3,
6628 virtual int32_t getBatchSize()
const noexcept = 0;
6643 virtual bool getBatch(
void* bindings[],
const char* names[], int32_t nbBindings) noexcept = 0;
6659 virtual const void* readCalibrationCache(std::size_t& length) noexcept = 0;
6669 virtual void writeCalibrationCache(
const void* ptr, std::size_t length) noexcept = 0;
6693 return CalibrationAlgoType::kENTROPY_CALIBRATION;
6711 return CalibrationAlgoType::kENTROPY_CALIBRATION_2;
6728 return CalibrationAlgoType::kMINMAX_CALIBRATION;
6746 return CalibrationAlgoType::kLEGACY_CALIBRATION;
6755 virtual double getQuantile() const noexcept = 0;
6763 virtual
double getRegressionCutoff() const noexcept = 0;
6777 virtual const
void* readHistogramCache(std::
size_t& length) noexcept = 0;
6787 virtual
void writeHistogramCache(const
void* ptr, std::
size_t length) noexcept = 0;
6810 return mImpl->getTensorFormat();
6818 return mImpl->getDataType();
6826 return mImpl->getStrides();
6831 apiv::VAlgorithmIOInfo* mImpl;
6853 return mImpl->getImplementation();
6861 return mImpl->getTactic();
6866 apiv::VAlgorithmVariant* mImpl;
6886 return mImpl->getName();
6897 return mImpl->getDimensions(index, select);
6905 return mImpl->getNbInputs();
6913 return mImpl->getNbOutputs();
6918 apiv::VAlgorithmContext* mImpl;
6945 return mImpl->getAlgorithmIOInfo(index);
6953 return mImpl->getAlgorithmVariant();
6961 return mImpl->getTimingMSec();
6969 return mImpl->getWorkspaceSize();
6982 return mImpl->getAlgorithmIOInfoByIndex(index);
6987 apiv::VAlgorithm* mImpl;
7015 int32_t nbChoices, int32_t* selection) noexcept
7028 int32_t nbAlgorithms) noexcept
7180 return mImpl->serialize();
7204 return mImpl->combine(inputCache, ignoreMismatch);
7214 return mImpl->reset();
7243 mImpl->setMinTimingIterations(minTiming);
7255 return mImpl->getMinTimingIterations();
7268 mImpl->setAvgTimingIterations(avgTiming);
7280 return mImpl->getAvgTimingIterations();
7293 mImpl->setEngineCapability(capability);
7305 return mImpl->getEngineCapability();
7315 mImpl->setInt8Calibrator(calibrator);
7323 return mImpl->getInt8Calibrator();
7335 mImpl->setMaxWorkspaceSize(workspaceSize);
7349 return mImpl->getMaxWorkspaceSize();
7366 mImpl->setFlags(builderFlags);
7378 return mImpl->getFlags();
7390 mImpl->clearFlag(builderFlag);
7402 mImpl->setFlag(builderFlag);
7414 return mImpl->getFlag(builderFlag);
7429 mImpl->setDeviceType(layer, deviceType);
7438 return mImpl->getDeviceType(layer);
7448 return mImpl->isDeviceTypeSet(layer);
7458 mImpl->resetDeviceType(layer);
7467 return mImpl->canRunOnDLA(layer);
7482 mImpl->setDLACore(dlaCore);
7493 return mImpl->getDLACore();
7503 mImpl->setDefaultDeviceType(deviceType);
7513 return mImpl->getDefaultDeviceType();
7549 return mImpl->setProfileStream(stream);
7561 return mImpl->getProfileStream();
7577 return mImpl->addOptimizationProfile(profile);
7590 return mImpl->getNbOptimizationProfiles();
7602 mImpl->setProfilingVerbosity(verbosity);
7615 return mImpl->getProfilingVerbosity();
7624 mImpl->setAlgorithmSelector(selector);
7632 return mImpl->getAlgorithmSelector();
7647 return mImpl->setCalibrationProfile(profile);
7657 return mImpl->getCalibrationProfile();
7674 mImpl->setQuantizationFlags(flags);
7686 return mImpl->getQuantizationFlags();
7698 mImpl->clearQuantizationFlag(flag);
7710 mImpl->setQuantizationFlag(flag);
7722 return mImpl->getQuantizationFlag(flag);
7747 return mImpl->setTacticSources(tacticSources);
7762 return mImpl->getTacticSources();
7781 return mImpl->createTimingCache(blob, size);
7804 return mImpl->setTimingCache(cache, ignoreMismatch);
7814 return mImpl->getTimingCache();
7878 virtual ~
IBuilder() noexcept =
default;
7890 mImpl->setMaxBatchSize(batchSize);
7903 return mImpl->getMaxBatchSize();
7911 return mImpl->platformHasFastFp16();
7919 return mImpl->platformHasFastInt8();
7943 return mImpl->getMaxDLABatchSize();
7951 return mImpl->getNbDLACores();
7967 mImpl->setGpuAllocator(allocator);
7977 return mImpl->createBuilderConfig();
7993 return mImpl->buildEngineWithConfig(network, config);
8009 return mImpl->createNetworkV2(flags);
8023 return mImpl->createOptimizationProfile();
8042 mImpl->setErrorRecorder(recorder);
8057 return mImpl->getErrorRecorder();
8073 return mImpl->platformHasTf32();
8092 return mImpl->buildSerializedNetwork(network, config);
8114 return mImpl->isNetworkSupported(network, config);
8127 extern "C" TENSORRTAPI
void* createInferBuilder_INTERNAL(
void* logger, int32_t version) noexcept;
8143 return static_cast<IBuilder*>(createInferBuilder_INTERNAL(&logger, NV_TENSORRT_VERSION));
8149 #endif // NV_INFER_H
Dims getStrideNd() const noexcept
Get the multi-dimension stride of the deconvolution.
Definition: NvInfer.h:2687
void removeTensor(ITensor &tensor) noexcept
remove a tensor from the network definition.
Definition: NvInfer.h:6097
TRT_DEPRECATED void setKernelSize(DimsHW kernelSize) noexcept
Set the HW kernel size of the convolution.
Definition: NvInfer.h:999
void setDimensions(Dims dimensions) noexcept
Set the output tensor's dimensions.
Definition: NvInfer.h:5116
LoopOutput
Enum that describes kinds of loop outputs.
Definition: NvInfer.h:4754
TRT_DEPRECATED DimsHW getPadding() const noexcept
Get the padding for pooling.
Definition: NvInfer.h:1754
float getAlpha() const noexcept
Get the alpha parameter.
Definition: NvInfer.h:1598
Dims getStrideNd() const noexcept
Get the multi-dimension stride of the convolution.
Definition: NvInfer.h:1326
IErrorRecorder * getErrorRecorder() const noexcept
get the ErrorRecorder assigned to this interface.
Definition: NvInfer.h:6542
TRT_DEPRECATED void destroy() noexcept
Destroy this INetworkDefinition object.
Definition: NvInfer.h:5856
void setStride(Dims stride) noexcept
Set the stride for computing the output slice data.
Definition: NvInfer.h:3991
ITensor * addInput(const char *name, DataType type, Dims dimensions) noexcept
Add an input tensor to the network.
Definition: NvInfer.h:5490
Dims getPostPadding() const noexcept
Get the padding.
Definition: NvInfer.h:2613
Hard sigmoid activation: max(0, min(1, alpha*x+beta))
void setDilationNd(Dims dilation) noexcept
Set the multi-dimension dilation of the convolution.
Definition: NvInfer.h:1370
Definition: NvInferImpl.h:607
void setReduceAxes(uint32_t reduceAxes) noexcept
Set which axes to reduce for the layer.
Definition: NvInfer.h:4150
TRT_DEPRECATED DimsHW getPadding() const noexcept
Get the padding of the deconvolution.
Definition: NvInfer.h:2478
void setWindowSize(int32_t windowSize) noexcept
Set the LRN window size.
Definition: NvInfer.h:2000
Use explicit padding, rounding output size up.
TRT_DEPRECATED void setAlignCorners(bool alignCorners) noexcept
Set whether to align corners while resizing.
Definition: NvInfer.h:4636
void setAlgorithmSelector(IAlgorithmSelector *selector) noexcept
Set Algorithm Selector.
Definition: NvInfer.h:7622
void setPrePaddingNd(Dims padding) noexcept
Set the padding that is applied at the start of the tensor.
Definition: NvInfer.h:3642
void setFirstTranspose(Permutation permutation) noexcept
Set the permutation applied by the first transpose operation.
Definition: NvInfer.h:3725
SliceMode
Controls how ISliceLayer handles out of bounds coordinates.
Definition: NvInfer.h:3878
Use SAME padding, with prePadding >= postPadding.
void reset() noexcept
Resets the builder configuration to defaults.
Definition: NvInfer.h:7521
TRT_DEPRECATED void setStride(DimsHW stride) noexcept
Get the stride of the convolution.
Definition: NvInfer.h:1049
Parametric softplus activation: alpha*log(exp(beta*x)+1)
UnaryOperation getOperation() const noexcept
Get the unary operation for the layer.
Definition: NvInfer.h:3449
void setFlag(BuilderFlag builderFlag) noexcept
Set a single build mode flag.
Definition: NvInfer.h:7400
void setPaddingNd(Dims padding) noexcept
Set the multi-dimension padding of the deconvolution.
Definition: NvInfer.h:2705
void setOperation(RNNOperation op) noexcept
Set the operation of the RNN layer.
Definition: NvInfer.h:3141
A fully connected layer in a network definition. This layer expects an input tensor of three or more ...
Definition: NvInfer.h:1439
Definition: NvInfer.h:3690
ISoftMaxLayer * addSoftMax(ITensor &input) noexcept
Add a SoftMax layer to the network.
Definition: NvInfer.h:5644
void setPostPaddingNd(Dims padding) noexcept
Set the padding that is applied at the end of the tensor.
Definition: NvInfer.h:3668
No operation is performed on the first recurrent layer.
TRT_DEPRECATED DimsHW getKernelSize() const noexcept
Get the HW kernel size of the convolution.
Definition: NvInfer.h:1011
Layer that represents a constant value.
Definition: NvInfer.h:4305
Use CAFFE padding, rounding output size down, uses prePadding value.
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:5684
Layer that represents a parametric ReLU operation.
Definition: NvInfer.h:4368
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:6128
const IOptimizationProfile * getCalibrationProfile() noexcept
Get the current calibration profile.
Definition: NvInfer.h:7655
constexpr int32_t EnumMax< RNNDirection >() noexcept
Maximum number of elements in RNNDirection enum.
Definition: NvInfer.h:3018
TRT_DEPRECATED IPaddingLayer * addPadding(ITensor &input, DimsHW prePadding, DimsHW postPadding) noexcept
Add a padding layer to the network.
Definition: NvInfer.h:5746
Minimum of the two elements.
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:6290
IErrorRecorder * getErrorRecorder() const noexcept
get the ErrorRecorder assigned to this interface.
Definition: NvInfer.h:8055
uint32_t getAxes() const noexcept
Get the axis along which softmax occurs.
Definition: NvInfer.h:2306
A network definition for input to the builder.
Definition: NvInfer.h:5450
bool getQuantizationFlag(QuantizationFlag flag) const noexcept
Returns true if the quantization flag is set.
Definition: NvInfer.h:7720
bool getFlag(BuilderFlag builderFlag) const noexcept
Returns true if the build mode flag is set.
Definition: NvInfer.h:7412
void setAxis(int32_t axis) noexcept
Set the quantization axis.
Definition: NvInfer.h:5334
Per-channel coefficients.
void setPostPadding(Dims padding) noexcept
Set the multi-dimension post-padding for pooling.
Definition: NvInfer.h:1856
void setDimensions(Dims dimensions) noexcept
Set the dimensions of a tensor.
Definition: NvInfer.h:233
An array of weights used as a layer parameter.
Definition: NvInferRuntime.h:152
provides a unique 128-bit identifier, which along with the input and output information denotes the v...
Definition: NvInfer.h:6845
int32_t getGatherAxis() const noexcept
Get the axis to gather on.
Definition: NvInfer.h:2877
bool isShapeTensor() const noexcept
Whether the tensor is a shape tensor.
Definition: NvInfer.h:471
bool combine(const ITimingCache &inputCache, bool ignoreMismatch) noexcept
Combine input timing cache into local instance.
Definition: NvInfer.h:7202
int32_t getNbGroups() const noexcept
Get the number of groups of the convolution.
Definition: NvInfer.h:1121
Layer that represents a padding operation.
Definition: NvInfer.h:3578
bool canRunOnDLA(const ILayer *layer) const noexcept
Checks if a layer can run on DLA.
Definition: NvInfer.h:7465
IFullyConnectedLayer * addFullyConnected(ITensor &input, int32_t nbOutputs, Weights kernelWeights, Weights biasWeights) noexcept
Add a fully connected layer to the network.
Definition: NvInfer.h:5546
A convolution layer in a network definition.
Definition: NvInfer.h:987
bool setDynamicRange(float min, float max) noexcept
Set dynamic range for the tensor.
Definition: NvInfer.h:288
DataType getOutputType(int32_t index) const noexcept
get the output type of this layer
Definition: NvInfer.h:698
TRT_DEPRECATED void setStride(DimsHW stride) noexcept
Set the stride for pooling.
Definition: NvInfer.h:1712
void setActivationType(ActivationType type) noexcept
Set the type of activation to be performed.
Definition: NvInfer.h:1550
void setDirection(RNNDirection op) noexcept
Set the direction of the RNN layer.
Definition: NvInfer.h:3183
Class to handle library allocated memory that is accessible to the user.
Definition: NvInferRuntime.h:170
Dims getOutputDimensions() const noexcept
Get the output dimensions.
Definition: NvInfer.h:4555
A layer that represents the identity function.
Definition: NvInfer.h:4291
void setPostPadding(Dims padding) noexcept
Set the multi-dimension post-padding of the convolution.
Definition: NvInfer.h:1242
RNNOperation getOperation() const noexcept
Get the operation of the RNN layer.
Definition: NvInfer.h:3150
DeviceType
The device that this layer/network will execute on.
Definition: NvInferRuntime.h:628
int32_t getNbOutputs() const noexcept
Get the number of outputs in the network.
Definition: NvInfer.h:5828
Network iterates from first to last and vice versa and outputs concatenated.
ProfilingVerbosity
List of verbosity levels of layer information exposed in NVTX annotations.
Definition: NvInfer.h:7120
const IAlgorithmIOInfo * getAlgorithmIOInfoByIndex(int32_t index) const noexcept
Returns the format of an Algorithm input or output. Algorithm inputs are incrementally numbered first...
Definition: NvInfer.h:6980
void setAxes(uint32_t axes) noexcept
Set the axis along which softmax is computed. Currently, only one axis can be set.
Definition: NvInfer.h:2296
Selu activation: x>0 ? beta * x : beta * (alpha*exp(x) - alpha)
Register layer names in NVTX message field and register layer detail in NVTX JSON payload field.
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:6691
Interface implemented by application for selecting and reporting algorithms of a layer provided by th...
Definition: NvInfer.h:6998
RNNInputMode
Enumerates the RNN input modes that may occur with an RNN layer.
Definition: NvInfer.h:3038
void setBiasWeights(Weights weights) noexcept
Set the bias weights.
Definition: NvInfer.h:1491
int32_t getNbOptimizationProfiles() const noexcept
Get number of optimization profiles.
Definition: NvInfer.h:7588
void clearFlag(BuilderFlag builderFlag) noexcept
clear a single build mode flag.
Definition: NvInfer.h:7388
Declaration of EnumMaxImpl struct to store maximum number of elements in an enumeration type.
Definition: NvInferRuntimeCommon.h:141
int32_t getAvgTimingIterations() const noexcept
Query the number of averaging iterations.
Definition: NvInfer.h:7278
IDequantizeLayer * addDequantize(ITensor &input, ITensor &scale) noexcept
Add a dequantization layer to the network.
Definition: NvInfer.h:6561
void setDilationNd(Dims dilation) noexcept
Set the multi-dimension dilation of the deconvolution.
Definition: NvInfer.h:2747
Dims getPostPaddingNd() const noexcept
Get the padding that is applied at the end of the tensor.
Definition: NvInfer.h:3680
TopKOperation getOperation() const noexcept
Get the operation for the layer.
Definition: NvInfer.h:4118
int32_t getNbGroups() const noexcept
Get the number of groups for a deconvolution.
Definition: NvInfer.h:2508
bool hasImplicitBatchDimension() const noexcept
Query whether the network was created with an implicit batch dimension.
Definition: NvInfer.h:6220
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:4597
Definition: NvInfer.h:4907
Clip activation: max(alpha, min(beta, x))
Permutation getFirstTranspose() const noexcept
Get the permutation applied by the first transpose operation.
Definition: NvInfer.h:3737
int32_t addOptimizationProfile(const IOptimizationProfile *profile) noexcept
Add an optimization profile.
Definition: NvInfer.h:7575
IElementWiseLayer * addElementWise(ITensor &input1, ITensor &input2, ElementWiseOperation op) noexcept
Add an elementwise layer to the network.
Definition: NvInfer.h:5710
Plugin class for user-implemented layers.
Definition: NvInferRuntimeCommon.h:414
Carries information about input or output of the algorithm. IAlgorithmIOInfo for all the input and ou...
Definition: NvInfer.h:6802
float getTimingMSec() const noexcept
The time in milliseconds to execute the algorithm.
Definition: NvInfer.h:6959
void setK(float k) noexcept
Set the LRN K value.
Definition: NvInfer.h:2063
Definition: NvInferRuntimeCommon.h:194
Definition: NvInfer.h:6720
TensorFormat getTensorFormat() const noexcept
Return TensorFormat of the input/output of algorithm.
Definition: NvInfer.h:6808
void markOutput(ITensor &tensor) noexcept
Mark a tensor as a network output.
Definition: NvInfer.h:5502
Scaled tanh activation: alpha*tanh(beta*x)
constexpr int32_t EnumMax< FillOperation >() noexcept
Maximum number of elements in FillOperation enum.
Definition: NvInfer.h:5074
uint32_t TensorFormats
It is capable of representing one or more TensorFormat by binary OR operations, e....
Definition: NvInfer.h:141
Definition: NvInferImpl.h:530
int32_t getNbInputs() const noexcept
Return number of inputs of the algorithm.
Definition: NvInfer.h:6903
void setType(DataType type) noexcept
Set the data type of a tensor.
Definition: NvInfer.h:261
void setPoolingType(PoolingType type) noexcept
Set the type of activation to be performed.
Definition: NvInfer.h:1660
Softsign activation: x / (1+|x|)
Enable layers marked to execute on GPU if layer cannot execute on DLA.
Fail with error when the coordinates are out of bounds. This is the default.
void resetDynamicRange() noexcept
Undo effect of setDynamicRange.
Definition: NvInfer.h:385
TRT_DEPRECATED void destroy() noexcept
Destroy this object.
Definition: NvInfer.h:7929
IRaggedSoftMaxLayer * addRaggedSoftMax(ITensor &input, ITensor &bounds) noexcept
Add a RaggedSoftMax layer to the network.
Definition: NvInfer.h:5952
void setLocation(TensorLocation location) noexcept
Set the storage location of a tensor.
Definition: NvInfer.h:367
Like kNONE, but transpose the matrix dimensions.
Enable building a refittable engine.
nvinfer1::IHostMemory * buildSerializedNetwork(INetworkDefinition &network, IBuilderConfig &config) noexcept
Builds and serializes a network for the given INetworkDefinition and IBuilderConfig.
Definition: NvInfer.h:8090
ISelectLayer * addSelect(ITensor &condition, ITensor &thenInput, ITensor &elseInput) noexcept
Add a select layer to the network.
Definition: NvInfer.h:6455
void setPrePadding(Dims padding) noexcept
Set the multi-dimension pre-padding of the deconvolution.
Definition: NvInfer.h:2575
An RNN layer in a network definition, version 2.
Definition: NvInfer.h:3086
float getAlpha() const noexcept
Get the LRN alpha value.
Definition: NvInfer.h:2031
DataType getDataType() const noexcept
Return DataType of the input/output of algorithm.
Definition: NvInfer.h:6816
void setBroadcastAcrossBatch(bool broadcastAcrossBatch) noexcept
Set whether to enable broadcast of tensor across the batch.
Definition: NvInfer.h:326
void setPaddingMode(PaddingMode paddingMode) noexcept
Set the padding mode.
Definition: NvInfer.h:1879
Logical AND of two elements.
int32_t getNbOutputChannels() const noexcept
Get the number of output channels K from the fully connected layer.
Definition: NvInfer.h:1459
Enable Int8 layer selection, with FP32 fallback with FP16 fallback if kFP16 also specified.
Dims getPostPadding() const noexcept
Get the post-padding.
Definition: NvInfer.h:1252
TRT_DEPRECATED bool getAlignCorners() const noexcept
True if align corners has been set.
Definition: NvInfer.h:4649
TRT_DEPRECATED DimsHW getWindowSize() const noexcept
Get the window size for pooling.
Definition: NvInfer.h:1696
IRecurrenceLayer * addRecurrence(ITensor &initialValue) noexcept
Create a recurrence layer for this loop with initialValue as its first input.
Definition: NvInfer.h:4970
Dims getDimensions() const noexcept
Get the output tensor's dimensions.
Definition: NvInfer.h:5131
Descriptor for two-dimensional spatial data.
Definition: NvInferLegacyDims.h:110
cudaStream_t getProfileStream() const noexcept
Get the cuda stream that is used to profile this network.
Definition: NvInfer.h:7559
Dims getStrideNd() const noexcept
Get the multi-dimension stride for pooling.
Definition: NvInfer.h:1938
Definition: NvInferImpl.h:718
Dims getReshapeDimensions() const noexcept
Get the reshaped dimensions.
Definition: NvInfer.h:3775
bool platformHasFastInt8() const noexcept
Determine whether the platform has fast native int8.
Definition: NvInfer.h:7917
ResizeMode
Enumerates various modes of resize in the resize layer. Resize mode set using setResizeMode().
Definition: NvInfer.h:4380
TensorFormat
Format of the input/output tensors.
Definition: NvInferRuntimeCommon.h:225
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:3251
Definition: NvInfer.h:4849
constexpr int32_t EnumMax< MatrixOperation >() noexcept
Maximum number of elements in MatrixOperation enum.
Definition: NvInfer.h:4200
IUnaryLayer * addUnary(ITensor &input, UnaryOperation operation) noexcept
Add a unary layer to the network.
Definition: NvInfer.h:5729
void setBiasWeights(Weights weights) noexcept
Set the bias weights for the convolution.
Definition: NvInfer.h:1160
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:6332
nvinfer1::INetworkDefinition * createNetworkV2(NetworkDefinitionCreationFlags flags) noexcept
Create a network definition object.
Definition: NvInfer.h:8007
Dims getStrides() const noexcept
Return strides of the input/output tensor of algorithm.
Definition: NvInfer.h:6824
void setPaddingNd(Dims padding) noexcept
Set the multi-dimension padding for pooling.
Definition: NvInfer.h:1957
int32_t getNbOutputs() const noexcept
Get the number of outputs of a layer.
Definition: NvInfer.h:571
PoolingType getPoolingType() const noexcept
Get the type of activation to be performed.
Definition: NvInfer.h:1670
void setNbOutputMaps(int32_t nbOutputMaps) noexcept
Set the number of output feature maps for the deconvolution.
Definition: NvInfer.h:2406
bool markOutputForShapes(ITensor &tensor) noexcept
Enable tensor's value to be computed by IExecutionContext::getShapeBinding.
Definition: NvInfer.h:6238
Output value is concatenation of values of tensor for each iteration, in reverse order.
bool getKeepDimensions() const noexcept
Get the boolean that specifies whether or not to keep the reduced dimensions for the layer.
Definition: NvInfer.h:3558
Application-implemented logging interface for the builder, engine and runtime.
Definition: NvInferRuntimeCommon.h:1194
Describes the context and requirements, that could be fulfilled by one or more instances of IAlgorith...
Definition: NvInfer.h:6877
Dims getStart() const noexcept
Get the start offset for the slice layer.
Definition: NvInfer.h:3948
bool outputTypeIsSet(int32_t index) const noexcept
whether the output type has been set for this layer
Definition: NvInfer.h:711
Dims getPrePadding() const noexcept
Get the pre-padding.
Definition: NvInfer.h:1225
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:1091
constexpr int32_t EnumMax< QuantizationFlag >() noexcept
Maximum number of quantization flags in QuantizationFlag enum.
Definition: NvInfer.h:7059
void setChannelAxis(int32_t channelAxis) noexcept
Set the channel axis.
Definition: NvInfer.h:2242
void setReverse(bool reverse) noexcept
Definition: NvInfer.h:4940
IShuffleLayer * addShuffle(ITensor &input) noexcept
Add a shuffle layer to the network.
Definition: NvInfer.h:5760
void unmarkOutput(ITensor &tensor) noexcept
unmark a tensor as a network output.
Definition: NvInfer.h:6109
An engine for executing inference on a built network, with functionally unsafe features.
Definition: NvInferRuntime.h:1258
uint32_t QuantizationFlags
Represents one or more QuantizationFlag values using binary OR operations.
Definition: NvInfer.h:7040
Enable FP16 layer selection, with FP32 fallback.
ActivationType
Enumerates the types of activation to perform in an activation layer.
Definition: NvInfer.h:148
void setPrecision(DataType dataType) noexcept
Set the computational precision of this layer.
Definition: NvInfer.h:619
Dims getWindowSizeNd() const noexcept
Get the multi-dimension window size for pooling.
Definition: NvInfer.h:1913
Permutation getSecondTranspose() const noexcept
Get the permutation applied by the second transpose operation.
Definition: NvInfer.h:3834
A RaggedSoftmax layer in a network definition.
Definition: NvInfer.h:4273
void setKeepDimensions(bool keepDimensions) noexcept
Set the boolean that specifies whether or not to keep the reduced dimensions for the layer.
Definition: NvInfer.h:3548
void setScale(Weights scale) noexcept
Set the scale value.
Definition: NvInfer.h:2176
Dims getDilationNd() const noexcept
Get the multi-dimension dilation of the convolution.
Definition: NvInfer.h:1380
void setReshapeDimensions(Dims dimensions) noexcept
Set the reshaped dimensions.
Definition: NvInfer.h:3762
void clearQuantizationFlag(QuantizationFlag flag) noexcept
clear a quantization flag.
Definition: NvInfer.h:7696
float getK() const noexcept
Get the LRN K value.
Definition: NvInfer.h:2073
Use SAME padding, with prePadding <= postPadding.
int32_t getAxis() const noexcept
Get axis being concatenated over.
Definition: NvInfer.h:4875
constexpr int32_t EnumMax< RNNInputMode >() noexcept
Maximum number of elements in RNNInputMode enum.
Definition: NvInfer.h:3046
void setKernelWeights(Weights weights) noexcept
Set the kernel weights, given as a KxC matrix in row-major order.
Definition: NvInfer.h:1469
RNNDirection getDirection() const noexcept
Get the direction of the RNN layer.
Definition: NvInfer.h:3192
Dims getDimensions(int32_t index, OptProfileSelector select) const noexcept
Get the minimum / optimum / maximum dimensions for input or output tensor.
Definition: NvInfer.h:6895
void setKernelWeights(Weights weights) noexcept
Set the kernel weights for the deconvolution.
Definition: NvInfer.h:2522
Definition: NvInferImpl.h:661
float getDynamicRangeMin() const noexcept
Get minimum of dynamic range.
Definition: NvInfer.h:395
ElementWiseOperation
Enumerates the binary operations that may be performed by an ElementWise layer.
Definition: NvInfer.h:2774
Definition: NvInferImpl.h:583
Generate a tensor with random values drawn from a uniform distribution.
ScaleMode getMode() const noexcept
Get the scale mode.
Definition: NvInfer.h:2146
IQuantizeLayer * addQuantize(ITensor &input, ITensor &scale) noexcept
Add a quantization layer to the network.
Definition: NvInfer.h:6580
Divide the first element by the second.
int32_t getChannelAxis() const noexcept
Get the channel axis.
Definition: NvInfer.h:2221
IInt8Calibrator * getInt8Calibrator() const noexcept
Get Int8 Calibration interface.
Definition: NvInfer.h:7321
Weights getKernelWeights() const noexcept
Get the kernel weights for the deconvolution.
Definition: NvInfer.h:2532
Slices an input tensor into an output tensor based on the offset and strides.
Definition: NvInfer.h:3921
int32_t uint32_t TacticSources
Represents a collection of one or more TacticSource values combine using bitwise-OR operations.
Definition: NvInferImpl.h:152
TRT_DEPRECATED DimsHW getDilation() const noexcept
Get the dilation for a convolution.
Definition: NvInfer.h:1198
void setOperation(int32_t index, MatrixOperation op) noexcept
Set the operation for an input tensor.
Definition: NvInfer.h:4239
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:424
TRT_DEPRECATED void setWindowSize(DimsHW windowSize) noexcept
Set the window size for pooling.
Definition: NvInfer.h:1684
ScaleMode
Controls how shift, scale and power are applied in a Scale layer.
Definition: NvInfer.h:2088
TacticSources getTacticSources() const noexcept
Get tactic sources.
Definition: NvInfer.h:7760
Use explicit padding, rounding output size down.
Definition: NvInferImpl.h:620
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:6967
Logical OR of two elements.
IConstantLayer * addConstant(Dims dimensions, Weights weights) noexcept
Add a constant layer to the network.
Definition: NvInfer.h:5997
ReduceOperation
Enumerates the reduce operations that may be performed by a Reduce layer.
Definition: NvInfer.h:3477
Rectified linear activation.
constexpr int32_t EnumMax< SliceMode >() noexcept
Maximum number of elements in SliceMode enum.
Definition: NvInfer.h:3886
Definition: NvInfer.h:4920
void setProfilingVerbosity(ProfilingVerbosity verbosity) noexcept
Set verbosity level of layer information exposed in NVTX annotations.
Definition: NvInfer.h:7600
void setZeroIsPlaceholder(bool zeroIsPlaceholder) noexcept
Set meaning of 0 in reshape dimensions.
Definition: NvInfer.h:3850
void setOutputDimensions(Dims dimensions) noexcept
Set the output dimensions.
Definition: NvInfer.h:4545
bool unmarkOutputForShapes(ITensor &tensor) noexcept
Undo markOutputForShapes.
Definition: NvInfer.h:6250
int32_t getNbLayers() const noexcept
Get the number of layers in the network.
Definition: NvInfer.h:5772
BuilderFlag
List of valid modes that the builder can enable when creating an engine from a network definition.
Definition: NvInfer.h:7079
void setAlpha(float alpha) noexcept
Set the alpha parameter (must be finite).
Definition: NvInfer.h:1575
void setBeta(float beta) noexcept
Set the beta parameter (must be finite).
Definition: NvInfer.h:1589
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:7501
constexpr int32_t EnumMax< TopKOperation >() noexcept
Maximum number of elements in TopKOperation enum.
Definition: NvInfer.h:4088
The first element to the power of the second element.
PaddingMode getPaddingMode() const noexcept
Get the padding mode.
Definition: NvInfer.h:2639
bool getReverse() const noexcept
True if and only if reversing input.
Definition: NvInfer.h:4946
void setAxis(int32_t axis) noexcept
Set where to insert the contenation axis. Ignored if getLoopOutput() is kLAST_VALUE.
Definition: NvInfer.h:4869
void setBlendFactor(float blendFactor) noexcept
Set the blending factor for the max_average_blend mode: max_average_blendPool = (1-blendFactor)*maxPo...
Definition: NvInfer.h:1769
Weights getBiasWeights() const noexcept
Get the bias weights for the deconvolution.
Definition: NvInfer.h:2557
constexpr int32_t EnumMax< ScaleMode >() noexcept
Maximum number of elements in ScaleMode enum.
Definition: NvInfer.h:2097
ElementWiseOperation getOperation() const noexcept
Get the binary operation for the layer.
Definition: NvInfer.h:2845
void setNbOutputMaps(int32_t nbOutputMaps) noexcept
Set the number of output maps for the convolution.
Definition: NvInfer.h:1023
bool precisionIsSet() const noexcept
whether the computational precision has been set for this layer
Definition: NvInfer.h:643
bool setWeightsName(Weights weights, const char *name) noexcept
Associate a name with all current uses of the given weights.
Definition: NvInfer.h:6508
The TensorRT API version 1 namespace.
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:6488
constexpr int32_t EnumMax< RNNGateType >() noexcept
Maximum number of elements in RNNGateType enum.
Definition: NvInfer.h:3070
TRT_DEPRECATED void setPadding(DimsHW padding) noexcept
Set the padding of the convolution.
Definition: NvInfer.h:1079
Base class for all layer classes in a network definition.
Definition: NvInfer.h:511
ITensor * getCellState() const noexcept
Get the initial cell state of the RNN.
Definition: NvInfer.h:3348
constexpr int32_t EnumMax< ReduceOperation >() noexcept
Maximum number of elements in ReduceOperation enum.
Definition: NvInfer.h:3488
TensorFormats getAllowedFormats() const noexcept
Get a bitmask of TensorFormat values that the tensor supports. For a shape tensor,...
Definition: NvInfer.h:437
void setStrideNd(Dims stride) noexcept
Set the multi-dimension stride of the convolution.
Definition: NvInfer.h:1316
void setWindowSizeNd(Dims windowSize) noexcept
Set the multi-dimension window size for pooling.
Definition: NvInfer.h:1903
void setStart(Dims start) noexcept
Set the start offset that the slice layer uses to create the output slice.
Definition: NvInfer.h:3933
constexpr int32_t EnumMax< ProfilingVerbosity >() noexcept
Maximum number of profile verbosity levels in ProfilingVerbosity enum.
Definition: NvInfer.h:7129
Inverse hyperbolic tangent.
void setCoordinateTransformation(ResizeCoordinateTransformation coordTransform) noexcept
Set coordinate transformation function.
Definition: NvInfer.h:4684
void setKernelSizeNd(Dims kernelSize) noexcept
Set the multi-dimension kernel size of the convolution.
Definition: NvInfer.h:1291
void setReduceAxes(uint32_t reduceAxes) noexcept
Set the axes over which to reduce.
Definition: NvInfer.h:3528
IReduceLayer * addReduce(ITensor &input, ReduceOperation operation, uint32_t reduceAxes, bool keepDimensions) noexcept
Add a reduce layer to the network.
Definition: NvInfer.h:5884
Definition: NvInfer.h:4961
Thresholded ReLU activation: x>alpha ? x : 0.
void setBiasWeights(Weights weights) noexcept
Set the bias weights for the deconvolution.
Definition: NvInfer.h:2547
Layer type for pluginV2.
Definition: NvInfer.h:3367
void setMode(SliceMode mode) noexcept
Set the slice mode.
Definition: NvInfer.h:4016
Definition: NvInferImpl.h:666
void setOperation(ReduceOperation op) noexcept
Set the reduce operation for the layer.
Definition: NvInfer.h:3508
Enables strict type constraints.
void setPrePadding(Dims padding) noexcept
Set the multi-dimension pre-padding for pooling.
Definition: NvInfer.h:1828
IConcatenationLayer * addConcatenation(ITensor *const *inputs, int32_t nbInputs) noexcept
Add a concatenation layer to the network.
Definition: NvInfer.h:5661
constexpr int32_t EnumMax< LoopOutput >() noexcept
Maximum number of elements in LoopOutput enum.
Definition: NvInfer.h:4768
bool platformHasFastFp16() const noexcept
Determine whether the platform has fast native fp16.
Definition: NvInfer.h:7909
int32_t getNbOutputMaps() const noexcept
Get the number of output feature maps for the deconvolution.
Definition: NvInfer.h:2416
nvinfer1::IHostMemory * serialize() const noexcept
Serialize a timing cache to IHostMemory object.
Definition: NvInfer.h:7178
int32_t getHiddenSize() const noexcept
Get the hidden size of the RNN.
Definition: NvInfer.h:3093
IResizeLayer * addResize(ITensor &input) noexcept
Add a resize layer to the network.
Definition: NvInfer.h:6384
Dims getPrePaddingNd() const noexcept
Get the padding that is applied at the start of the tensor.
Definition: NvInfer.h:3654
A elementwise layer in a network definition.
Definition: NvInfer.h:2821
bool setTacticSources(TacticSources tacticSources) noexcept
Set tactic sources.
Definition: NvInfer.h:7745
Network iterations from first input to last input.
MatrixOperation getOperation(int32_t index) const noexcept
Get the operation for an input tensor.
Definition: NvInfer.h:4249
Inverse hyperbolic cosine.
Weights getBiasWeights() const noexcept
Get the bias weights for the convolution.
Definition: NvInfer.h:1170
void setErrorRecorder(IErrorRecorder *recorder) noexcept
Set the ErrorRecorder for this interface.
Definition: NvInfer.h:6527
Definition: NvInferImpl.h:656
void setPostPadding(Dims padding) noexcept
Set the multi-dimension post-padding of the deconvolution.
Definition: NvInfer.h:2603
Tensor is a scalar of type kBOOL. Loop terminates when value is false.
const char * getName() const noexcept
Return the name of a layer.
Definition: NvInfer.h:542
RNNGateType
Identifies an individual gate within an RNN cell.
Definition: NvInfer.h:3058
void setOutputType(int32_t index, DataType dataType) noexcept
Set the output type of this layer.
Definition: NvInfer.h:684
ILoop * getLoop() const noexcept
Return pointer to ILoop associated with this boundary layer.
Definition: NvInfer.h:4794
void setKernelWeights(Weights weights) noexcept
Set the kernel weights for the convolution.
Definition: NvInfer.h:1135
Check if two elements are equal.
void setNbOutputChannels(int32_t nbOutputs) noexcept
Set the number of output channels K from the fully connected layer.
Definition: NvInfer.h:1449
Weights getKernelWeights() const noexcept
Get the kernel weights.
Definition: NvInfer.h:1479
Definition: NvInferImpl.h:316
constexpr int32_t EnumMax< CalibrationAlgoType >() noexcept
Maximum number of elements in CalibrationAlgoType enum.
Definition: NvInfer.h:6604
void setK(int32_t k) noexcept
Set the k value for the layer.
Definition: NvInfer.h:4130
Dims getDimensions() const noexcept
Get the dimensions of a tensor.
Definition: NvInfer.h:246
ISliceLayer * addSlice(ITensor &input, Dims start, Dims size, Dims stride) noexcept
Add a slice layer to the network.
Definition: NvInfer.h:6147
int32_t getDLACore() const noexcept
Get the DLA core that the engine executes on.
Definition: NvInfer.h:7491
ResizeCoordinateTransformation getCoordinateTransformation() const noexcept
Get coordinate transformation function.
Definition: NvInfer.h:4694
void setNearestRounding(ResizeRoundMode value) noexcept
Set rounding mode for nearest neighbor resize.
Definition: NvInfer.h:4733
EngineCapability getEngineCapability() const noexcept
Query EngineCapability flow configured for the builder.
Definition: NvInfer.h:7303
void setNbGroups(int32_t nbGroups) noexcept
Set the number of groups for a deconvolution.
Definition: NvInfer.h:2498
ITripLimitLayer * addTripLimit(ITensor &tensor, TripLimit limit) noexcept
Add a trip-count limiter, based on the given tensor.
Definition: NvInfer.h:4991
IFillLayer * addFill(Dims dimensions, FillOperation op) noexcept
Add a fill layer to the network.
Definition: NvInfer.h:6473
void setName(const char *name) noexcept
Set the name of a layer.
Definition: NvInfer.h:531
void setOperation(UnaryOperation op) noexcept
Set the unary operation for the layer.
Definition: NvInfer.h:3439
UnaryOperation
Enumerates the unary operations that may be performed by a Unary layer.
Definition: NvInfer.h:3392
Disable reuse of timing information across identical layers.
Class to handle tactic timing info collected from builder.
Definition: NvInfer.h:7164
Output value is value of tensor for last iteration.
int32_t getWindowSize() const noexcept
Get the LRN window size.
Definition: NvInfer.h:2010
Dims getPrePadding() const noexcept
Get the pre-padding.
Definition: NvInfer.h:1838
ITensor * getInput(int32_t index) const noexcept
Get the layer input corresponding to the given index.
Definition: NvInfer.h:563
ResizeRoundMode getNearestRounding() const noexcept
Get rounding mode for nearest neighbor resize.
Definition: NvInfer.h:4743
Dims getSize() const noexcept
Get dimensions of the output slice.
Definition: NvInfer.h:3977
constexpr int32_t EnumMax< RNNOperation >() noexcept
Maximum number of elements in RNNOperation enum.
Definition: NvInfer.h:2998
void setPower(Weights power) noexcept
Set the power value.
Definition: NvInfer.h:2196
Layer type for getting shape of a tensor.
Definition: NvInfer.h:4068
Coordinates wrap around periodically.
const char * getName() const noexcept
Get the tensor name.
Definition: NvInfer.h:214
ITopKLayer * addTopK(ITensor &input, TopKOperation op, int32_t k, uint32_t reduceAxes) noexcept
Add a TopK layer to the network.
Definition: NvInfer.h:5918
void setAxis(int32_t axis) noexcept
Set the quantization axis.
Definition: NvInfer.h:5421
float getBeta() const noexcept
Get the LRN beta value.
Definition: NvInfer.h:2052
Definition: NvInferImpl.h:638
void setInput(int32_t index, ITensor &tensor) noexcept
Replace an input of this layer with a specific tensor.
Definition: NvInfer.h:599
NetworkDefinitionCreationFlag
List of immutable network properties expressed at network creation time. NetworkDefinitionCreationFla...
Definition: NvInfer.h:7838
std::size_t getMaxWorkspaceSize() const noexcept
Get the maximum workspace size.
Definition: NvInfer.h:7347
ND (0 < N <= 8) nearest neighbor resizing.
const char * getName() const noexcept
Return the name of the loop.
Definition: NvInfer.h:5039
void setShift(Weights shift) noexcept
Set the shift value.
Definition: NvInfer.h:2156
Generate evenly spaced numbers over a specified interval.
PaddingMode
Enumerates the modes of padding to perform in convolution, deconvolution and pooling layer,...
Definition: NvInfer.h:955
void setAlpha(double alpha) noexcept
Set the alpha parameter.
Definition: NvInfer.h:5169
TRT_DEPRECATED DimsHW getPostPadding() const noexcept
Get the padding that is applied at the end of the tensor.
Definition: NvInfer.h:3628
void setEngineCapability(EngineCapability capability) noexcept
Configure the builder to target specified EngineCapability flow.
Definition: NvInfer.h:7291
DeviceType getDeviceType(const ILayer *layer) const noexcept
Get the device that this layer executes on.
Definition: NvInfer.h:7436
Perform the normal matrix multiplication in the first recurrent layer.
Dims getPostPadding() const noexcept
Get the padding.
Definition: NvInfer.h:1866
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:6744
void setMode(ScaleMode mode) noexcept
Set the scale mode.
Definition: NvInfer.h:2136
PaddingMode getPaddingMode() const noexcept
Get the padding mode.
Definition: NvInfer.h:1278
IActivationLayer * addActivation(ITensor &input, ActivationType type) noexcept
Add an activation layer to the network.
Definition: NvInfer.h:5566
bool reset() noexcept
Empty the timing cache.
Definition: NvInfer.h:7212
MatrixOperation
Enumerates the operations that may be performed on a tensor by IMatrixMultiplyLayer before multiplica...
Definition: NvInfer.h:4176
ITensor * getOutput(int32_t index) const noexcept
Get the layer output corresponding to the given index.
Definition: NvInfer.h:582
const char * getName() const noexcept
Return name of the algorithm node. This is a unique identifier for the IAlgorithmContext.
Definition: NvInfer.h:6884
Use formula to map the original index.
Optimization profile for dynamic input dimensions and shape tensors.
Definition: NvInferRuntime.h:1067
PoolingType
The type of pooling to perform in a pooling layer.
Definition: NvInfer.h:1622
float getDynamicRangeMax() const noexcept
Get maximum of dynamic range.
Definition: NvInfer.h:405
ILRNLayer * addLRN(ITensor &input, int32_t window, float alpha, float beta, float k) noexcept
Add a LRN layer to the network.
Definition: NvInfer.h:5604
void setBeta(float beta) noexcept
Set the LRN beta value.
Definition: NvInfer.h:2042
void resetOutputType(int32_t index) noexcept
reset the output type for this layer
Definition: NvInfer.h:723
DataType getType() const noexcept
Get the data type of a tensor.
Definition: NvInfer.h:273
ActivationType getActivationType() const noexcept
Get the type of activation to be performed.
Definition: NvInfer.h:1560
ResizeSelector
The coordinate selector when resize to single pixel output.
Definition: NvInfer.h:4453
Output value is concatenation of values of tensor for each iteration, in forward order.
void setHiddenState(ITensor &hidden) noexcept
Set the initial hidden state of the RNN with the provided hidden ITensor.
Definition: NvInfer.h:3311
DataType getPrecision() const noexcept
get the computational precision of this layer
Definition: NvInfer.h:631
ITensor * getSequenceLengths() const noexcept
Get the sequence lengths specified for the RNN.
Definition: NvInfer.h:3132
DataType
The type of weights and tensors.
Definition: NvInferRuntimeCommon.h:155
void setMaxBatchSize(int32_t batchSize) noexcept
Set the maximum batch size.
Definition: NvInfer.h:7888
Dims getKernelSizeNd() const noexcept
Get the multi-dimension kernel size of the deconvolution.
Definition: NvInfer.h:2662
void setQuantizationFlags(QuantizationFlags flags) noexcept
Set the quantization flags.
Definition: NvInfer.h:7672
Elu activation: x>=0 ? x : alpha * (exp(x) - 1).
int32_t getMaxBatchSize() const noexcept
Get the maximum batch size.
Definition: NvInfer.h:7901
void resetDeviceType(const ILayer *layer) noexcept
reset the DeviceType for this layer
Definition: NvInfer.h:7456
A concatenation layer in a network definition.
Definition: NvInfer.h:2328
Layer type for shuffling data.
Definition: NvInfer.h:3713
TRT_DEPRECATED DimsHW getPrePadding() const noexcept
Get the padding that is applied at the start of the tensor.
Definition: NvInfer.h:3602
bool getZeroIsPlaceholder() const noexcept
Get meaning of 0 in reshape dimensions.
Definition: NvInfer.h:3863
Weights getShift() const noexcept
Get the shift value.
Definition: NvInfer.h:2166
RNNOperation
Enumerates the RNN operations that may be performed by an RNN layer.
Definition: NvInfer.h:2988
A tensor in a network definition.
Definition: NvInfer.h:187
Dims getKernelSizeNd() const noexcept
Get the multi-dimension kernel size of the convolution.
Definition: NvInfer.h:1301
Definition: NvInfer.h:4804
BuilderFlags getFlags() const noexcept
Get the build mode flags for this builder config. Defaults to 0.
Definition: NvInfer.h:7376
ILoop * addLoop() noexcept
Add a loop to the network.
Definition: NvInfer.h:6417
Select the upper left pixel.
Definition: NvInferImpl.h:594
Definition: NvInferImpl.h:576
int32_t getDataLength() const noexcept
Get the maximum data length of the RNN.
Definition: NvInfer.h:3101
PaddingMode getPaddingMode() const noexcept
Get the padding mode.
Definition: NvInfer.h:1890
Floor division of the first element by the second.
void setGatherAxis(int32_t axis) noexcept
Set the axis to gather on. The axis must be less than the number of dimensions in the data input.
Definition: NvInfer.h:2867
LayerType
The type values of layer classes.
Definition: NvInfer.h:89
int32_t getNbElementWiseDims() const noexcept
Get the number of leading dimensions of indices tensor to be handled elementwise.
Definition: NvInfer.h:2899
void setProfileStream(const cudaStream_t stream) noexcept
Set the cuda stream that is used to profile this network.
Definition: NvInfer.h:7547
Definition: NvInferImpl.h:894
nvinfer1::IBuilderConfig * createBuilderConfig() noexcept
Create a builder configuration object.
Definition: NvInfer.h:7975
void resetPrecision() noexcept
reset the computational precision for this layer
Definition: NvInfer.h:653
Identical coefficients across all elements of the tensor.
A LRN layer in a network definition.
Definition: NvInfer.h:1988
RNNDirection
Enumerates the RNN direction that may be performed by an RNN layer.
Definition: NvInfer.h:3010
DeviceType getDefaultDeviceType() const noexcept
Get the default DeviceType which was set by setDefaultDeviceType.
Definition: NvInfer.h:7511
Four-gate LSTM network w/o peephole connections.
ILayer * getLayer(int32_t index) const noexcept
Get the layer specified by the given index.
Definition: NvInfer.h:5786
Check if element in first tensor is greater than corresponding element in second tensor.
FillOperation getOperation() const noexcept
Get the fill operation for the layer.
Definition: NvInfer.h:5151
void reset() noexcept
Resets the builder state to default values.
Definition: NvInfer.h:8063
TripLimit
Enum that describes kinds of trip limits.
Definition: NvInfer.h:4774
Layer that represents a reduction operator across Shape, Int32, Float, and Half tensors.
Definition: NvInfer.h:3500
ResizeMode getResizeMode() const noexcept
Get resize mode for an input tensor.
Definition: NvInfer.h:4619
Layer that represents a Matrix Multiplication.
Definition: NvInfer.h:4230
float getBlendFactor() const noexcept
Get the blending factor for the max_average_blend mode: max_average_blendPool = (1-blendFactor)*maxPo...
Definition: NvInfer.h:1782
Logical XOR of two elements.
int32_t getAxis() const noexcept
Get the quantization axis.
Definition: NvInfer.h:5410
Substract the second element from the first.
void setDimensions(Dims dimensions) noexcept
Set the dimensions for the layer.
Definition: NvInfer.h:4339
LeakyRelu activation: x>=0 ? x : alpha * x.
FillOperation
Enumerates the tensor fill operations that may performed by a fill layer.
Definition: NvInfer.h:5066
TRT_DEPRECATED IPoolingLayer * addPooling(ITensor &input, PoolingType type, DimsHW windowSize) noexcept
Add a pooling layer to the network.
Definition: NvInfer.h:5585
Forward declaration of IPluginFactory for use by other interfaces.
Definition: NvInferRuntime.h:78
void setSize(Dims size) noexcept
Set the dimensions of the output slice.
Definition: NvInfer.h:3962
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:3285
Definition: NvInfer.h:6703
Definition: NvInferImpl.h:649
TRT_DEPRECATED bool hasExplicitPrecision() const noexcept
True if network is an explicit precision network.
Definition: NvInfer.h:6401
Generate an output tensor with specified mode.
Definition: NvInfer.h:5104
QuantizationFlag
List of valid flags for quantizing the network to int8.
Definition: NvInfer.h:7049
double getAlpha() const noexcept
Get the value of alpha parameter.
Definition: NvInfer.h:5184
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:5016
void setQuantizationFlag(QuantizationFlag flag) noexcept
Set a single quantization flag.
Definition: NvInfer.h:7708
Application-implemented interface for calibration.
Definition: NvInfer.h:6620
Reference counted application-implemented error reporting interface for TensorRT objects.
Definition: NvInferRuntimeCommon.h:1373
void setGpuAllocator(IGpuAllocator *allocator) noexcept
Set the GPU allocator.
Definition: NvInfer.h:7965
Layer that represents a TopK reduction.
Definition: NvInfer.h:4100
void setResizeMode(ResizeMode resizeMode) noexcept
Set resize mode for an input tensor.
Definition: NvInfer.h:4609
TRT_DEPRECATED void setPrePadding(DimsHW padding) noexcept
Set the padding that is applied at the start of the tensor.
Definition: NvInfer.h:3590
ResizeRoundMode
The rounding mode for nearest neighbor resize.
Definition: NvInfer.h:4480
Dims getPaddingNd() const noexcept
Get the multi-dimension padding of the convolution.
Definition: NvInfer.h:1356
int32_t getAxis() const noexcept
Get axis being iterated over.
Definition: NvInfer.h:4930
OptProfileSelector
When setting or querying optimization profile parameters (such as shape tensor inputs or dynamic dime...
Definition: NvInferRuntime.h:1031
TensorLocation getLocation() const noexcept
Get the storage location of a tensor.
Definition: NvInfer.h:352
void setPrePadding(Dims padding) noexcept
Set the multi-dimension pre-padding of the convolution.
Definition: NvInfer.h:1215
Use CAFFE padding, rounding output size up, uses prePadding value.
Definition: NvInfer.h:5052
void setAxis(int32_t axis) noexcept
Set the axis along which concatenation occurs.
Definition: NvInfer.h:2341
int32_t getNbInputs() const noexcept
Get the number of inputs of a layer.
Definition: NvInfer.h:550
Weights getWeights() const noexcept
Get the weights for the layer.
Definition: NvInfer.h:4327
IGatherLayer * addGather(ITensor &data, ITensor &indices, int32_t axis) noexcept
Add a gather layer to the network.
Definition: NvInfer.h:5934
Dims getDilationNd() const noexcept
Get the multi-dimension dilation of the deconvolution.
Definition: NvInfer.h:2757
ITensor * getOutput(int32_t index) const noexcept
Get the output tensor specified by the given index.
Definition: NvInfer.h:5844
IIdentityLayer * addIdentity(ITensor &input) noexcept
Add an identity layer.
Definition: NvInfer.h:6082
TRT_DEPRECATED void setStride(DimsHW stride) noexcept
Get the stride of the deconvolution.
Definition: NvInfer.h:2432
Definition: NvInferImpl.h:742
bool dynamicRangeIsSet() const noexcept
Query whether dynamic range is set.
Definition: NvInfer.h:377
uint32_t NetworkDefinitionCreationFlags
Represents one or more NetworkDefinitionCreationFlag flags using binary OR operations....
Definition: NvInfer.h:7827
Definition: NvInfer.h:6685
constexpr int32_t EnumMax< UnaryOperation >() noexcept
Maximum number of elements in UnaryOperation enum.
Definition: NvInfer.h:3419
Definition: NvInfer.h:2858
Dims getPrePadding() const noexcept
Get the pre-padding.
Definition: NvInfer.h:2585
Definition: NvInferImpl.h:523
Dims getDimensions() const noexcept
Get the dimensions for the layer.
Definition: NvInfer.h:4351
void setOperation(FillOperation op) noexcept
Set the fill operation for the layer.
Definition: NvInfer.h:5141
virtual int32_t getMinTimingIterations() const noexcept
Query the number of minimization iterations.
Definition: NvInfer.h:7253
void setAxis(int32_t axis) noexcept
Set axis to iterate over.
Definition: NvInfer.h:4924
Dims getPaddingNd() const noexcept
Get the multi-dimension padding of the deconvolution.
Definition: NvInfer.h:2717
uint32_t getReduceAxes() const noexcept
Get the axes over which to reduce for the layer.
Definition: NvInfer.h:3538
static constexpr int32_t MAX_DIMS
The maximum number of dimensions supported for a tensor.
Definition: NvInferRuntimeCommon.h:198
void setOperation(ElementWiseOperation op) noexcept
Set the binary operation for the layer.
Definition: NvInfer.h:2833
virtual void setAvgTimingIterations(int32_t avgTiming) noexcept
Set the number of averaging iterations used when timing layers.
Definition: NvInfer.h:7266
void setName(const char *name) noexcept
Set the tensor name.
Definition: NvInfer.h:202
Dims getPaddingNd() const noexcept
Get the multi-dimension padding for pooling.
Definition: NvInfer.h:1969
Definition: NvInferImpl.h:539
virtual void setMinTimingIterations(int32_t minTiming) noexcept
Set the number of minimization iterations used when timing layers.
Definition: NvInfer.h:7241
constexpr int32_t EnumMax< LayerType >() noexcept
Maximum number of elements in LayerType enum.
Definition: NvInfer.h:130
uint32_t BuilderFlags
Represents one or more QuantizationFlag values using binary OR operations, e.g., 1U << BuilderFlag::k...
Definition: NvInfer.h:7070
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:6726
Product of the two elements.
double getBeta() const noexcept
Get the value of beta parameter.
Definition: NvInfer.h:5217
void setNbGroups(int32_t nbGroups) noexcept
Set the number of groups for a convolution.
Definition: NvInfer.h:1111
Definition: NvInfer.h:4790
void setPaddingNd(Dims padding) noexcept
Set the multi-dimension padding of the convolution.
Definition: NvInfer.h:1344
EngineCapability
List of supported engine capability flows.
Definition: NvInferRuntime.h:104
Describes a variation of execution of a layer. An algorithm is represented by IAlgorithmVariant and t...
Definition: NvInfer.h:6930
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:3260
Definition: NvInferImpl.h:705
nvinfer1::IOptimizationProfile * createOptimizationProfile() noexcept
Create a new optimization profile.
Definition: NvInfer.h:8021
void setDLACore(int32_t dlaCore) noexcept
Sets the DLA core used by the network.
Definition: NvInfer.h:7480
void setSelectorForSinglePixel(ResizeSelector selector) noexcept
Set coordinate selector function when resized to single pixel.
Definition: NvInfer.h:4709
IAlgorithmSelector * getAlgorithmSelector() const noexcept
Get Algorithm Selector.
Definition: NvInfer.h:7630
int32_t getNbInputs() const noexcept
Get the number of inputs in the network.
Definition: NvInfer.h:5798
bool isExecutionTensor() const noexcept
Whether the tensor is an execution tensor.
Definition: NvInfer.h:494
void setName(const char *name) noexcept
Set the name of the loop.
Definition: NvInfer.h:5029
void setFlags(BuilderFlags builderFlags) noexcept
Set the build mode flags to turn on builder options for this network.
Definition: NvInfer.h:7364
ResizeCoordinateTransformation
The resize coordinate transformation function.
Definition: NvInfer.h:4403
TRT_DEPRECATED DimsHW getStride() const noexcept
Get the stride of the convolution.
Definition: NvInfer.h:1059
const nvinfer1::ITimingCache * getTimingCache() const noexcept
Get the pointer to the timing cache from current IBuilderConfig.
Definition: NvInfer.h:7812
void setSecondTranspose(Permutation permutation) noexcept
Set the permutation applied by the second transpose operation.
Definition: NvInfer.h:3822
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:6368
void setDeviceType(const ILayer *layer, DeviceType deviceType) noexcept
Set the device that this layer must execute on.
Definition: NvInfer.h:7427
Definition: NvInferImpl.h:886
Definition: NvInfer.h:6738
IParametricReLULayer * addParametricReLU(ITensor &input, ITensor &slope) noexcept
Add a parametric ReLU layer to the network.
Definition: NvInfer.h:6268
bool getAverageCountExcludesPadding() const noexcept
Get whether average pooling uses as a denominator the overlap area between the window and the unpadde...
Definition: NvInfer.h:1810
Check if element in first tensor is less than corresponding element in second tensor.
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:5525
Definition: NvInferImpl.h:570
TRT_DEPRECATED void setDilation(DimsHW dilation) noexcept
Set the dilation for a convolution.
Definition: NvInfer.h:1186
Weights getPower() const noexcept
Get the power value.
Definition: NvInfer.h:2206
struct CUstream_st * cudaStream_t
Forward declaration of cudaStream_t.
Definition: NvInferRuntimeCommon.h:109
int32_t getAxis() const noexcept
Get the quantization axis.
Definition: NvInfer.h:5323
IPluginV2 & getPlugin() noexcept
Get the plugin for the layer.
Definition: NvInfer.h:3375
Weights getKernelWeights() const noexcept
Get the kernel weights of the convolution.
Definition: NvInfer.h:1145
A Scale layer in a network definition.
Definition: NvInfer.h:2128
const char * getName() const noexcept
Returns the name associated with the network.
Definition: NvInfer.h:6183
TRT_DEPRECATED void setPadding(DimsHW padding) noexcept
Set the padding of the deconvolution.
Definition: NvInfer.h:2464
void setWeights(Weights weights) noexcept
Set the weights for the layer.
Definition: NvInfer.h:4317
Three-gate network consisting of Gated Recurrent Units.
Dims getStride() const noexcept
Get the stride for the output slice.
Definition: NvInfer.h:4006
void setAlpha(float alpha) noexcept
Set the LRN alpha value.
Definition: NvInfer.h:2021
TRT_DEPRECATED nvinfer1::ICudaEngine * buildEngineWithConfig(INetworkDefinition &network, IBuilderConfig &config) noexcept
Builds an engine for the given INetworkDefinition and given IBuilderConfig.
Definition: NvInfer.h:7990
A Softmax layer in a network definition.
Definition: NvInfer.h:2263
TRT_DEPRECATED DimsHW getStride() const noexcept
Get the stride for pooling.
Definition: NvInfer.h:1724
IPoolingLayer * addPoolingNd(ITensor &input, PoolingType type, Dims windowSize) noexcept
Add a multi-dimension pooling layer to the network.
Definition: NvInfer.h:6310
void setName(const char *name) noexcept
Sets the name of the network.
Definition: NvInfer.h:6169
Tensor is scalar of type kINT32 that contains the trip count.
bool platformHasTf32() const noexcept
Determine whether the platform has TF32 support.
Definition: NvInfer.h:8071
int32_t getMaxSeqLength() const noexcept
Get the maximum sequence length of the RNN.
Definition: NvInfer.h:3097
Weights getBiasWeights() const noexcept
Get the bias weights.
Definition: NvInfer.h:1501
Holds properties for configuring a builder to produce an engine.
Definition: NvInfer.h:7226
void setPaddingMode(PaddingMode paddingMode) noexcept
Set the padding mode.
Definition: NvInfer.h:2627
bool setCalibrationProfile(const IOptimizationProfile *profile) noexcept
Add a calibration profile.
Definition: NvInfer.h:7645
TRT_DEPRECATED void setKernelSize(DimsHW kernelSize) noexcept
Set the HW kernel size of the convolution.
Definition: NvInfer.h:2382
ITensor * getInput(int32_t index) const noexcept
Get the input tensor specified by the given index.
Definition: NvInfer.h:5814
TRT_DEPRECATED DimsHW getStride() const noexcept
Get the stride of the deconvolution.
Definition: NvInfer.h:2444
ITensor * getHiddenState() const noexcept
Get the initial hidden state of the RNN.
Definition: NvInfer.h:3320
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:1799
Definition: NvInferImpl.h:675
constexpr int32_t EnumMax< NetworkDefinitionCreationFlag >() noexcept
Maximum number of elements in NetworkDefinitionCreationFlag enum.
Definition: NvInfer.h:7863
constexpr int32_t EnumMax< TripLimit >() noexcept
Maximum number of elements in TripLimit enum.
Definition: NvInfer.h:4783
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:3294
A resize layer in a network definition.
Definition: NvInfer.h:4526
Definition: NvInferImpl.h:633
IMatrixMultiplyLayer * addMatrixMultiply(ITensor &input0, MatrixOperation op0, ITensor &input1, MatrixOperation op1) noexcept
Add a MatrixMultiply layer to the network.
Definition: NvInfer.h:5971
Enable debugging of layers via synchronizing after every layer.
const TRT_DEPRECATED IAlgorithmIOInfo & getAlgorithmIOInfo(int32_t index) const noexcept
Returns the format of an Algorithm input or output. Algorithm inputs are incrementally numbered first...
Definition: NvInfer.h:6943
ProfilingVerbosity getProfilingVerbosity() const noexcept
Get verbosity level of layer information exposed in NVTX annotations.
Definition: NvInfer.h:7613
void setKernelSizeNd(Dims kernelSize) noexcept
Set the multi-dimension kernel size of the deconvolution.
Definition: NvInfer.h:2652
bool isDeviceTypeSet(const ILayer *layer) const noexcept
whether the DeviceType has been explicitly set for this layer
Definition: NvInfer.h:7446
bool isNetworkInput() const noexcept
Whether the tensor is a network input.
Definition: NvInfer.h:296
void setNbElementWiseDims(int32_t k) noexcept
Set the number of leading dimensions of indices tensor to be handled elementwise. k must be 0 if ther...
Definition: NvInfer.h:2889
void setScales(const float *scales, int32_t nbScales) noexcept
Set the resize scales.
Definition: NvInfer.h:4578
LayerType getType() const noexcept
Return the type of a layer.
Definition: NvInfer.h:519
A Pooling layer in a network definition.
Definition: NvInfer.h:1650
Definition: NvInferImpl.h:773
void setPaddingMode(PaddingMode paddingMode) noexcept
Set the padding mode.
Definition: NvInfer.h:1266
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:6709
uint32_t getReduceAxes() const noexcept
Get the axes to reduce for the layer.
Definition: NvInfer.h:4160
nvinfer1::ITimingCache * createTimingCache(const void *blob, std::size_t size) const noexcept
Create timing cache.
Definition: NvInfer.h:7779
void setCellState(ITensor &cell) noexcept
Set the initial cell state of the LSTM with the provided cell ITensor.
Definition: NvInfer.h:3339
Application-implemented class for controlling allocation on the GPU.
Definition: NvInferRuntimeCommon.h:1103
void setStrideNd(Dims stride) noexcept
Set the multi-dimension stride for pooling.
Definition: NvInfer.h:1928
IBuilder * createInferBuilder(ILogger &logger) noexcept
Create an instance of an IBuilder class.
Definition: NvInfer.h:8141
Layer that represents an unary operation.
Definition: NvInfer.h:3431
int64_t getTactic() const noexcept
Return tactic of the algorithm.
Definition: NvInfer.h:6859
ReduceOperation getOperation() const noexcept
Get the reduce operation for the layer.
Definition: NvInfer.h:3518
TRT_DEPRECATED void setPadding(DimsHW padding) noexcept
Set the padding for pooling.
Definition: NvInfer.h:1740
SliceMode getMode() const noexcept
Get the slice mode.
Definition: NvInfer.h:4026
Builds an engine from a network definition.
Definition: NvInfer.h:7875
TRT_DEPRECATED void setPostPadding(DimsHW padding) noexcept
Set the padding that is applied at the end of the tensor.
Definition: NvInfer.h:3616
void setInt8Calibrator(IInt8Calibrator *calibrator) noexcept
Set Int8 Calibration interface.
Definition: NvInfer.h:7313
void setBeta(double beta) noexcept
Set the beta parameter.
Definition: NvInfer.h:5202
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:8112
Definition: NvInferImpl.h:944
Allow the builder to examine weights and use optimized functions when weights have suitable sparsity.
void setOperation(TopKOperation op) noexcept
Set the operation for the layer.
Definition: NvInfer.h:4108
const IAlgorithmVariant & getAlgorithmVariant() const noexcept
Returns the algorithm variant.
Definition: NvInfer.h:6951
Mark the network to be an explicit batch network.
TRT_DEPRECATED DimsHW getKernelSize() const noexcept
Get the HW kernel size of the deconvolution.
Definition: NvInfer.h:2394
A Quantize layer in a network definition.
Definition: NvInfer.h:5312
IShapeLayer * addShape(ITensor &input) noexcept
Add a shape layer to the network.
Definition: NvInfer.h:6201
TopKOperation
Enumerates the operations that may be performed by a TopK layer.
Definition: NvInfer.h:4080
int64_t getImplementation() const noexcept
Return implementation of the algorithm.
Definition: NvInfer.h:6851
IIteratorLayer * addIterator(ITensor &tensor, int32_t axis=0, bool reverse=false) noexcept
Return layer that subscripts tensor by loop iteration.
Definition: NvInfer.h:5004
int32_t getLayerCount() const noexcept
Get the layer count of the RNN.
Definition: NvInfer.h:3089
TensorLocation
The location for tensor data storage, device or host.
Definition: NvInferRuntime.h:242
int32_t getK() const noexcept
Get the k value for the layer.
Definition: NvInfer.h:4140
float getBeta() const noexcept
Get the beta parameter.
Definition: NvInfer.h:1607
int32_t getNbDLACores() const noexcept
Return the number of DLA engines available to this builder.
Definition: NvInfer.h:7949
ResizeSelector getSelectorForSinglePixel() const noexcept
Get the coordinate selector function when resized to single pixel.
Definition: NvInfer.h:4719
void setMaxWorkspaceSize(std::size_t workspaceSize) noexcept
Set the maximum workspace size.
Definition: NvInfer.h:7333
int32_t getNbOutputMaps() const noexcept
Get the number of output maps for the convolution.
Definition: NvInfer.h:1033
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:7941
RNNInputMode getInputMode() const noexcept
Get the input mode of the RNN layer.
Definition: NvInfer.h:3168
int32_t getNbOutputs() const noexcept
Return number of outputs of the algorithm.
Definition: NvInfer.h:6911
QuantizationFlags getQuantizationFlags() const noexcept
Get the quantization flags.
Definition: NvInfer.h:7684
int32_t getAxis() const noexcept
Get the axis along which concatenation occurs.
Definition: NvInfer.h:2351
IScaleLayer * addScale(ITensor &input, ScaleMode mode, Weights shift, Weights scale, Weights power) noexcept
Add a Scale layer to the network.
Definition: NvInfer.h:5631
#define TRT_DEPRECATED
< Items that are marked as deprecated will be removed in a future release.
Definition: NvInferRuntimeCommon.h:76
void setStrideNd(Dims stride) noexcept
Set the multi-dimension stride of the deconvolution.
Definition: NvInfer.h:2677
bool getBroadcastAcrossBatch() const noexcept
Check if tensor is broadcast across the batch.
Definition: NvInfer.h:342
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:6065
bool isNetworkOutput() const noexcept
Whether the tensor is a network output.
Definition: NvInfer.h:304
bool setTimingCache(const ITimingCache &cache, bool ignoreMismatch) noexcept
Attach a timing cache to IBuilderConfig.
Definition: NvInfer.h:7802
Weights getScale() const noexcept
Get the scale value.
Definition: NvInfer.h:2186
An Activation layer in a network definition.
Definition: NvInfer.h:1540
CalibrationAlgoType
Version of calibration algorithm to use.
Definition: NvInfer.h:6594
void setErrorRecorder(IErrorRecorder *recorder) noexcept
Set the ErrorRecorder for this interface.
Definition: NvInfer.h:8040
A Dequantize layer in a network definition.
Definition: NvInfer.h:5399
TRT_DEPRECATED void destroy() noexcept
De-allocates any internally allocated memory.
Definition: NvInfer.h:7535
constexpr int32_t EnumMax< BuilderFlag >() noexcept
Maximum number of builder flags in BuilderFlag enum.
Definition: NvInfer.h:7107
void setInputMode(RNNInputMode op) noexcept
Set the input mode of the RNN layer.
Definition: NvInfer.h:3159
A deconvolution layer in a network definition.
Definition: NvInfer.h:2368
void setSequenceLengths(ITensor &seqLengths) noexcept
Specify individual sequence lengths in the batch with the ITensor pointed to by seqLengths.
Definition: NvInfer.h:3120
Copyright © 2024 NVIDIA Corporation
Privacy Policy |
Manage My Privacy |
Do Not Sell or Share My Data |
Terms of Service |
Accessibility |
Corporate Policies |
Product Security |
Contact