164 static constexpr int32_t kVALUE = 14;
204 mImpl->setName(name);
216 return mImpl->getName();
235 mImpl->setDimensions(dimensions);
249 return mImpl->getDimensions();
264 mImpl->setType(type);
276 return mImpl->getType();
293 return mImpl->setDynamicRange(min, max);
301 return mImpl->isNetworkInput();
309 return mImpl->isNetworkOutput();
326 mImpl->setBroadcastAcrossBatch(broadcastAcrossBatch);
340 return mImpl->getBroadcastAcrossBatch();
352 return mImpl->getLocation();
371 mImpl->setLocation(location);
383 return mImpl->dynamicRangeIsSet();
391 mImpl->resetDynamicRange();
401 return mImpl->getDynamicRangeMin();
411 return mImpl->getDynamicRangeMax();
433 mImpl->setAllowedFormats(formats);
446 return mImpl->getAllowedFormats();
477 return mImpl->isShapeTensor();
498 return mImpl->isExecutionTensor();
524 mImpl->setDimensionName(index, name);
539 return mImpl->getDimensionName(index);
564 return mLayer->getType();
578 mLayer->setName(name);
588 return mLayer->getName();
596 return mLayer->getNbInputs();
609 return mLayer->getInput(index);
617 return mLayer->getNbOutputs();
627 return mLayer->getOutput(index);
644 return mLayer->setInput(index, tensor);
675 mLayer->setPrecision(dataType);
687 return mLayer->getPrecision();
699 return mLayer->precisionIsSet();
709 mLayer->resetPrecision();
756 mLayer->setOutputType(index, dataType);
771 return mLayer->getOutputType(index);
785 return mLayer->outputTypeIsSet(index);
797 return mLayer->resetOutputType(index);
815 mLayer->setMetadata(metadata);
828 return mLayer->getMetadata();
833 apiv::VLayer* mLayer;
1010 static constexpr int32_t kVALUE = 4;
1038 mImpl->setNbOutputMaps(nbOutputMaps);
1048 return mImpl->getNbOutputMaps();
1068 mImpl->setNbGroups(nbGroups);
1078 return mImpl->getNbGroups();
1092 mImpl->setKernelWeights(weights);
1102 return mImpl->getKernelWeights();
1117 mImpl->setBiasWeights(weights);
1127 return mImpl->getBiasWeights();
1144 mImpl->setPrePadding(padding);
1154 return mImpl->getPrePadding();
1171 mImpl->setPostPadding(padding);
1181 return mImpl->getPostPadding();
1195 mImpl->setPaddingMode(paddingMode);
1207 return mImpl->getPaddingMode();
1220 mImpl->setKernelSizeNd(kernelSize);
1230 return mImpl->getKernelSizeNd();
1245 mImpl->setStrideNd(stride);
1255 return mImpl->getStrideNd();
1273 mImpl->setPaddingNd(padding);
1285 return mImpl->getPaddingNd();
1299 mImpl->setDilationNd(dilation);
1309 return mImpl->getDilationNd();
1358 mImpl->setActivationType(type);
1368 return mImpl->getActivationType();
1383 mImpl->setAlpha(alpha);
1397 mImpl->setBeta(beta);
1406 return mImpl->getAlpha();
1415 return mImpl->getBeta();
1445 static constexpr int32_t kVALUE = 3;
1472 mImpl->setPoolingType(type);
1482 return mImpl->getPoolingType();
1497 mImpl->setBlendFactor(blendFactor);
1510 return mImpl->getBlendFactor();
1524 mImpl->setAverageCountExcludesPadding(exclusive);
1535 return mImpl->getAverageCountExcludesPadding();
1553 mImpl->setPrePadding(padding);
1563 return mImpl->getPrePadding();
1581 mImpl->setPostPadding(padding);
1591 return mImpl->getPostPadding();
1604 mImpl->setPaddingMode(paddingMode);
1615 return mImpl->getPaddingMode();
1628 mImpl->setWindowSizeNd(windowSize);
1638 return mImpl->getWindowSizeNd();
1653 mImpl->setStrideNd(stride);
1663 return mImpl->getStrideNd();
1682 mImpl->setPaddingNd(padding);
1694 return mImpl->getPaddingNd();
1725 mImpl->setWindowSize(windowSize);
1735 return mImpl->getWindowSize();
1747 mImpl->setAlpha(alpha);
1757 return mImpl->getAlpha();
1769 mImpl->setBeta(beta);
1779 return mImpl->getBeta();
1801 return mImpl->getK();
1867 mImpl->setMode(mode);
1877 return mImpl->getMode();
1887 mImpl->setShift(shift);
1897 return mImpl->getShift();
1907 mImpl->setScale(scale);
1917 return mImpl->getScale();
1927 mImpl->setPower(power);
1937 return mImpl->getPower();
1952 return mImpl->getChannelAxis();
1973 mImpl->setChannelAxis(channelAxis);
2026 mImpl->setAxes(axes);
2036 return mImpl->getAxes();
2072 mImpl->setAxis(axis);
2082 return mImpl->getAxis();
2109 mImpl->setNbOutputMaps(nbOutputMaps);
2119 return mImpl->getNbOutputMaps();
2139 mImpl->setNbGroups(nbGroups);
2149 return mImpl->getNbGroups();
2163 mImpl->setKernelWeights(weights);
2173 return mImpl->getKernelWeights();
2188 mImpl->setBiasWeights(weights);
2198 return mImpl->getBiasWeights();
2215 mImpl->setPrePadding(padding);
2225 return mImpl->getPrePadding();
2242 mImpl->setPostPadding(padding);
2252 return mImpl->getPostPadding();
2266 mImpl->setPaddingMode(paddingMode);
2278 return mImpl->getPaddingMode();
2293 mImpl->setKernelSizeNd(kernelSize);
2303 return mImpl->getKernelSizeNd();
2320 mImpl->setStrideNd(stride);
2330 return mImpl->getStrideNd();
2348 mImpl->setPaddingNd(padding);
2360 return mImpl->getPaddingNd();
2386 mImpl->setDilationNd(dilation);
2396 return mImpl->getDilationNd();
2446 static constexpr int32_t kVALUE = 14;
2483 return mImpl->setOperation(op);
2495 return mImpl->getOperation();
2616 mImpl->setGatherAxis(axis);
2628 return mImpl->getGatherAxis();
2651 mImpl->setNbElementWiseDims(elementWiseDims);
2661 return mImpl->getNbElementWiseDims();
2671 mImpl->setMode(mode);
2681 return mImpl->getMode();
2710 return mImpl->getPlugin();
2737 return mImpl->getPlugin();
2820 mImpl->setOperation(op);
2830 return mImpl->getOperation();
2893 mImpl->setOperation(op);
2903 return mImpl->getOperation();
2913 mImpl->setReduceAxes(reduceAxes);
2923 return mImpl->getReduceAxes();
2933 mImpl->setKeepDimensions(keepDimensions);
2943 return mImpl->getKeepDimensions();
2977 mImpl->setPrePaddingNd(padding);
2989 return mImpl->getPrePaddingNd();
3003 mImpl->setPostPaddingNd(padding);
3015 return mImpl->getPostPaddingNd();
3065 mImpl->setFirstTranspose(permutation);
3077 return mImpl->getFirstTranspose();
3105 mImpl->setReshapeDimensions(dimensions);
3118 return mImpl->getReshapeDimensions();
3165 mImpl->setSecondTranspose(permutation);
3177 return mImpl->getSecondTranspose();
3193 return mImpl->setZeroIsPlaceholder(zeroIsPlaceholder);
3206 return mImpl->getZeroIsPlaceholder();
3317 mImpl->setStart(start);
3332 return mImpl->getStart();
3346 return mImpl->setSize(size);
3361 return mImpl->getSize();
3375 mImpl->setStride(stride);
3390 return mImpl->getStride();
3400 mImpl->setMode(mode);
3410 return mImpl->getMode();
3453 mImpl->setAxes(axes);
3468 return mImpl->getAxes();
3538 mImpl->setOperation(op);
3548 return mImpl->getOperation();
3576 return mImpl->getK();
3586 mImpl->setReduceAxes(reduceAxes);
3596 return mImpl->getReduceAxes();
3698 mImpl->setOperation(index, op);
3710 return mImpl->getOperation(index);
3818 mImpl->setToType(toType);
3829 return mImpl->getToType();
3858 mImpl->setWeights(weights);
3868 return mImpl->getWeights();
3880 mImpl->setDimensions(dimensions);
3892 return mImpl->getDimensions();
3938 static constexpr int32_t kVALUE = 3;
3992 static constexpr int32_t kVALUE = 3;
4022 static constexpr int32_t kVALUE = 2;
4058 static constexpr int32_t kVALUE = 4;
4121 return mImpl->setOutputDimensions(dimensions);
4131 return mImpl->getOutputDimensions();
4159 void setScales(
float const* scales, int32_t nbScales)
noexcept
4161 mImpl->setScales(scales, nbScales);
4178 int32_t
getScales(int32_t size,
float* scales)
const noexcept
4180 return mImpl->getScales(size, scales);
4192 mImpl->setResizeMode(interpolationMode);
4202 return mImpl->getResizeMode();
4237 mImpl->setCoordinateTransformation(coordTransform);
4247 return mImpl->getCoordinateTransformation();
4262 mImpl->setSelectorForSinglePixel(selector);
4272 return mImpl->getSelectorForSinglePixel();
4286 mImpl->setNearestRounding(value);
4296 return mImpl->getNearestRounding();
4318 mImpl->setCubicCoeff(A);
4328 return mImpl->getCubicCoeff();
4341 mImpl->setExcludeOutside(excludeFlag);
4351 return mImpl->getExcludeOutside();
4433 return mBoundary->getLoop();
4438 apiv::VLoopBoundaryLayer* mBoundary;
4456 return mBoundary->getConditional();
4461 apiv::VConditionalBoundaryLayer* mBoundary;
4544 return mImpl->setCondition(condition);
4562 return mImpl->addOutput(trueSubgraphOutput, falseSubgraphOutput);
4574 return mImpl->addInput(input);
4589 mImpl->setName(name);
4599 return mImpl->getName();
4669 return mImpl->getLoopOutput();
4686 mImpl->setAxis(axis);
4694 return mImpl->getAxis();
4743 return mImpl->getTripLimit();
4769 mImpl->setAxis(axis);
4777 return mImpl->getAxis();
4791 mImpl->setReverse(reverse);
4801 return mImpl->getReverse();
4829 return mImpl->addRecurrence(initialValue);
4850 return mImpl->addTripLimit(tensor, limit);
4863 return mImpl->addIterator(tensor, axis, reverse);
4876 return mImpl->addLoopOutput(tensor, outputKind, axis);
4891 mImpl->setName(name);
4901 return mImpl->getName();
4956 mImpl->setMessage(message);
4966 return mImpl->getMessage();
5068 mImpl->setDimensions(dimensions);
5083 return mImpl->getDimensions();
5093 mImpl->setOperation(op);
5103 return mImpl->getOperation();
5122 mImpl->setAlpha(alpha);
5137 return mImpl->getAlpha();
5156 mImpl->setBeta(beta);
5171 return mImpl->getBeta();
5232 mImpl->setAlphaInt64(alpha);
5247 return mImpl->getAlphaInt64();
5266 mImpl->setBetaInt64(beta);
5281 return mImpl->getBetaInt64();
5289 return mImpl->isAlphaBetaInt64();
5306 mImpl->setToType(toType);
5318 return mImpl->getToType();
5413 return mImpl->getAxis();
5424 mImpl->setAxis(axis);
5440 mImpl->setToType(toType);
5452 return mImpl->getToType();
5544 return mImpl->getAxis();
5555 mImpl->setAxis(axis);
5571 mImpl->setToType(toType);
5583 return mImpl->getToType();
5638 mImpl->setToType(toType);
5651 return mImpl->getToType();
5663 mImpl->setScaleType(scaleType);
5676 return mImpl->getScaleType();
5689 mImpl->setAxis(axis);
5699 return mImpl->getAxis();
5712 mImpl->setBlockSize(size);
5722 return mImpl->getBlockSize();
5780 return mImpl->setEquation(equation);
5790 return mImpl->getEquation();
5889 mImpl->setMode(mode);
5899 return mImpl->getMode();
5909 mImpl->setAxis(axis);
5917 return mImpl->getAxis();
5961 mImpl->setAxis(axis);
5969 return mImpl->getAxis();
5997 mImpl->setInterpolationMode(mode);
6009 return mImpl->getInterpolationMode();
6019 mImpl->setAlignCorners(alignCorners);
6031 return mImpl->getAlignCorners();
6043 return mImpl->setSampleMode(mode);
6055 return mImpl->getSampleMode();
6149 mImpl->setBoundingBoxFormat(fmt);
6161 return mImpl->getBoundingBoxFormat();
6175 mImpl->setTopKBoxLimit(limit);
6185 return mImpl->getTopKBoxLimit();
6238 mImpl->setBatchAxis(batchAxis);
6248 return mImpl->getBatchAxis();
6261 mImpl->setSequenceAxis(sequenceAxis);
6271 return mImpl->getSequenceAxis();
6309 return mImpl->setEpsilon(eps);
6319 return mImpl->getEpsilon();
6329 return mImpl->setAxes(axesMask);
6339 return mImpl->getAxes();
6360 return mImpl->setNbGroups(nbGroups);
6370 return mImpl->getNbGroups();
6396 return mImpl->setComputePrecision(type);
6406 return mImpl->getComputePrecision();
6501 static constexpr int32_t kVALUE = 1;
6547 return mImpl->setOperation(op);
6559 return mImpl->getOperation();
6571 mImpl->setExclusive(exclusive);
6583 return mImpl->getExclusive();
6595 mImpl->setReverse(reverse);
6607 return mImpl->getReverse();
6675 return mImpl->addInput(name, type, dimensions);
6689 mImpl->markOutput(tensor);
6707 return mImpl->markDebug(tensor);
6723 return mImpl->unmarkDebug(tensor);
6733 return mImpl->isDebugTensor(tensor);
6753 return mImpl->addActivation(input, type);
6772 return mImpl->addLRN(input, window, alpha, beta, k);
6798 return mImpl->addScale(input, mode, shift, scale, power);
6811 return mImpl->addSoftMax(input);
6828 return mImpl->addConcatenation(inputs, nbInputs);
6855 return mImpl->addElementWise(input1, input2, op);
6877 return mImpl->addUnary(input, operation);
6891 return mImpl->addShuffle(input);
6908 return mImpl->addOneHot(indices, values, depth, axis);
6920 return mImpl->getNbLayers();
6934 return mImpl->getLayer(index);
6946 return mImpl->getNbInputs();
6962 return mImpl->getInput(index);
6976 return mImpl->getNbOutputs();
6992 return mImpl->getOutput(index);
7019 return mImpl->addReduce(input, operation, reduceAxes, keepDimensions);
7051 return mImpl->addTopK(input, op, k, reduceAxes);
7067 return mImpl->addGather(data, indices, axis);
7083 return mImpl->addGatherV2(data, indices, mode);
7102 return mImpl->addRaggedSoftMax(input, bounds);
7124 return mImpl->addMatrixMultiply(input0, op0, input1, op1);
7138 return mImpl->addNonZero(input);
7162 return mImpl->addConstant(dimensions, weights);
7176 return mImpl->addIdentity(input);
7191 return mImpl->addCast(input, toType);
7206 mImpl->removeTensor(tensor);
7218 mImpl->unmarkOutput(tensor);
7239 return mImpl->addPluginV2(inputs, nbInputs, plugin);
7256 int32_t nbShapeInputs,
IPluginV3& plugin)
noexcept
7258 return mImpl->addPluginV3(inputs, nbInputs, shapeInputs, nbShapeInputs, plugin);
7277 return mImpl->addSlice(input, start, size, stride);
7301 mImpl->setName(name);
7315 return mImpl->getName();
7331 return mImpl->addShape(input);
7345 return mImpl->hasImplicitBatchDimension();
7355 return mImpl->getFlags();
7367 return mImpl->getFlag(networkDefinitionCreationFlag);
7384 return mImpl->markOutputForShapes(tensor);
7396 return mImpl->unmarkOutputForShapes(tensor);
7414 return mImpl->addParametricReLU(input, slope);
7437 return mImpl->addConvolutionNd(input, nbOutputMaps, kernelSize, kernelWeights, biasWeights);
7456 return mImpl->addPoolingNd(input, type, windowSize);
7479 return mImpl->addDeconvolutionNd(input, nbOutputMaps, kernelSize, kernelWeights, biasWeights);
7516 return mImpl->addScaleNd(input, mode, shift, scale, power, channelAxis);
7532 return mImpl->addResize(input);
7546 return mImpl->addLoop();
7561 return mImpl->addIfConditional();
7600 return mImpl->addSelect(condition, thenInput, elseInput);
7617 return mImpl->addAssertion(condition, message);
7642 return mImpl->addFill(dimensions, op);
7668 return mImpl->addFillV2(dimensions, op, outputType);
7684 return mImpl->addPaddingNd(input, prePadding, postPadding);
7708 return mImpl->setWeightsName(weights, name);
7727 mImpl->setErrorRecorder(recorder);
7742 return mImpl->getErrorRecorder();
7763 return mImpl->addDequantize(input, scale);
7785 return mImpl->addDequantizeV2(input, scale, outputType);
7805 return mImpl->addScatter(data, indices, updates, mode);
7826 return mImpl->addQuantize(input, scale);
7848 return mImpl->addQuantizeV2(input, scale, outputType);
7874 return mImpl->addDynamicQuantize(input, axis, blockSize, outputType, scaleType);
7889 return mImpl->addEinsum(inputs, nbInputs, equation);
7907 return mImpl->addGridSample(input, grid);
7925 return mImpl->addNMS(boxes, scores, maxOutputBoxesPerClass);
7942 return mImpl->addReverseSequence(input, sequenceLens);
7968 return mImpl->addNormalization(input, scale, bias, axesMask);
7990 return mImpl->addCumulative(input, axis, operation, exclusive, reverse);
8001 return mImpl->getBuilder();
8014 return mImpl->markWeightsRefittable(name);
8026 return mImpl->unmarkWeightsRefittable(name);
8039 return mImpl->areWeightsMarkedRefittable(name);
8058 return mImpl->addSqueeze(input, axes);
8079 return mImpl->addUnsqueeze(input, axes);
8151 virtual
bool getBatch(
void* bindings[],
char const* names[], int32_t nbBindings) noexcept = 0;
8167 virtual
void const* readCalibrationCache(std::
size_t& length) noexcept = 0;
8177 virtual
void writeCalibrationCache(
void const* ptr, std::
size_t length) noexcept = 0;
8343 virtual
double getRegressionCutoff() const noexcept = 0;
8357 virtual
void const* readHistogramCache(std::
size_t& length) noexcept = 0;
8367 virtual
void writeHistogramCache(
void const* ptr, std::
size_t length) noexcept = 0;
8410 return mImpl->getDataType();
8421 return mImpl->getStrides();
8431 return mImpl->getVectorizedDim();
8442 return mImpl->getComponentsPerElement();
8471 return mImpl->getImplementation();
8479 return mImpl->getTactic();
8507 return mImpl->getName();
8519 return mImpl->getDimensions(index, select);
8527 return mImpl->getNbInputs();
8535 return mImpl->getNbOutputs();
8564 return mImpl->getAlgorithmVariant();
8572 return mImpl->getTimingMSec();
8580 return mImpl->getWorkspaceSize();
8594 return mImpl->getAlgorithmIOInfoByIndex(index);
8629 int32_t nbChoices, int32_t* selection)
noexcept = 0;
8642 int32_t nbAlgorithms)
noexcept = 0;
8736 static constexpr int32_t kVALUE = 2;
8955 static constexpr uint64_t kINVALID_TACTIC_HASH = UINT64_MAX;
8988 return mImpl->serialize();
9012 return mImpl->combine(inputCache, ignoreMismatch);
9022 return mImpl->reset();
9039 int64_t
queryKeys(TimingCacheKey* keyBuffer, int64_t capacity)
const noexcept
9041 return mImpl->queryKeys(keyBuffer, capacity);
9056 TimingCacheValue
query(TimingCacheKey
const& key)
const noexcept
9058 return mImpl->query(key);
9078 bool update(TimingCacheKey
const& key, TimingCacheValue
const& value)
noexcept
9080 return mImpl->update(key, value);
9194 static constexpr int32_t kVALUE = 2;
9246 static constexpr int32_t kVALUE = 3;
9285 static constexpr int32_t kVALUE = 4;
9324 virtual void phaseStart(
char const* phaseName,
char const* parentPhase, int32_t nbSteps)
noexcept = 0;
9338 virtual bool stepComplete(
char const* phaseName, int32_t step)
noexcept = 0;
9397 virtual
void setAvgTimingIterations(int32_t avgTiming) noexcept
9399 mImpl->setAvgTimingIterations(avgTiming);
9411 return mImpl->getAvgTimingIterations();
9424 mImpl->setEngineCapability(capability);
9436 return mImpl->getEngineCapability();
9448 mImpl->setInt8Calibrator(calibrator);
9458 return mImpl->getInt8Calibrator();
9475 mImpl->setFlags(builderFlags);
9487 return mImpl->getFlags();
9499 mImpl->clearFlag(builderFlag);
9511 mImpl->setFlag(builderFlag);
9523 return mImpl->getFlag(builderFlag);
9540 mImpl->setDeviceType(layer, deviceType);
9550 return mImpl->getDeviceType(layer);
9562 return mImpl->isDeviceTypeSet(layer);
9572 mImpl->resetDeviceType(layer);
9582 return mImpl->canRunOnDLA(layer);
9598 mImpl->setDLACore(dlaCore);
9608 return mImpl->getDLACore();
9619 mImpl->setDefaultDeviceType(deviceType);
9629 return mImpl->getDefaultDeviceType();
9651 return mImpl->setProfileStream(stream);
9663 return mImpl->getProfileStream();
9680 return mImpl->addOptimizationProfile(profile);
9693 return mImpl->getNbOptimizationProfiles();
9705 mImpl->setProfilingVerbosity(verbosity);
9718 return mImpl->getProfilingVerbosity();
9730 mImpl->setAlgorithmSelector(selector);
9740 return mImpl->getAlgorithmSelector();
9758 return mImpl->setCalibrationProfile(profile);
9770 return mImpl->getCalibrationProfile();
9787 mImpl->setQuantizationFlags(flags);
9799 return mImpl->getQuantizationFlags();
9811 mImpl->clearQuantizationFlag(flag);
9823 mImpl->setQuantizationFlag(flag);
9835 return mImpl->getQuantizationFlag(flag);
9857 return mImpl->setTacticSources(tacticSources);
9872 return mImpl->getTacticSources();
9891 return mImpl->createTimingCache(blob, size);
9914 return mImpl->setTimingCache(cache, ignoreMismatch);
9924 return mImpl->getTimingCache();
9956 mImpl->setMemoryPoolLimit(pool, poolSize);
9975 return mImpl->getMemoryPoolLimit(pool);
9993 mImpl->setPreviewFeature(feature, enable);
10007 return mImpl->getPreviewFeature(feature);
10040 mImpl->setBuilderOptimizationLevel(level);
10052 return mImpl->getBuilderOptimizationLevel();
10069 mImpl->setHardwareCompatibilityLevel(hardwareCompatibilityLevel);
10082 return mImpl->getHardwareCompatibilityLevel();
10095 mImpl->setPluginsToSerialize(paths, nbPaths);
10108 return mImpl->getPluginToSerialize(index);
10118 return mImpl->getNbPluginsToSerialize();
10147 mImpl->setMaxAuxStreams(nbStreams);
10157 return mImpl->getMaxAuxStreams();
10173 return mImpl->setProgressMonitor(monitor);
10183 return mImpl->getProgressMonitor();
10199 mImpl->setRuntimePlatform(runtimePlatform);
10211 return mImpl->getRuntimePlatform();
10223 mImpl->setMaxNbTactics(maxNbTactics);
10235 return mImpl->getMaxNbTactics();
10251 return mImpl->setTilingOptimizationLevel(level);
10263 return mImpl->getTilingOptimizationLevel();
10279 return mImpl->setL2LimitForTiling(size);
10291 return mImpl->getL2LimitForTiling();
10368 return mImpl->platformHasFastFp16();
10378 return mImpl->platformHasFastInt8();
10390 return mImpl->getMaxDLABatchSize();
10398 return mImpl->getNbDLACores();
10416 mImpl->setGpuAllocator(allocator);
10426 return mImpl->createBuilderConfig();
10448 return mImpl->createNetworkV2(flags);
10463 return mImpl->createOptimizationProfile();
10482 mImpl->setErrorRecorder(recorder);
10497 return mImpl->getErrorRecorder();
10515 return mImpl->platformHasTf32();
10534 return mImpl->buildSerializedNetwork(network, config);
10554 return mImpl->buildEngineWithConfig(network, config);
10576 return mImpl->isNetworkSupported(network, config);
10586 return mImpl->getLogger();
10602 return mImpl->setMaxThreads(maxThreads);
10616 return mImpl->getMaxThreads();
10626 return mImpl->getPluginRegistry();
10639extern "C" TENSORRTAPI void* createInferBuilder_INTERNAL(
void* logger, int32_t version)
noexcept;
#define TENSORRTAPI
Definition: NvInferRuntimeBase.h:59
#define NV_TENSORRT_VERSION
Definition: NvInferRuntimeBase.h:91
#define TRT_DEPRECATED
Definition: NvInferRuntimeBase.h:45
#define TRT_DEPRECATED_ENUM
Definition: NvInferRuntimeBase.h:46
Definition: NvInferRuntimeBase.h:204
static constexpr int32_t MAX_DIMS
The maximum rank (number of dimensions) supported for a tensor.
Definition: NvInferRuntimeBase.h:207
An Activation layer in a network definition.
Definition: NvInfer.h:1347
void setBeta(float beta) noexcept
Set the beta parameter (must be finite).
Definition: NvInfer.h:1395
void setActivationType(ActivationType type) noexcept
Set the type of activation to be performed.
Definition: NvInfer.h:1356
ActivationType getActivationType() const noexcept
Get the type of activation to be performed.
Definition: NvInfer.h:1366
float getAlpha() const noexcept
Get the alpha parameter.
Definition: NvInfer.h:1404
virtual ~IActivationLayer() noexcept=default
float getBeta() const noexcept
Get the beta parameter.
Definition: NvInfer.h:1413
void setAlpha(float alpha) noexcept
Set the alpha parameter (must be finite).
Definition: NvInfer.h:1381
Describes the context and requirements, that could be fulfilled by one or more instances of IAlgorith...
Definition: NvInfer.h:8498
int32_t getNbOutputs() const noexcept
Return number of outputs of the algorithm.
Definition: NvInfer.h:8533
int32_t getNbInputs() const noexcept
Return number of inputs of the algorithm.
Definition: NvInfer.h:8525
char const * getName() const noexcept
Return name of the algorithm node.
Definition: NvInfer.h:8505
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:8517
Describes a variation of execution of a layer. An algorithm is represented by IAlgorithmVariant and t...
Definition: NvInfer.h:8557
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:8578
float getTimingMSec() const noexcept
The time in milliseconds to execute the algorithm.
Definition: NvInfer.h:8570
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:8592
virtual ~IAlgorithm() noexcept=default
IAlgorithmVariant const & getAlgorithmVariant() const noexcept
Returns the algorithm variant.
Definition: NvInfer.h:8562
Carries information about input or output of the algorithm. IAlgorithmIOInfo for all the input and ou...
Definition: NvInfer.h:8401
virtual ~IAlgorithmIOInfo() noexcept=default
int64_t getVectorizedDim() const noexcept
Return the index of the vectorized dimension or -1 for non-vectorized formats.
Definition: NvInfer.h:8429
Dims getStrides() const noexcept
Return strides of the input/output tensor of algorithm. For vectorized formats, strides are given in ...
Definition: NvInfer.h:8419
DataType getDataType() const noexcept
Return DataType of the input/output of algorithm.
Definition: NvInfer.h:8408
int64_t getComponentsPerElement() const noexcept
Return the number of components per element. This is always 1 for non-vectorized formats.
Definition: NvInfer.h:8440
provides a unique 128-bit identifier, which along with the input and output information denotes the v...
Definition: NvInfer.h:8464
virtual ~IAlgorithmVariant() noexcept=default
int64_t getTactic() const noexcept
Return tactic of the algorithm.
Definition: NvInfer.h:8477
int64_t getImplementation() const noexcept
Return implementation of the algorithm.
Definition: NvInfer.h:8469
An assertion layer in a network.
Definition: NvInfer.h:4944
void setMessage(char const *message) noexcept
Set the message to print if the assertion fails.
Definition: NvInfer.h:4954
char const * getMessage() const noexcept
Return the assertion message.
Definition: NvInfer.h:4964
virtual ~IAssertionLayer() noexcept=default
Holds properties for configuring a builder to produce an engine.
Definition: NvInfer.h:9385
void setMemoryPoolLimit(MemoryPoolType pool, std::size_t poolSize) noexcept
Set the memory size for the memory pool.
Definition: NvInfer.h:9954
void setQuantizationFlag(QuantizationFlag flag) noexcept
Set a single quantization flag.
Definition: NvInfer.h:9821
nvinfer1::ITimingCache * createTimingCache(void const *blob, std::size_t size) const noexcept
Create timing cache.
Definition: NvInfer.h:9889
void setPreviewFeature(PreviewFeature feature, bool enable) noexcept
Enable or disable a specific preview feature.
Definition: NvInfer.h:9991
TRT_DEPRECATED void setAlgorithmSelector(IAlgorithmSelector *selector) noexcept
Set Algorithm Selector.
Definition: NvInfer.h:9728
TRT_DEPRECATED void setInt8Calibrator(IInt8Calibrator *calibrator) noexcept
Set Int8 Calibration interface.
Definition: NvInfer.h:9446
bool getPreviewFeature(PreviewFeature feature) const noexcept
Get status of preview feature.
Definition: NvInfer.h:10005
void clearQuantizationFlag(QuantizationFlag flag) noexcept
clear a quantization flag.
Definition: NvInfer.h:9809
int32_t getBuilderOptimizationLevel() noexcept
Get builder optimization level.
Definition: NvInfer.h:10050
bool setTacticSources(TacticSources tacticSources) noexcept
Set tactic sources.
Definition: NvInfer.h:9855
void setPluginsToSerialize(char const *const *paths, int32_t nbPaths) noexcept
Set the plugin libraries to be serialized with version-compatible engines.
Definition: NvInfer.h:10093
bool setTilingOptimizationLevel(TilingOptimizationLevel level) noexcept
Set the Tiling optimization level.
Definition: NvInfer.h:10249
bool setL2LimitForTiling(int64_t size) noexcept
Set the L2 cache usage limit for Tiling optimization.
Definition: NvInfer.h:10277
bool getQuantizationFlag(QuantizationFlag flag) const noexcept
Returns true if the quantization flag is set.
Definition: NvInfer.h:9833
TRT_DEPRECATED IInt8Calibrator * getInt8Calibrator() const noexcept
Get Int8 Calibration interface.
Definition: NvInfer.h:9456
std::size_t getMemoryPoolLimit(MemoryPoolType pool) const noexcept
Get the memory size limit of the memory pool.
Definition: NvInfer.h:9973
int32_t getDLACore() const noexcept
Get the DLA core that the engine executes on.
Definition: NvInfer.h:9606
int32_t getNbPluginsToSerialize() const noexcept
Get the number of plugin library paths to be serialized with version-compatible engines.
Definition: NvInfer.h:10116
void setDeviceType(ILayer const *layer, DeviceType deviceType) noexcept
Set the device that this layer must execute on.
Definition: NvInfer.h:9538
void setEngineCapability(EngineCapability capability) noexcept
Configure the builder to target specified EngineCapability flow.
Definition: NvInfer.h:9422
int32_t getMaxAuxStreams() const noexcept
Get the maximum number of auxiliary streams that TRT is allowed to use.
Definition: NvInfer.h:10155
bool getFlag(BuilderFlag builderFlag) const noexcept
Returns true if the build mode flag is set.
Definition: NvInfer.h:9521
void setQuantizationFlags(QuantizationFlags flags) noexcept
Set the quantization flags.
Definition: NvInfer.h:9785
void setMaxNbTactics(int32_t maxNbTactics) noexcept
Set the maximum number of tactics to time when there is a choice of tactics.
Definition: NvInfer.h:10221
int64_t getL2LimitForTiling() const noexcept
Get the L2 cache usage limit for tiling optimization.
Definition: NvInfer.h:10289
void setProgressMonitor(IProgressMonitor *monitor) noexcept
Sets the progress monitor for building a network.
Definition: NvInfer.h:10171
void setProfilingVerbosity(ProfilingVerbosity verbosity) noexcept
Set verbosity level of layer information exposed in NVTX annotations and IEngineInspector.
Definition: NvInfer.h:9703
int32_t getNbOptimizationProfiles() const noexcept
Get number of optimization profiles.
Definition: NvInfer.h:9691
QuantizationFlags getQuantizationFlags() const noexcept
Get the quantization flags.
Definition: NvInfer.h:9797
nvinfer1::ITimingCache const * getTimingCache() const noexcept
Get the pointer to the timing cache from current IBuilderConfig.
Definition: NvInfer.h:9922
void reset() noexcept
Resets the builder configuration to defaults.
Definition: NvInfer.h:9637
bool setTimingCache(ITimingCache const &cache, bool ignoreMismatch) noexcept
Attach a timing cache to IBuilderConfig.
Definition: NvInfer.h:9912
char const * getPluginToSerialize(int32_t index) const noexcept
Get the plugin library path to be serialized with version-compatible engines.
Definition: NvInfer.h:10106
EngineCapability getEngineCapability() const noexcept
Query EngineCapability flow configured for the builder.
Definition: NvInfer.h:9434
RuntimePlatform getRuntimePlatform() const noexcept
Get the target platform for runtime execution.
Definition: NvInfer.h:10209
DeviceType getDefaultDeviceType() const noexcept
Get the default DeviceType which was set by setDefaultDeviceType.
Definition: NvInfer.h:9627
void setRuntimePlatform(RuntimePlatform runtimePlatform) noexcept
Set the target platform for runtime execution.
Definition: NvInfer.h:10197
int32_t getMaxNbTactics() const noexcept
Query the maximum number of tactics timed when there is a choice.
Definition: NvInfer.h:10233
BuilderFlags getFlags() const noexcept
Get the build mode flags for this builder config. Defaults to 0.
Definition: NvInfer.h:9485
void setFlags(BuilderFlags builderFlags) noexcept
Set the build mode flags to turn on builder options for this network.
Definition: NvInfer.h:9473
TacticSources getTacticSources() const noexcept
Get tactic sources.
Definition: NvInfer.h:9870
void resetDeviceType(ILayer const *layer) noexcept
reset the DeviceType for this layer
Definition: NvInfer.h:9570
void setDLACore(int32_t dlaCore) noexcept
Sets the DLA core used by the network. Defaults to -1.
Definition: NvInfer.h:9596
HardwareCompatibilityLevel getHardwareCompatibilityLevel() const noexcept
Get the hardware compatibility level.
Definition: NvInfer.h:10080
void clearFlag(BuilderFlag builderFlag) noexcept
clear a single build mode flag.
Definition: NvInfer.h:9497
int32_t addOptimizationProfile(IOptimizationProfile const *profile) noexcept
Add an optimization profile.
Definition: NvInfer.h:9678
IProgressMonitor * getProgressMonitor() const noexcept
Definition: NvInfer.h:10181
apiv::VBuilderConfig * mImpl
Definition: NvInfer.h:10295
TRT_DEPRECATED IOptimizationProfile const * getCalibrationProfile() noexcept
Get the current calibration profile.
Definition: NvInfer.h:9768
int32_t getAvgTimingIterations() const noexcept
Query the number of averaging iterations.
Definition: NvInfer.h:9409
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:9617
void setFlag(BuilderFlag builderFlag) noexcept
Set a single build mode flag.
Definition: NvInfer.h:9509
TRT_DEPRECATED bool setCalibrationProfile(IOptimizationProfile const *profile) noexcept
Add a calibration profile.
Definition: NvInfer.h:9756
virtual ~IBuilderConfig() noexcept=default
DeviceType getDeviceType(ILayer const *layer) const noexcept
Get the device that this layer executes on.
Definition: NvInfer.h:9548
bool canRunOnDLA(ILayer const *layer) const noexcept
Checks if a layer can run on DLA.
Definition: NvInfer.h:9580
cudaStream_t getProfileStream() const noexcept
Get the CUDA stream that is used to profile this network.
Definition: NvInfer.h:9661
void setHardwareCompatibilityLevel(HardwareCompatibilityLevel hardwareCompatibilityLevel) noexcept
Set the hardware compatibility level.
Definition: NvInfer.h:10067
TilingOptimizationLevel getTilingOptimizationLevel() const noexcept
Get the Tiling optimization level.
Definition: NvInfer.h:10261
void setMaxAuxStreams(int32_t nbStreams) noexcept
Set the maximum number of auxiliary streams that TRT is allowed to use.
Definition: NvInfer.h:10145
ProfilingVerbosity getProfilingVerbosity() const noexcept
Get verbosity level of layer information exposed in NVTX annotations and IEngineInspector.
Definition: NvInfer.h:9716
bool isDeviceTypeSet(ILayer const *layer) const noexcept
whether the DeviceType has been explicitly set for this layer
Definition: NvInfer.h:9560
void setBuilderOptimizationLevel(int32_t level) noexcept
Set builder optimization level.
Definition: NvInfer.h:10038
void setProfileStream(const cudaStream_t stream) noexcept
Set the CUDA stream that is used to profile this network.
Definition: NvInfer.h:9649
TRT_DEPRECATED IAlgorithmSelector * getAlgorithmSelector() const noexcept
Get Algorithm Selector.
Definition: NvInfer.h:9738
Builds an engine from a network definition.
Definition: NvInfer.h:10357
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:10388
int32_t getNbDLACores() const noexcept
Return the number of DLA engines available to this builder.
Definition: NvInfer.h:10396
IErrorRecorder * getErrorRecorder() const noexcept
get the ErrorRecorder assigned to this interface.
Definition: NvInfer.h:10495
apiv::VBuilder * mImpl
Definition: NvInfer.h:10630
ILogger * getLogger() const noexcept
get the logger with which the builder was created
Definition: NvInfer.h:10584
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:10574
int32_t getMaxThreads() const noexcept
get the maximum number of threads that can be used by the builder.
Definition: NvInfer.h:10614
IPluginRegistry & getPluginRegistry() noexcept
get the local plugin registry that can be used by the builder.
Definition: NvInfer.h:10624
TRT_DEPRECATED bool platformHasFastInt8() const noexcept
Determine whether the platform has fast native int8.
Definition: NvInfer.h:10376
nvinfer1::IOptimizationProfile * createOptimizationProfile() noexcept
Create a new optimization profile.
Definition: NvInfer.h:10461
void setGpuAllocator(IGpuAllocator *allocator) noexcept
Set the GPU allocator.
Definition: NvInfer.h:10414
nvinfer1::INetworkDefinition * createNetworkV2(NetworkDefinitionCreationFlags flags) noexcept
Create a network definition object.
Definition: NvInfer.h:10446
nvinfer1::IBuilderConfig * createBuilderConfig() noexcept
Create a builder configuration object.
Definition: NvInfer.h:10424
void reset() noexcept
Resets the builder state to default values.
Definition: NvInfer.h:10503
bool setMaxThreads(int32_t maxThreads) noexcept
Set the maximum number of threads.
Definition: NvInfer.h:10600
void setErrorRecorder(IErrorRecorder *recorder) noexcept
Set the ErrorRecorder for this interface.
Definition: NvInfer.h:10480
nvinfer1::IHostMemory * buildSerializedNetwork(INetworkDefinition &network, IBuilderConfig &config) noexcept
Builds and serializes a network for the given INetworkDefinition and IBuilderConfig.
Definition: NvInfer.h:10532
virtual ~IBuilder() noexcept=default
TRT_DEPRECATED bool platformHasTf32() const noexcept
Determine whether the platform has TF32 support.
Definition: NvInfer.h:10513
nvinfer1::ICudaEngine * buildEngineWithConfig(INetworkDefinition &network, IBuilderConfig &config) noexcept
Builds a network for the given INetworkDefinition and IBuilderConfig.
Definition: NvInfer.h:10552
A cast layer in a network.
Definition: NvInfer.h:3807
virtual ~ICastLayer() noexcept=default
apiv::VCastLayer * mImpl
Definition: NvInfer.h:3833
DataType getToType() const noexcept
Return cast layer output type.
Definition: NvInfer.h:3827
void setToType(DataType toType) noexcept
Set cast layer output type.
Definition: NvInfer.h:3816
A concatenation layer in a network definition.
Definition: NvInfer.h:2057
void setAxis(int32_t axis) noexcept
Set the axis along which concatenation occurs.
Definition: NvInfer.h:2070
int32_t getAxis() const noexcept
Get the axis along which concatenation occurs.
Definition: NvInfer.h:2080
virtual ~IConcatenationLayer() noexcept=default
This layer represents a condition input to an IIfConditional.
Definition: NvInfer.h:4470
virtual ~IConditionLayer() noexcept=default
Layer that represents a constant value.
Definition: NvInfer.h:3846
void setWeights(Weights weights) noexcept
Set the weights for the layer.
Definition: NvInfer.h:3856
Weights getWeights() const noexcept
Get the weights for the layer.
Definition: NvInfer.h:3866
void setDimensions(Dims const &dimensions) noexcept
Set the dimensions for the layer.
Definition: NvInfer.h:3878
apiv::VConstantLayer * mImpl
Definition: NvInfer.h:3896
virtual ~IConstantLayer() noexcept=default
Dims getDimensions() const noexcept
Get the dimensions for the layer.
Definition: NvInfer.h:3890
A convolution layer in a network definition.
Definition: NvInfer.h:1027
Dims getPrePadding() const noexcept
Get the pre-padding.
Definition: NvInfer.h:1152
Weights getBiasWeights() const noexcept
Get the bias weights for the convolution.
Definition: NvInfer.h:1125
void setPaddingMode(PaddingMode paddingMode) noexcept
Set the padding mode.
Definition: NvInfer.h:1193
void setDilationNd(Dims const &dilation) noexcept
Set the multi-dimension dilation of the convolution.
Definition: NvInfer.h:1297
Dims getPaddingNd() const noexcept
Get the multi-dimension padding of the convolution.
Definition: NvInfer.h:1283
Dims getStrideNd() const noexcept
Get the multi-dimension stride of the convolution.
Definition: NvInfer.h:1253
Weights getKernelWeights() const noexcept
Get the kernel weights of the convolution.
Definition: NvInfer.h:1100
void setStrideNd(Dims const &stride) noexcept
Set the multi-dimension stride of the convolution.
Definition: NvInfer.h:1243
Dims getDilationNd() const noexcept
Get the multi-dimension dilation of the convolution.
Definition: NvInfer.h:1307
int64_t getNbOutputMaps() const noexcept
Get the number of output maps for the convolution.
Definition: NvInfer.h:1046
void setKernelWeights(Weights weights) noexcept
Set the kernel weights for the convolution.
Definition: NvInfer.h:1090
Dims getPostPadding() const noexcept
Get the post-padding.
Definition: NvInfer.h:1179
int64_t getNbGroups() const noexcept
Get the number of groups of the convolution.
Definition: NvInfer.h:1076
PaddingMode getPaddingMode() const noexcept
Get the padding mode.
Definition: NvInfer.h:1205
virtual ~IConvolutionLayer() noexcept=default
void setNbGroups(int64_t nbGroups) noexcept
Set the number of groups for a convolution.
Definition: NvInfer.h:1066
void setNbOutputMaps(int64_t nbOutputMaps) noexcept
Set the number of output maps for the convolution.
Definition: NvInfer.h:1036
void setBiasWeights(Weights weights) noexcept
Set the bias weights for the convolution.
Definition: NvInfer.h:1115
Dims getKernelSizeNd() const noexcept
Get the multi-dimension kernel size of the convolution.
Definition: NvInfer.h:1228
void setPaddingNd(Dims const &padding) noexcept
Set the multi-dimension padding of the convolution.
Definition: NvInfer.h:1271
void setPrePadding(Dims const &padding) noexcept
Set the multi-dimension pre-padding of the convolution.
Definition: NvInfer.h:1142
void setPostPadding(Dims const &padding) noexcept
Set the multi-dimension post-padding of the convolution.
Definition: NvInfer.h:1169
void setKernelSizeNd(Dims const &kernelSize) noexcept
Set the multi-dimension kernel size of the convolution.
Definition: NvInfer.h:1218
An engine for executing inference on a built network, with functionally unsafe features.
Definition: NvInferRuntime.h:3004
Layer that represents a cumulative operation across a tensor.
Definition: NvInfer.h:6534
bool setOperation(CumulativeOperation op) noexcept
Set the cumulative operation for the layer.
Definition: NvInfer.h:6545
void setReverse(bool reverse) noexcept
Specify whether the cumulative operation should be applied backward.
Definition: NvInfer.h:6593
apiv::VCumulativeLayer * mImpl
Definition: NvInfer.h:6611
bool getExclusive() const noexcept
Get whether it is exclusive accumulation or inclusive accumulation.
Definition: NvInfer.h:6581
virtual ~ICumulativeLayer() noexcept=default
bool getReverse() const noexcept
Get the boolean that specifies whether the cumulative operation should be applied backward.
Definition: NvInfer.h:6605
void setExclusive(bool exclusive) noexcept
Set whether it is an exclusive accumulation or inclusive accumulation.
Definition: NvInfer.h:6569
CumulativeOperation getOperation() const noexcept
Get the cumulative operation for the layer.
Definition: NvInfer.h:6557
A deconvolution layer in a network definition.
Definition: NvInfer.h:2098
void setBiasWeights(Weights weights) noexcept
Set the bias weights for the deconvolution.
Definition: NvInfer.h:2186
int64_t getNbGroups() const noexcept
Get the number of groups for a deconvolution.
Definition: NvInfer.h:2147
Weights getKernelWeights() const noexcept
Get the kernel weights for the deconvolution.
Definition: NvInfer.h:2171
void setPrePadding(Dims const &padding) noexcept
Set the multi-dimension pre-padding of the deconvolution.
Definition: NvInfer.h:2213
Dims getStrideNd() const noexcept
Get the multi-dimension stride of the deconvolution.
Definition: NvInfer.h:2328
Dims getDilationNd() const noexcept
Get the multi-dimension dilation of the deconvolution.
Definition: NvInfer.h:2394
Weights getBiasWeights() const noexcept
Get the bias weights for the deconvolution.
Definition: NvInfer.h:2196
void setKernelWeights(Weights weights) noexcept
Set the kernel weights for the deconvolution.
Definition: NvInfer.h:2161
int64_t getNbOutputMaps() const noexcept
Get the number of output feature maps for the deconvolution.
Definition: NvInfer.h:2117
void setStrideNd(Dims const &stride) noexcept
Set the multi-dimension stride of the deconvolution.
Definition: NvInfer.h:2318
Dims getPostPadding() const noexcept
Get the padding.
Definition: NvInfer.h:2250
Dims getKernelSizeNd() const noexcept
Get the multi-dimension kernel size of the deconvolution.
Definition: NvInfer.h:2301
void setPostPadding(Dims const &padding) noexcept
Set the multi-dimension post-padding of the deconvolution.
Definition: NvInfer.h:2240
void setKernelSizeNd(Dims const &kernelSize) noexcept
Set the multi-dimension kernel size of the deconvolution.
Definition: NvInfer.h:2291
virtual ~IDeconvolutionLayer() noexcept=default
void setPaddingNd(Dims const &padding) noexcept
Set the multi-dimension padding of the deconvolution.
Definition: NvInfer.h:2346
void setNbOutputMaps(int64_t nbOutputMaps) noexcept
Set the number of output feature maps for the deconvolution.
Definition: NvInfer.h:2107
Dims getPaddingNd() const noexcept
Get the multi-dimension padding of the deconvolution.
Definition: NvInfer.h:2358
void setDilationNd(Dims const &dilation) noexcept
Set the multi-dimension dilation of the deconvolution.
Definition: NvInfer.h:2384
void setPaddingMode(PaddingMode paddingMode) noexcept
Set the padding mode.
Definition: NvInfer.h:2264
void setNbGroups(int64_t nbGroups) noexcept
Set the number of groups for a deconvolution.
Definition: NvInfer.h:2137
Dims getPrePadding() const noexcept
Get the pre-padding.
Definition: NvInfer.h:2223
PaddingMode getPaddingMode() const noexcept
Get the padding mode.
Definition: NvInfer.h:2276
A Dequantize layer in a network definition.
Definition: NvInfer.h:5532
void setToType(DataType toType) noexcept
Set the Dequantize layer output type.
Definition: NvInfer.h:5569
virtual ~IDequantizeLayer() noexcept=default
int32_t getAxis() const noexcept
Get the quantization axis.
Definition: NvInfer.h:5542
DataType getToType() const noexcept
Return the Dequantize layer output type.
Definition: NvInfer.h:5581
void setAxis(int32_t axis) noexcept
Set the quantization axis.
Definition: NvInfer.h:5553
A network layer to perform dynamic quantization.
Definition: NvInfer.h:5609
int32_t getAxis() const noexcept
Get the axis along which blocking occurs.
Definition: NvInfer.h:5697
int32_t getBlockSize() const noexcept
Get the size of the quantization block.
Definition: NvInfer.h:5720
DataType getScaleType() const noexcept
Return the scale factors data type.
Definition: NvInfer.h:5674
void setScaleType(DataType scaleType) noexcept
Set the data type of the scale factors used to quantize the data.
Definition: NvInfer.h:5661
DataType getToType() const noexcept
Return DynamicQuantizeLayer’s quantized output type.
Definition: NvInfer.h:5649
virtual ~IDynamicQuantizeLayer() noexcept=default
void setToType(DataType toType) noexcept
Set DynamicQuantizeLayer’s quantized output type.
Definition: NvInfer.h:5636
void setAxis(int32_t axis) noexcept
Set the axis along which block quantization occurs.
Definition: NvInfer.h:5687
void setBlockSize(int32_t size) noexcept
Set the size of the quantization block.
Definition: NvInfer.h:5710
An Einsum layer in a network.
Definition: NvInfer.h:5767
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:5778
virtual ~IEinsumLayer() noexcept=default
char const * getEquation() const noexcept
Return the equation.
Definition: NvInfer.h:5788
A elementwise layer in a network definition.
Definition: NvInfer.h:2470
virtual ~IElementWiseLayer() noexcept=default
apiv::VElementWiseLayer * mImpl
Definition: NvInfer.h:2499
ElementWiseOperation getOperation() const noexcept
Get the binary operation for the layer.
Definition: NvInfer.h:2493
void setOperation(ElementWiseOperation op) noexcept
Set the binary operation for the layer.
Definition: NvInfer.h:2481
Generate a tensor according to a specified mode.
Definition: NvInfer.h:5055
bool isAlphaBetaInt64() const noexcept
Return true if alpha/beta have type int64, false if they have type double.
Definition: NvInfer.h:5287
FillOperation getOperation() const noexcept
Get the fill operation for the layer.
Definition: NvInfer.h:5101
void setOperation(FillOperation op) noexcept
Set the fill operation for the layer.
Definition: NvInfer.h:5091
DataType getToType() const noexcept
Get the fill layer output type.
Definition: NvInfer.h:5316
void setAlphaInt64(int64_t alpha) noexcept
Set the alpha parameter with int64 datatype.
Definition: NvInfer.h:5230
void setBetaInt64(int64_t beta) noexcept
Set the beta parameter with int64 datatype.
Definition: NvInfer.h:5264
void setBeta(double beta) noexcept
Set the beta parameter.
Definition: NvInfer.h:5154
int64_t getAlphaInt64() const noexcept
Get the value of alpha parameter with int64 datatype.
Definition: NvInfer.h:5245
int64_t getBetaInt64() const noexcept
Get the value of beta parameter with int64 datatype.
Definition: NvInfer.h:5279
double getAlpha() const noexcept
Get the value of alpha parameter.
Definition: NvInfer.h:5135
void setDimensions(Dims const &dimensions) noexcept
Set the output tensor's dimensions.
Definition: NvInfer.h:5066
void setAlpha(double alpha) noexcept
Set the alpha parameter.
Definition: NvInfer.h:5120
void setToType(DataType toType) noexcept
Set the fill layer output type.
Definition: NvInfer.h:5304
Dims getDimensions() const noexcept
Get the output tensor's dimensions.
Definition: NvInfer.h:5081
double getBeta() const noexcept
Get the value of beta parameter.
Definition: NvInfer.h:5169
virtual ~IFillLayer() noexcept=default
A Gather layer in a network definition. Supports several kinds of gathering.
Definition: NvInfer.h:2603
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:2614
void setNbElementWiseDims(int32_t elementWiseDims) noexcept
Set the number of leading dimensions of indices tensor to be handled elementwise.
Definition: NvInfer.h:2649
apiv::VGatherLayer * mImpl
Definition: NvInfer.h:2685
int32_t getNbElementWiseDims() const noexcept
Get the number of leading dimensions of indices tensor to be handled elementwise.
Definition: NvInfer.h:2659
void setMode(GatherMode mode) noexcept
Set the gather mode.
Definition: NvInfer.h:2669
int32_t getGatherAxis() const noexcept
Get the axis to gather on.
Definition: NvInfer.h:2626
GatherMode getMode() const noexcept
Get the gather mode.
Definition: NvInfer.h:2679
virtual ~IGatherLayer() noexcept=default
A GridSample layer in a network definition.
Definition: NvInfer.h:5988
void setInterpolationMode(InterpolationMode mode) noexcept
Set the grid sample interpolation mode.
Definition: NvInfer.h:5995
bool setSampleMode(SampleMode mode) noexcept
Set the sample mode.
Definition: NvInfer.h:6041
void setAlignCorners(bool alignCorners) noexcept
Set the align corners mode.
Definition: NvInfer.h:6017
apiv::VGridSampleLayer * mImpl
Definition: NvInfer.h:6059
SampleMode getSampleMode() const noexcept
Get the sample mode.
Definition: NvInfer.h:6053
InterpolationMode getInterpolationMode() const noexcept
Get the grid sample interpolation mode.
Definition: NvInfer.h:6007
bool getAlignCorners() const noexcept
Get the align corners mode.
Definition: NvInfer.h:6029
virtual ~IGridSampleLayer() noexcept=default
Class to handle library allocated memory that is accessible to the user.
Definition: NvInferRuntime.h:142
A layer that represents the identity function.
Definition: NvInfer.h:3794
apiv::VIdentityLayer * mImpl
Definition: NvInfer.h:3796
virtual ~IIdentityLayer() noexcept=default
This is a base class for Conditional boundary layers.
Definition: NvInfer.h:4449
IIfConditional * getConditional() const noexcept
Get a pointer to the IIfConditional associated with this boundary layer.
Definition: NvInfer.h:4454
virtual ~IIfConditionalBoundaryLayer() noexcept=default
Helper for constructing conditionally-executed subgraphs.
Definition: NvInfer.h:4531
IIfConditionalInputLayer * addInput(ITensor &input) noexcept
Add an If-conditional input.
Definition: NvInfer.h:4572
char const * getName() const noexcept
Return the name of the conditional.
Definition: NvInfer.h:4597
virtual ~IIfConditional() noexcept=default
IConditionLayer * setCondition(ITensor &condition) noexcept
Set the condition tensor for this If-Conditional construct.
Definition: NvInfer.h:4542
IIfConditionalOutputLayer * addOutput(ITensor &trueSubgraphOutput, ITensor &falseSubgraphOutput) noexcept
Add an If-conditional output.
Definition: NvInfer.h:4560
void setName(char const *name) noexcept
Set the name of the conditional.
Definition: NvInfer.h:4587
This layer represents an output of an IIfConditional.
Definition: NvInfer.h:4487
virtual ~IIfConditionalOutputLayer() noexcept=default
Application-implemented interface for calibration.
Definition: NvInfer.h:8126
virtual TRT_DEPRECATED int32_t getBatchSize() const noexcept=0
Get the batch size used for calibration batches.
A layer to do iterations.
Definition: NvInfer.h:4762
virtual ~IIteratorLayer() noexcept=default
void setReverse(bool reverse) noexcept
Set iteration order to be reverse.
Definition: NvInfer.h:4789
bool getReverse() const noexcept
Check if the iteration order is reverse.
Definition: NvInfer.h:4799
int32_t getAxis() const noexcept
Get axis being iterated over.
Definition: NvInfer.h:4775
void setAxis(int32_t axis) noexcept
Set axis to iterate over.
Definition: NvInfer.h:4767
A LRN layer in a network definition.
Definition: NvInfer.h:1712
int64_t getWindowSize() const noexcept
Get the LRN window size.
Definition: NvInfer.h:1733
float getAlpha() const noexcept
Get the LRN alpha value.
Definition: NvInfer.h:1755
void setWindowSize(int64_t windowSize) noexcept
Set the LRN window size.
Definition: NvInfer.h:1723
void setK(float k) noexcept
Set the LRN K value.
Definition: NvInfer.h:1789
void setAlpha(float alpha) noexcept
Set the LRN alpha value.
Definition: NvInfer.h:1745
void setBeta(float beta) noexcept
Set the LRN beta value.
Definition: NvInfer.h:1767
virtual ~ILRNLayer() noexcept=default
float getBeta() const noexcept
Get the LRN beta value.
Definition: NvInfer.h:1777
float getK() const noexcept
Get the LRN K value.
Definition: NvInfer.h:1799
Base class for all layer classes in a network definition.
Definition: NvInfer.h:555
bool precisionIsSet() const noexcept
whether the computational precision has been set for this layer
Definition: NvInfer.h:697
void setMetadata(char const *metadata) noexcept
Set the metadata for this layer.
Definition: NvInfer.h:813
void setPrecision(DataType dataType) noexcept
Set the preferred or required computational precision of this layer in a weakly-typed network.
Definition: NvInfer.h:673
void setName(char const *name) noexcept
Set the name of a layer.
Definition: NvInfer.h:576
void resetPrecision() noexcept
reset the computational precision for this layer
Definition: NvInfer.h:707
int32_t getNbInputs() const noexcept
Get the number of inputs of a layer.
Definition: NvInfer.h:594
char const * getMetadata() const noexcept
Get the metadata of the layer.
Definition: NvInfer.h:826
DataType getOutputType(int32_t index) const noexcept
get the output type of this layer
Definition: NvInfer.h:769
DataType getPrecision() const noexcept
get the computational precision of this layer
Definition: NvInfer.h:685
char const * getName() const noexcept
Return the name of a layer.
Definition: NvInfer.h:586
int32_t getNbOutputs() const noexcept
Get the number of outputs of a layer.
Definition: NvInfer.h:615
bool outputTypeIsSet(int32_t index) const noexcept
whether the output type has been set for this layer
Definition: NvInfer.h:783
ITensor * getOutput(int32_t index) const noexcept
Get the layer output corresponding to the given index.
Definition: NvInfer.h:625
void setInput(int32_t index, ITensor &tensor) noexcept
Replace an input of this layer with a specific tensor.
Definition: NvInfer.h:642
void resetOutputType(int32_t index) noexcept
reset the output type for this layer
Definition: NvInfer.h:795
ITensor * getInput(int32_t index) const noexcept
Get the layer input corresponding to the given index.
Definition: NvInfer.h:607
void setOutputType(int32_t index, DataType dataType) noexcept
Set the output type of this layer in a weakly-typed network.
Definition: NvInfer.h:754
LayerType getType() const noexcept
Return the type of a layer.
Definition: NvInfer.h:562
virtual ~ILayer() noexcept=default
Application-implemented logging interface for the builder, refitter and runtime.
Definition: NvInferRuntime.h:1542
This is a base class for Loop boundary layers.
Definition: NvInfer.h:4426
virtual ~ILoopBoundaryLayer() noexcept=default
ILoop * getLoop() const noexcept
Get a pointer to ILoop associated with this boundary layer.
Definition: NvInfer.h:4431
Helper for creating a recurrent subgraph.
Definition: NvInfer.h:4819
void setName(char const *name) noexcept
Set the name of the loop.
Definition: NvInfer.h:4889
ITripLimitLayer * addTripLimit(ITensor &tensor, TripLimit limit) noexcept
Add a trip-count limiter, based on the given tensor.
Definition: NvInfer.h:4848
IIteratorLayer * addIterator(ITensor &tensor, int32_t axis=0, bool reverse=false) noexcept
Return layer that subscripts tensor by loop iteration.
Definition: NvInfer.h:4861
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:4874
virtual ~ILoop() noexcept=default
char const * getName() const noexcept
Return the name of the loop.
Definition: NvInfer.h:4899
IRecurrenceLayer * addRecurrence(ITensor &initialValue) noexcept
Create a recurrence layer for this loop with initialValue as its first input.
Definition: NvInfer.h:4827
An ILoopOutputLayer is the sole way to get output from a loop.
Definition: NvInfer.h:4662
virtual ~ILoopOutputLayer() noexcept=default
int32_t getAxis() const noexcept
Get axis being concatenated over.
Definition: NvInfer.h:4692
LoopOutput getLoopOutput() const noexcept
Get which kind a loop output has.
Definition: NvInfer.h:4667
void setAxis(int32_t axis) noexcept
Set where to insert the contenation axis. Ignored if getLoopOutput() is kLAST_VALUE.
Definition: NvInfer.h:4684
Layer that represents a Matrix Multiplication.
Definition: NvInfer.h:3686
apiv::VMatrixMultiplyLayer * mImpl
Definition: NvInfer.h:3714
virtual ~IMatrixMultiplyLayer() noexcept=default
MatrixOperation getOperation(int32_t index) const noexcept
Get the operation for an input tensor.
Definition: NvInfer.h:3708
void setOperation(int32_t index, MatrixOperation op) noexcept
Set the operation for an input tensor.
Definition: NvInfer.h:3696
A non-maximum suppression layer in a network definition.
Definition: NvInfer.h:6136
virtual ~INMSLayer() noexcept=default
void setTopKBoxLimit(int32_t limit) noexcept
Set the TopK box limit parameter for the layer.
Definition: NvInfer.h:6173
void setBoundingBoxFormat(BoundingBoxFormat fmt) noexcept
Set the bounding box format parameter for the layer.
Definition: NvInfer.h:6147
BoundingBoxFormat getBoundingBoxFormat() const noexcept
Get the bounding box format parameter for the layer.
Definition: NvInfer.h:6159
apiv::VNMSLayer * mImpl
Definition: NvInfer.h:6209
int32_t getTopKBoxLimit() const noexcept
Get the TopK box limit parameter for the layer.
Definition: NvInfer.h:6183
A network definition for input to the builder.
Definition: NvInfer.h:6633
IConcatenationLayer * addConcatenation(ITensor *const *inputs, int32_t nbInputs) noexcept
Add a concatenation layer to the network.
Definition: NvInfer.h:6826
IShuffleLayer * addShuffle(ITensor &input) noexcept
Add a shuffle layer to the network.
Definition: NvInfer.h:6889
INormalizationLayer * addNormalization(ITensor &input, ITensor &scale, ITensor &bias, uint32_t axesMask) noexcept
Add a normalization layer to the network.
Definition: NvInfer.h:7966
void setName(char const *name) noexcept
Sets the name of the network.
Definition: NvInfer.h:7299
bool markDebug(ITensor &tensor) noexcept
Mark a tensor as a debug tensor.
Definition: NvInfer.h:6705
ILRNLayer * addLRN(ITensor &input, int64_t window, float alpha, float beta, float k) noexcept
Add a LRN layer to the network.
Definition: NvInfer.h:6770
ITopKLayer * addTopK(ITensor &input, TopKOperation op, int32_t k, uint32_t reduceAxes) noexcept
Add a TopK layer to the network.
Definition: NvInfer.h:7049
ICumulativeLayer * addCumulative(ITensor &input, ITensor &axis, CumulativeOperation operation, bool exclusive, bool reverse) noexcept
Add a cumulative layer to the network.
Definition: NvInfer.h:7988
IAssertionLayer * addAssertion(ITensor &condition, char const *message) noexcept
Add an assertion layer to the network.
Definition: NvInfer.h:7615
IConvolutionLayer * addConvolutionNd(ITensor &input, int64_t nbOutputMaps, Dims const &kernelSize, Weights kernelWeights, Weights biasWeights) noexcept
Add a multi-dimension convolution layer to the network.
Definition: NvInfer.h:7434
ICastLayer * addCast(ITensor &input, DataType toType) noexcept
Add a cast layer.
Definition: NvInfer.h:7189
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:7513
char const * getName() const noexcept
Returns the name associated with the network.
Definition: NvInfer.h:7313
IParametricReLULayer * addParametricReLU(ITensor &input, ITensor &slope) noexcept
Add a parametric ReLU layer to the network.
Definition: NvInfer.h:7412
ITensor * getOutput(int32_t index) const noexcept
Get the output tensor specified by the given index.
Definition: NvInfer.h:6990
ITensor * getInput(int32_t index) const noexcept
Get the input tensor specified by the given index.
Definition: NvInfer.h:6960
IDequantizeLayer * addDequantize(ITensor &input, ITensor &scale, DataType outputType) noexcept
Add a dequantization layer to the network.
Definition: NvInfer.h:7783
bool unmarkOutputForShapes(ITensor &tensor) noexcept
Undo markOutputForShapes.
Definition: NvInfer.h:7394
IFillLayer * addFill(Dims const &dimensions, FillOperation op, DataType outputType) noexcept
Add a fill layer to the network.
Definition: NvInfer.h:7666
ILoop * addLoop() noexcept
Add a loop to the network.
Definition: NvInfer.h:7544
IDynamicQuantizeLayer * addDynamicQuantize(ITensor &input, int32_t axis, int32_t blockSize, DataType outputType, DataType scaleType) noexcept
Add a dynamic quantization layer to the network.
Definition: NvInfer.h:7871
IActivationLayer * addActivation(ITensor &input, ActivationType type) noexcept
Add an activation layer to the network.
Definition: NvInfer.h:6751
TRT_DEPRECATED IFillLayer * addFill(Dims const &dimensions, FillOperation op) noexcept
Add a fill layer to the network.
Definition: NvInfer.h:7640
ISliceLayer * addSlice(ITensor &input, Dims const &start, Dims const &size, Dims const &stride) noexcept
Add a slice layer to the network.
Definition: NvInfer.h:7275
virtual ~INetworkDefinition() noexcept=default
TRT_DEPRECATED IQuantizeLayer * addQuantize(ITensor &input, ITensor &scale) noexcept
Add a quantization layer to the network.
Definition: NvInfer.h:7824
virtual IBuilder & getBuilder() const noexcept
Return the builder from which this INetworkDefinition was created.
Definition: NvInfer.h:7999
INMSLayer * addNMS(ITensor &boxes, ITensor &scores, ITensor &maxOutputBoxesPerClass) noexcept
Add a non-maximum suppression layer to the network.
Definition: NvInfer.h:7923
ILayer * getLayer(int32_t index) const noexcept
Get the layer specified by the given index.
Definition: NvInfer.h:6932
bool getFlag(NetworkDefinitionCreationFlag networkDefinitionCreationFlag) const noexcept
Returns true if the network definition creation flag is set.
Definition: NvInfer.h:7365
IIfConditional * addIfConditional() noexcept
Add an if-then-else to the network.
Definition: NvInfer.h:7559
IErrorRecorder * getErrorRecorder() const noexcept
get the ErrorRecorder assigned to this interface.
Definition: NvInfer.h:7740
ISqueezeLayer * addSqueeze(ITensor &input, ITensor &axes) noexcept
Add a squeeze layer to the network.
Definition: NvInfer.h:8056
IReverseSequenceLayer * addReverseSequence(ITensor &input, ITensor &sequenceLens) noexcept
Add a ReverseSequence layer to the network.
Definition: NvInfer.h:7940
int32_t getNbInputs() const noexcept
Get the number of inputs in the network.
Definition: NvInfer.h:6944
NetworkDefinitionCreationFlags getFlags() const noexcept
Get the network definition creation flags for this network definition object. Defaults to 0.
Definition: NvInfer.h:7353
IQuantizeLayer * addQuantize(ITensor &input, ITensor &scale, DataType outputType) noexcept
Add a quantization layer to the network.
Definition: NvInfer.h:7846
IReduceLayer * addReduce(ITensor &input, ReduceOperation operation, uint32_t reduceAxes, bool keepDimensions) noexcept
Add a reduce layer to the network.
Definition: NvInfer.h:7016
IUnaryLayer * addUnary(ITensor &input, UnaryOperation operation) noexcept
Add a unary layer to the network.
Definition: NvInfer.h:6875
IGridSampleLayer * addGridSample(ITensor &input, ITensor &grid) noexcept
Add a GridSample layer to the network.
Definition: NvInfer.h:7905
void removeTensor(ITensor &tensor) noexcept
remove a tensor from the network definition.
Definition: NvInfer.h:7204
bool areWeightsMarkedRefittable(char const *name) const noexcept
Whether the weight has been marked as refittable.
Definition: NvInfer.h:8037
ISelectLayer * addSelect(ITensor &condition, ITensor &thenInput, ITensor &elseInput) noexcept
Add a select layer to the network.
Definition: NvInfer.h:7598
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:7803
int32_t getNbLayers() const noexcept
Get the number of layers in the network.
Definition: NvInfer.h:6918
TRT_DEPRECATED bool hasImplicitBatchDimension() const noexcept
Query whether the network was created with an implicit batch dimension.
Definition: NvInfer.h:7343
apiv::VNetworkDefinition * mImpl
Definition: NvInfer.h:8083
bool markOutputForShapes(ITensor &tensor) noexcept
Enable tensor's value to be computed by IExecutionContext::getShapeBinding.
Definition: NvInfer.h:7382
IOneHotLayer * addOneHot(ITensor &indices, ITensor &values, ITensor &depth, int32_t axis) noexcept
Add a OneHot layer to the network.
Definition: NvInfer.h:6906
IScaleLayer * addScale(ITensor &input, ScaleMode mode, Weights shift, Weights scale, Weights power) noexcept
Add a Scale layer to the network.
Definition: NvInfer.h:6796
IPluginV3Layer * addPluginV3(ITensor *const *inputs, int32_t nbInputs, ITensor *const *shapeInputs, int32_t nbShapeInputs, IPluginV3 &plugin) noexcept
Add a plugin layer implementing the IPluginV3 interface to the network.
Definition: NvInfer.h:7255
void unmarkOutput(ITensor &tensor) noexcept
unmark a tensor as a network output.
Definition: NvInfer.h:7216
IIdentityLayer * addIdentity(ITensor &input) noexcept
Add an identity layer.
Definition: NvInfer.h:7174
IGatherLayer * addGatherV2(ITensor &data, ITensor &indices, GatherMode mode) noexcept
Add gather with specified mode, axis=0 and nbElementWiseDims=0.
Definition: NvInfer.h:7081
IElementWiseLayer * addElementWise(ITensor &input1, ITensor &input2, ElementWiseOperation op) noexcept
Add an elementwise layer to the network.
Definition: NvInfer.h:6853
IConstantLayer * addConstant(Dims const &dimensions, Weights weights) noexcept
Add a constant layer to the network.
Definition: NvInfer.h:7160
void setErrorRecorder(IErrorRecorder *recorder) noexcept
Set the ErrorRecorder for this interface.
Definition: NvInfer.h:7725
IPoolingLayer * addPoolingNd(ITensor &input, PoolingType type, Dims const &windowSize) noexcept
Add a multi-dimension pooling layer to the network.
Definition: NvInfer.h:7454
IRaggedSoftMaxLayer * addRaggedSoftMax(ITensor &input, ITensor &bounds) noexcept
Add a RaggedSoftMax layer to the network.
Definition: NvInfer.h:7100
IShapeLayer * addShape(ITensor &input) noexcept
Add a shape layer to the network.
Definition: NvInfer.h:7329
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:7065
bool unmarkWeightsRefittable(char const *name) noexcept
Unmark weights as refittable when the builder flag kREFIT_INDIVIDUAL is set.
Definition: NvInfer.h:8024
bool markWeightsRefittable(char const *name) noexcept
Mark weights as refittable when the builder flag kREFIT_INDIVIDUAL is set.
Definition: NvInfer.h:8012
IDeconvolutionLayer * addDeconvolutionNd(ITensor &input, int64_t nbOutputMaps, Dims kernelSize, Weights kernelWeights, Weights biasWeights) noexcept
Add a multi-dimension deconvolution layer to the network.
Definition: NvInfer.h:7476
IResizeLayer * addResize(ITensor &input) noexcept
Add a resize layer to the network.
Definition: NvInfer.h:7530
IUnsqueezeLayer * addUnsqueeze(ITensor &input, ITensor &axes) noexcept
Add an unsqueeze layer to the network.
Definition: NvInfer.h:8077
IMatrixMultiplyLayer * addMatrixMultiply(ITensor &input0, MatrixOperation op0, ITensor &input1, MatrixOperation op1) noexcept
Add a MatrixMultiply layer to the network.
Definition: NvInfer.h:7121
ISoftMaxLayer * addSoftMax(ITensor &input) noexcept
Add a SoftMax layer to the network.
Definition: NvInfer.h:6809
bool isDebugTensor(nvinfer1::ITensor const &tensor) const noexcept
Check if a tensor is marked as debug tensor.
Definition: NvInfer.h:6731
bool unmarkDebug(ITensor &tensor) noexcept
Unmark a tensor as a debug tensor.
Definition: NvInfer.h:6721
IEinsumLayer * addEinsum(ITensor *const *inputs, int32_t nbInputs, char const *equation) noexcept
Add an Einsum layer to the network.
Definition: NvInfer.h:7887
void markOutput(ITensor &tensor) noexcept
Mark a tensor as a network output.
Definition: NvInfer.h:6687
TRT_DEPRECATED IPluginV2Layer * addPluginV2(ITensor *const *inputs, int32_t nbInputs, IPluginV2 &plugin) noexcept
Add a plugin layer to the network using the IPluginV2 interface.
Definition: NvInfer.h:7237
IPaddingLayer * addPaddingNd(ITensor &input, Dims const &prePadding, Dims const &postPadding) noexcept
Add a padding layer to the network. Only 2D padding is currently supported.
Definition: NvInfer.h:7682
INonZeroLayer * addNonZero(ITensor &input) noexcept
Add a nonzero layer to the network.
Definition: NvInfer.h:7136
TRT_DEPRECATED IDequantizeLayer * addDequantize(ITensor &input, ITensor &scale) noexcept
Add a dequantization layer to the network.
Definition: NvInfer.h:7761
int32_t getNbOutputs() const noexcept
Get the number of outputs in the network.
Definition: NvInfer.h:6974
bool setWeightsName(Weights weights, char const *name) noexcept
Associate a name with all current uses of the given weights.
Definition: NvInfer.h:7706
Forward declaration of IEngineInspector for use by other interfaces.
Definition: NvInferRuntime.h:51
Definition: NvInfer.h:3740
virtual ~INonZeroLayer() noexcept=default
A normalization layer in a network definition.
Definition: NvInfer.h:6298
float getEpsilon() const noexcept
Get the epsilon value used for the normalization calculation.
Definition: NvInfer.h:6317
uint32_t getAxes() const noexcept
Get the axes value used for the normalization calculation.
Definition: NvInfer.h:6337
virtual ~INormalizationLayer() noexcept=default
void setEpsilon(float eps) noexcept
Set the epsilon value used for the normalization calculation.
Definition: NvInfer.h:6307
DataType getComputePrecision() const noexcept
Get the compute precision of this layer.
Definition: NvInfer.h:6404
apiv::VNormalizationLayer * mImpl
Definition: NvInfer.h:6410
int64_t getNbGroups() const noexcept
Get the number of groups used to split the channels for the normalization calculation.
Definition: NvInfer.h:6368
void setAxes(uint32_t axesMask) noexcept
Set the reduction axes for the normalization calculation.
Definition: NvInfer.h:6327
void setComputePrecision(DataType type) noexcept
Set the compute precision of this layer.
Definition: NvInfer.h:6394
void setNbGroups(int64_t nbGroups) noexcept
Set the number of groups used to split the channels in the normalization calculation.
Definition: NvInfer.h:6358
A OneHot layer in a network definition.
Definition: NvInfer.h:5952
apiv::VOneHotLayer * mImpl
Definition: NvInfer.h:5973
void setAxis(int32_t axis) noexcept
Set the axis parameter.
Definition: NvInfer.h:5959
int32_t getAxis() const noexcept
Get the value of the axis parameter.
Definition: NvInfer.h:5967
Optimization profile for dynamic input dimensions and shape tensors.
Definition: NvInferRuntime.h:2618
Layer that represents a padding operation.
Definition: NvInfer.h:2964
Dims getPostPaddingNd() const noexcept
Get the padding that is applied at the end of the tensor.
Definition: NvInfer.h:3013
void setPrePaddingNd(Dims const &padding) noexcept
Set the padding that is applied at the start of the tensor.
Definition: NvInfer.h:2975
virtual ~IPaddingLayer() noexcept=default
void setPostPaddingNd(Dims const &padding) noexcept
Set the padding that is applied at the end of the tensor.
Definition: NvInfer.h:3001
Dims getPrePaddingNd() const noexcept
Get the padding that is applied at the start of the tensor.
Definition: NvInfer.h:2987
apiv::VPaddingLayer * mImpl
Definition: NvInfer.h:3019
Layer that represents a parametric ReLU operation.
Definition: NvInfer.h:3910
apiv::VParametricReLULayer * mImpl
Definition: NvInfer.h:3912
virtual ~IParametricReLULayer() noexcept=default
Single registration point for all plugins in an application. It is used to find plugin implementation...
Definition: NvInferRuntimeCommon.h:56
Plugin class for user-implemented layers.
Definition: NvInferRuntimePlugin.h:133
Layer type for pluginV2.
Definition: NvInfer.h:2701
virtual ~IPluginV2Layer() noexcept=default
apiv::VPluginV2Layer * mImpl
Definition: NvInfer.h:2714
IPluginV2 & getPlugin() noexcept
Get the plugin for the layer.
Definition: NvInfer.h:2708
Layer type for V3 plugins.
Definition: NvInfer.h:2728
virtual ~IPluginV3Layer() noexcept=default
IPluginV3 & getPlugin() noexcept
Get the plugin for the layer.
Definition: NvInfer.h:2735
apiv::VPluginV3Layer * mImpl
Definition: NvInfer.h:2741
A Pooling layer in a network definition.
Definition: NvInfer.h:1461
PoolingType getPoolingType() const noexcept
Get the type of activation to be performed.
Definition: NvInfer.h:1480
PaddingMode getPaddingMode() const noexcept
Get the padding mode.
Definition: NvInfer.h:1613
Dims getPostPadding() const noexcept
Get the padding.
Definition: NvInfer.h:1589
bool getAverageCountExcludesPadding() const noexcept
Get whether average pooling uses as a denominator the overlap area between the window and the unpadde...
Definition: NvInfer.h:1533
Dims getPrePadding() const noexcept
Get the pre-padding.
Definition: NvInfer.h:1561
void setPoolingType(PoolingType type) noexcept
Set the type of activation to be performed.
Definition: NvInfer.h:1470
void setWindowSizeNd(Dims const &windowSize) noexcept
Set the multi-dimension window size for pooling.
Definition: NvInfer.h:1626
void setPaddingMode(PaddingMode paddingMode) noexcept
Set the padding mode.
Definition: NvInfer.h:1602
Dims getWindowSizeNd() const noexcept
Get the multi-dimension window size for pooling.
Definition: NvInfer.h:1636
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:1522
void setPaddingNd(Dims const &padding) noexcept
Set the multi-dimension padding for pooling.
Definition: NvInfer.h:1680
float getBlendFactor() const noexcept
Get the blending factor for the max_average_blend mode: max_average_blendPool = (1-blendFactor)*maxPo...
Definition: NvInfer.h:1508
void setStrideNd(Dims const &stride) noexcept
Set the multi-dimension stride for pooling.
Definition: NvInfer.h:1651
Dims getStrideNd() const noexcept
Get the multi-dimension stride for pooling.
Definition: NvInfer.h:1661
virtual ~IPoolingLayer() noexcept=default
Dims getPaddingNd() const noexcept
Get the multi-dimension padding for pooling.
Definition: NvInfer.h:1692
void setPostPadding(Dims const &padding) noexcept
Set the multi-dimension post-padding for pooling.
Definition: NvInfer.h:1579
void setPrePadding(Dims const &padding) noexcept
Set the multi-dimension pre-padding for pooling.
Definition: NvInfer.h:1551
void setBlendFactor(float blendFactor) noexcept
Set the blending factor for the max_average_blend mode: max_average_blendPool = (1-blendFactor)*maxPo...
Definition: NvInfer.h:1495
A Quantize layer in a network definition.
Definition: NvInfer.h:5401
void setToType(DataType toType) noexcept
Set the Quantize layer output type.
Definition: NvInfer.h:5438
void setAxis(int32_t axis) noexcept
Set the quantization axis.
Definition: NvInfer.h:5422
int32_t getAxis() const noexcept
Get the quantization axis.
Definition: NvInfer.h:5411
virtual ~IQuantizeLayer() noexcept=default
DataType getToType() const noexcept
Return the Quantize layer output type.
Definition: NvInfer.h:5450
A RaggedSoftmax layer in a network definition.
Definition: NvInfer.h:3761
apiv::VRaggedSoftMaxLayer * mImpl
Definition: NvInfer.h:3763
virtual ~IRaggedSoftMaxLayer() noexcept=default
A recurrence layer in a network definition.
Definition: NvInfer.h:4615
virtual ~IRecurrenceLayer() noexcept=default
Layer that represents a reduction across a non-bool tensor.
Definition: NvInfer.h:2884
void setKeepDimensions(bool keepDimensions) noexcept
Set the boolean that specifies whether or not to keep the reduced dimensions for the layer.
Definition: NvInfer.h:2931
void setOperation(ReduceOperation op) noexcept
Set the reduce operation for the layer.
Definition: NvInfer.h:2891
ReduceOperation getOperation() const noexcept
Get the reduce operation for the layer.
Definition: NvInfer.h:2901
virtual ~IReduceLayer() noexcept=default
uint32_t getReduceAxes() const noexcept
Get the axes over which to reduce for the layer.
Definition: NvInfer.h:2921
void setReduceAxes(uint32_t reduceAxes) noexcept
Set the axes over which to reduce.
Definition: NvInfer.h:2911
apiv::VReduceLayer * mImpl
Definition: NvInfer.h:2947
bool getKeepDimensions() const noexcept
Get the boolean that specifies whether or not to keep the reduced dimensions for the layer.
Definition: NvInfer.h:2941
A resize layer in a network definition.
Definition: NvInfer.h:4099
void setSelectorForSinglePixel(ResizeSelector selector) noexcept
Set coordinate selector function when resized to single pixel.
Definition: NvInfer.h:4260
void setNearestRounding(ResizeRoundMode value) noexcept
Set rounding mode for nearest neighbor resize.
Definition: NvInfer.h:4284
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:4178
void setOutputDimensions(Dims const &dimensions) noexcept
Set the output dimensions.
Definition: NvInfer.h:4119
void setCubicCoeff(float A) noexcept
Set the coefficient 'A' used in cubic interpolation.
Definition: NvInfer.h:4316
void setScales(float const *scales, int32_t nbScales) noexcept
Set the resize scales.
Definition: NvInfer.h:4159
float getCubicCoeff() const noexcept
Get the coefficient 'A' used in cubic interpolation.
Definition: NvInfer.h:4326
ResizeSelector getSelectorForSinglePixel() const noexcept
Get the coordinate selector function when resized to single pixel.
Definition: NvInfer.h:4270
InterpolationMode getResizeMode() const noexcept
Get resize mode for an input tensor.
Definition: NvInfer.h:4200
void setCoordinateTransformation(ResizeCoordinateTransformation coordTransform) noexcept
Set coordinate transformation function.
Definition: NvInfer.h:4235
void setExcludeOutside(bool excludeFlag) noexcept
Set the state for excluding outside pixels.
Definition: NvInfer.h:4339
void setResizeMode(InterpolationMode interpolationMode) noexcept
Set resize mode for an input tensor.
Definition: NvInfer.h:4190
Dims getOutputDimensions() const noexcept
Get the output dimensions.
Definition: NvInfer.h:4129
ResizeRoundMode getNearestRounding() const noexcept
Get rounding mode for nearest neighbor resize.
Definition: NvInfer.h:4294
bool getExcludeOutside() const noexcept
Get the state for excluding outside pixels.
Definition: NvInfer.h:4349
ResizeCoordinateTransformation getCoordinateTransformation() const noexcept
Get coordinate transformation function.
Definition: NvInfer.h:4245
A ReverseSequence layer in a network definition.
Definition: NvInfer.h:6226
void setSequenceAxis(int32_t sequenceAxis) noexcept
Set the sequence axis. Default is 0.
Definition: NvInfer.h:6259
int32_t getBatchAxis() const noexcept
Return the batch axis. Return 1 if no batch axis was set.
Definition: NvInfer.h:6246
apiv::VReverseSequenceLayer * mImpl
Definition: NvInfer.h:6275
int32_t getSequenceAxis() const noexcept
Return the sequence axis. Return 0 if no sequence axis was set.
Definition: NvInfer.h:6269
void setBatchAxis(int32_t batchAxis) noexcept
Set the batch axis. Default is 1.
Definition: NvInfer.h:6236
virtual ~IReverseSequenceLayer() noexcept=default
A Scale layer in a network definition.
Definition: NvInfer.h:1858
Weights getScale() const noexcept
Get the scale value.
Definition: NvInfer.h:1915
Weights getPower() const noexcept
Get the power value.
Definition: NvInfer.h:1935
void setScale(Weights scale) noexcept
Set the scale value.
Definition: NvInfer.h:1905
void setPower(Weights power) noexcept
Set the power value.
Definition: NvInfer.h:1925
ScaleMode getMode() const noexcept
Get the scale mode.
Definition: NvInfer.h:1875
void setShift(Weights shift) noexcept
Set the shift value.
Definition: NvInfer.h:1885
void setChannelAxis(int32_t channelAxis) noexcept
Set the channel axis.
Definition: NvInfer.h:1971
Weights getShift() const noexcept
Get the shift value.
Definition: NvInfer.h:1895
virtual ~IScaleLayer() noexcept=default
void setMode(ScaleMode mode) noexcept
Set the scale mode.
Definition: NvInfer.h:1865
int32_t getChannelAxis() const noexcept
Get the channel axis.
Definition: NvInfer.h:1950
A scatter layer in a network definition. Supports several kinds of scattering.
Definition: NvInfer.h:5880
void setMode(ScatterMode mode) noexcept
Set the scatter mode.
Definition: NvInfer.h:5887
apiv::VScatterLayer * mImpl
Definition: NvInfer.h:5921
void setAxis(int32_t axis) noexcept
Set the axis used by ScatterMode::kELEMENTS.
Definition: NvInfer.h:5907
int32_t getAxis() const noexcept
Get the axis.
Definition: NvInfer.h:5915
ScatterMode getMode() const noexcept
Get the scatter mode.
Definition: NvInfer.h:5897
virtual ~IScatterLayer() noexcept=default
Select elements from two data tensors based on a condition tensor.
Definition: NvInfer.h:4922
virtual ~ISelectLayer() noexcept=default
Layer type for getting shape of a tensor.
Definition: NvInfer.h:3489
virtual ~IShapeLayer() noexcept=default
apiv::VShapeLayer * mImpl
Definition: NvInfer.h:3491
Layer type for shuffling data.
Definition: NvInfer.h:3052
apiv::VShuffleLayer * mImpl
Definition: NvInfer.h:3210
void setFirstTranspose(Permutation permutation) noexcept
Set the permutation applied by the first transpose operation.
Definition: NvInfer.h:3063
void setSecondTranspose(Permutation permutation) noexcept
Set the permutation applied by the second transpose operation.
Definition: NvInfer.h:3163
Dims getReshapeDimensions() const noexcept
Get the reshaped dimensions.
Definition: NvInfer.h:3116
void setReshapeDimensions(Dims const &dimensions) noexcept
Set the reshaped dimensions.
Definition: NvInfer.h:3103
Permutation getFirstTranspose() const noexcept
Get the permutation applied by the first transpose operation.
Definition: NvInfer.h:3075
virtual ~IShuffleLayer() noexcept=default
Permutation getSecondTranspose() const noexcept
Get the permutation applied by the second transpose operation.
Definition: NvInfer.h:3175
bool getZeroIsPlaceholder() const noexcept
Get meaning of 0 in reshape dimensions.
Definition: NvInfer.h:3204
void setZeroIsPlaceholder(bool zeroIsPlaceholder) noexcept
Set meaning of 0 in reshape dimensions.
Definition: NvInfer.h:3191
Slices an input tensor into an output tensor based on the offset and strides.
Definition: NvInfer.h:3304
void setStride(Dims const &stride) noexcept
Set the stride for computing the output slice data.
Definition: NvInfer.h:3373
apiv::VSliceLayer * mImpl
Definition: NvInfer.h:3472
virtual ~ISliceLayer() noexcept=default
void setSize(Dims const &size) noexcept
Set the dimensions of the output slice.
Definition: NvInfer.h:3344
void setAxes(Dims const &axes) noexcept
Set the axes for this ISliceLayer.
Definition: NvInfer.h:3451
void setStart(Dims const &start) noexcept
Set the start offset that the slice layer uses to create the output slice.
Definition: NvInfer.h:3315
Dims getStart() const noexcept
Get the start offset for the slice layer.
Definition: NvInfer.h:3330
void setMode(SampleMode mode) noexcept
Set the slice mode.
Definition: NvInfer.h:3398
Dims getSize() const noexcept
Get dimensions of the output slice.
Definition: NvInfer.h:3359
SampleMode getMode() const noexcept
Get the slice mode.
Definition: NvInfer.h:3408
Dims getStride() const noexcept
Get the stride for the output slice.
Definition: NvInfer.h:3388
Dims getAxes() const noexcept
Get the axes for this ISliceLayer.
Definition: NvInfer.h:3466
A Softmax layer in a network definition.
Definition: NvInfer.h:2002
void setAxes(uint32_t axes) noexcept
Set the axis along which softmax is computed. Currently, only one axis can be set.
Definition: NvInfer.h:2024
uint32_t getAxes() const noexcept
Get the axis along which softmax occurs.
Definition: NvInfer.h:2034
virtual ~ISoftMaxLayer() noexcept=default
Layer that represents a squeeze operation, removing unit dimensions of the input tensor on a set of a...
Definition: NvInfer.h:6424
virtual ~ISqueezeLayer() noexcept=default
apiv::VSqueezeLayer * mImpl
Definition: NvInfer.h:6441
A tensor in a network definition.
Definition: NvInfer.h:185
void setAllowedFormats(TensorFormats formats) noexcept
Set allowed formats for an input or output tensor. By default all formats are allowed....
Definition: NvInfer.h:431
TensorLocation getLocation() const noexcept
Get the storage location of a tensor.
Definition: NvInfer.h:350
void setDimensions(Dims const &dimensions) noexcept
Set the dimensions of a tensor.
Definition: NvInfer.h:233
void resetDynamicRange() noexcept
Undo effect of setDynamicRange.
Definition: NvInfer.h:389
void setName(char const *name) noexcept
Set the tensor name.
Definition: NvInfer.h:202
bool isExecutionTensor() const noexcept
Whether the tensor is an execution tensor.
Definition: NvInfer.h:496
void setType(DataType type) noexcept
Set the data type of a tensor.
Definition: NvInfer.h:262
TRT_DEPRECATED bool dynamicRangeIsSet() const noexcept
Query whether dynamic range is set.
Definition: NvInfer.h:381
char const * getName() const noexcept
Get the tensor name.
Definition: NvInfer.h:214
bool isShapeTensor() const noexcept
Whether the tensor is a shape tensor.
Definition: NvInfer.h:475
float getDynamicRangeMax() const noexcept
Get maximum of dynamic range.
Definition: NvInfer.h:409
bool isNetworkInput() const noexcept
Whether the tensor is a network input.
Definition: NvInfer.h:299
TRT_DEPRECATED void setBroadcastAcrossBatch(bool broadcastAcrossBatch) noexcept
Set whether to enable broadcast of tensor across the implicit batch dimension.
Definition: NvInfer.h:324
TRT_DEPRECATED bool setDynamicRange(float min, float max) noexcept
Set dynamic range for the tensor.
Definition: NvInfer.h:291
TRT_DEPRECATED bool getBroadcastAcrossBatch() const noexcept
Check if tensor is broadcast across the implicit batch dimension.
Definition: NvInfer.h:338
bool isNetworkOutput() const noexcept
Whether the tensor is a network output.
Definition: NvInfer.h:307
DataType getType() const noexcept
Get the data type of a tensor.
Definition: NvInfer.h:274
apiv::VTensor * mImpl
Definition: NvInfer.h:543
float getDynamicRangeMin() const noexcept
Get minimum of dynamic range.
Definition: NvInfer.h:399
virtual ~ITensor() noexcept=default
void setDimensionName(int32_t index, char const *name) noexcept
Name a dimension of an input tensor.
Definition: NvInfer.h:522
char const * getDimensionName(int32_t index) const noexcept
Get the name of an input dimension.
Definition: NvInfer.h:537
TRT_DEPRECATED void setLocation(TensorLocation location) noexcept
Set the storage location of a tensor.
Definition: NvInfer.h:369
Dims getDimensions() const noexcept
Get the dimensions of a tensor.
Definition: NvInfer.h:247
TensorFormats getAllowedFormats() const noexcept
Get a bitmask of TensorFormat values that the tensor supports. For a shape tensor,...
Definition: NvInfer.h:444
Class to handle tactic timing info collected from builder.
Definition: NvInfer.h:8973
int64_t queryKeys(TimingCacheKey *keyBuffer, int64_t capacity) const noexcept
Query cache keys from Timing Cache.
Definition: NvInfer.h:9039
bool combine(ITimingCache const &inputCache, bool ignoreMismatch) noexcept
Combine input timing cache into local instance.
Definition: NvInfer.h:9010
TimingCacheValue query(TimingCacheKey const &key) const noexcept
Query value in a cache entry.
Definition: NvInfer.h:9056
virtual ~ITimingCache() noexcept=default
bool update(TimingCacheKey const &key, TimingCacheValue const &value) noexcept
Update values in a cache entry.
Definition: NvInfer.h:9078
apiv::VTimingCache * mImpl
Definition: NvInfer.h:9084
bool reset() noexcept
Empty the timing cache.
Definition: NvInfer.h:9020
Layer that represents a TopK reduction.
Definition: NvInfer.h:3529
void setK(int32_t k) noexcept
Set the static k value for the layer.
Definition: NvInfer.h:3560
void setReduceAxes(uint32_t reduceAxes) noexcept
Set which axes to reduce for the layer.
Definition: NvInfer.h:3584
TopKOperation getOperation() const noexcept
Get the operation for the layer.
Definition: NvInfer.h:3546
apiv::VTopKLayer * mImpl
Definition: NvInfer.h:3616
void setOperation(TopKOperation op) noexcept
Set the operation for the layer.
Definition: NvInfer.h:3536
int32_t getK() const noexcept
Get the k value for the layer.
Definition: NvInfer.h:3574
uint32_t getReduceAxes() const noexcept
Get the axes to reduce for the layer.
Definition: NvInfer.h:3594
virtual ~ITopKLayer() noexcept=default
A layer that represents a trip-count limiter.
Definition: NvInfer.h:4736
TripLimit getTripLimit() const noexcept
Get a trip limiter type.
Definition: NvInfer.h:4741
virtual ~ITripLimitLayer() noexcept=default
Layer that represents an unary operation.
Definition: NvInfer.h:2809
void setOperation(UnaryOperation op) noexcept
Set the unary operation for the layer.
Definition: NvInfer.h:2818
apiv::VUnaryLayer * mImpl
Definition: NvInfer.h:2834
UnaryOperation getOperation() const noexcept
Get the unary operation for the layer.
Definition: NvInfer.h:2828
virtual ~IUnaryLayer() noexcept=default
Layer that represents an unsqueeze operation, which reshapes the input tensor by inserting unit-lengt...
Definition: NvInfer.h:6453
virtual ~IUnsqueezeLayer() noexcept=default
apiv::VUnsqueezeLayer * mImpl
Definition: NvInfer.h:6470
An Interface class for version control.
Definition: NvInferRuntimeBase.h:264
Version information associated with a TRT interface.
Definition: NvInferRuntimeBase.h:229
An array of weights used as a layer parameter.
Definition: NvInferRuntime.h:124
Definition: NvInfer.h:8605
virtual int32_t selectAlgorithms(IAlgorithmContext const &context, IAlgorithm const *const *choices, int32_t nbChoices, int32_t *selection) noexcept=0
Select Algorithms for a layer from the given list of algorithm choices.
virtual void reportAlgorithms(IAlgorithmContext const *const *algoContexts, IAlgorithm const *const *algoChoices, int32_t nbAlgorithms) noexcept=0
Called by TensorRT to report choices it made.
InterfaceInfo getInterfaceInfo() const noexcept override
Return version information associated with this interface. Applications must not override this method...
Definition: NvInfer.h:8610
virtual ~IAlgorithmSelector() noexcept=default
Definition: NvInferRuntimeBase.h:401
Definition: NvInferRuntime.h:1610
Definition: NvInfer.h:8232
~IInt8EntropyCalibrator2() noexcept override=default
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:8245
InterfaceInfo getInterfaceInfo() const noexcept override
Return version information associated with this interface. Applications must not override this method...
Definition: NvInfer.h:8237
Definition: NvInfer.h:8192
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:8205
InterfaceInfo getInterfaceInfo() const noexcept override
Return version information associated with this interface. Applications must not override this method...
Definition: NvInfer.h:8197
~IInt8EntropyCalibrator() noexcept override=default
Definition: NvInfer.h:8311
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:8324
InterfaceInfo getInterfaceInfo() const noexcept override
Return version information associated with this interface. Applications must not override this method...
Definition: NvInfer.h:8316
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:8272
~IInt8MinMaxCalibrator() noexcept override=default
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:8285
InterfaceInfo getInterfaceInfo() const noexcept override
Return version information associated with this interface. Applications must not override this method...
Definition: NvInfer.h:8277
Definition: NvInferPluginBase.h:206
Definition: NvInfer.h:9292
virtual bool stepComplete(char const *phaseName, int32_t step) noexcept=0
Signal that a step of an optimizer phase has finished.
virtual ~IProgressMonitor() noexcept=default
IProgressMonitor()=default
virtual void phaseFinish(char const *phaseName) noexcept=0
Signal that a phase of the optimizer has finished.
virtual void phaseStart(char const *phaseName, char const *parentPhase, int32_t nbSteps) noexcept=0
Signal that a phase of the optimizer has started.
IBuilder * createInferBuilder(ILogger &logger) noexcept
Create an instance of an IBuilder class.
Definition: NvInfer.h:10653
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:2834
ResizeSelector
The coordinate selector when resize to single pixel output.
Definition: NvInfer.h:4004
@ kFORMULA
Use formula to map the original index.
@ kUPPER
Select the upper left pixel.
EngineCapability
List of supported engine capability flows.
Definition: NvInferRuntime.h:76
nvinfer1::IPluginRegistry * getBuilderPluginRegistry(nvinfer1::EngineCapability capability) noexcept
Return the plugin registry for building a Standard engine, or nullptr if no registry exists.
MemoryPoolType
The type for memory pools used by TensorRT.
Definition: NvInfer.h:9095
ScaleMode
Controls how shift, scale and power are applied in a Scale layer.
Definition: NvInfer.h:1815
@ kUNIFORM
Identical coefficients across all elements of the tensor.
@ kCHANNEL
Per-channel coefficients.
RuntimePlatform
Describes the intended runtime platform (operating system and CPU architecture) for the execution of ...
Definition: NvInfer.h:8716
uint32_t QuantizationFlags
Represents one or more QuantizationFlag values using binary OR operations.
Definition: NvInfer.h:8668
HardwareCompatibilityLevel
Describes requirements of compatibility with GPU architectures other than that of the GPU on which th...
Definition: NvInfer.h:9207
@ kSAME_COMPUTE_CAPABILITY
CumulativeOperation
Enumerates the cumulative operations that may be performed by a Cumulative layer.
Definition: NvInfer.h:6486
BoundingBoxFormat
Representation of bounding box data used for the Boxes input tensor in INMSLayer.
Definition: NvInfer.h:6071
@ 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:8917
constexpr int32_t EnumMax< LayerType >() noexcept
Definition: NvInfer.h:118
constexpr int32_t EnumMax< CalibrationAlgoType >() noexcept
Definition: NvInfer.h:8107
UnaryOperation
Enumerates the unary operations that may be performed by a Unary layer.
Definition: NvInfer.h:2762
@ kISINF
Return true if input value equals +/- infinity for floating-point data type.
@ kCOSH
Hyperbolic cosine.
@ kACOSH
Inverse hyperbolic cosine.
@ kERF
Gauss error function.
@ kISNAN
Return true if input value is a NaN for floating-point data type.
@ kROUND
Round to nearest even for floating-point data type.
@ kATANH
Inverse hyperbolic tangent.
@ kASINH
Inverse hyperbolic sine.
@ kSIGN
Sign, If input > 0, output 1; if input < 0, output -1; if input == 0, output 0.
constexpr int32_t EnumMax< ReduceOperation >() noexcept
Definition: NvInfer.h:2871
constexpr int32_t EnumMax< TripLimit >() noexcept
Definition: NvInfer.h:4405
ActivationType
Enumerates the types of activation to perform in an activation layer.
Definition: NvInfer.h:137
@ kSELU
Selu activation: x>0 ? beta * x : beta * (alpha*exp(x) - alpha)
@ kSCALED_TANH
Scaled tanh activation: alpha*tanh(beta*x)
@ kRELU
Rectified linear activation.
@ kELU
Elu activation: x>=0 ? x : alpha * (exp(x) - 1).
@ kLEAKY_RELU
LeakyRelu activation: x>=0 ? x : alpha * x.
@ kSOFTSIGN
Softsign activation: x / (1+|x|)
@ kHARD_SIGMOID
Hard sigmoid activation: max(0, min(1, alpha*x+beta))
@ kTHRESHOLDED_RELU
Thresholded ReLU activation: x>alpha ? x : 0.
@ kSIGMOID
Sigmoid activation.
@ kCLIP
Clip activation: max(alpha, min(beta, x))
@ kGELU_TANH
GELU tanh activation: 0.5 * x * (1 + tanh(sqrt(2/pi) * (0.044715F * pow(x, 3) + x)))
@ kGELU_ERF
GELU erf activation: 0.5 * x * (1 + erf(sqrt(0.5) * x))
@ kSOFTPLUS
Parametric softplus activation: alpha*log(exp(beta*x)+1)
FillOperation
Enumerates the tensor fill operations that may performed by a fill layer.
Definition: NvInfer.h:4983
@ kRANDOM_UNIFORM
Randomly draw values from a uniform distribution.
@ kRANDOM_NORMAL
Randomly draw values from a normal distribution.
ResizeRoundMode
The rounding mode for nearest neighbor resize.
Definition: NvInfer.h:4034
@ kHALF_DOWN
Round half down.
nvinfer1::safe::IPluginRegistry * getBuilderSafePluginRegistry(nvinfer1::EngineCapability capability) noexcept
Return the plugin registry for building a Safety engine, or nullptr if no registry exists.
PaddingMode
Enumerates the modes of padding to perform in convolution, deconvolution and pooling layer,...
Definition: NvInfer.h:993
@ kSAME_LOWER
Use SAME padding, with prePadding >= postPadding.
@ kEXPLICIT_ROUND_DOWN
Use explicit padding, rounding output size down.
@ kEXPLICIT_ROUND_UP
Use explicit padding, rounding output size up.
@ kSAME_UPPER
Use SAME padding, with prePadding <= postPadding.
TripLimit
Enum that describes kinds of trip limits.
Definition: NvInfer.h:4393
@ kWHILE
Tensor is a scalar of type kBOOL. Loop terminates when value is false.
@ kCOUNT
Tensor is a scalar of type kINT32 or kINT64 that contains the trip count.
uint32_t NetworkDefinitionCreationFlags
Represents one or more NetworkDefinitionCreationFlag flags using binary OR operations....
Definition: NvInfer.h:10305
PreviewFeature
Define preview features.
Definition: NvInfer.h:9170
@ kALIASED_PLUGIN_IO_10_03
TilingOptimizationLevel
Define the optimization levels for Tiling.
Definition: NvInfer.h:9259
@ kFAST
Use a fast algorithm and heuristic based strategy. Slightly increases engine build time.
@ kFULL
Increase search space even wider. Significantly increases engine build time.
constexpr int32_t EnumMax< GatherMode >() noexcept
Definition: NvInfer.h:2521
DataType
The type of weights and tensors.
Definition: NvInferRuntimeBase.h:133
uint32_t BuilderFlags
Represents one or more BuilderFlag values using binary OR operations, e.g., 1U << BuilderFlag::kFP16 ...
Definition: NvInfer.h:8746
DeviceType
The device that this layer/network will execute on.
Definition: NvInferRuntime.h:1304
constexpr int32_t EnumMax< ScaleMode >() noexcept
Definition: NvInfer.h:1827
CalibrationAlgoType
Version of calibration algorithm to use.
Definition: NvInfer.h:8094
@ kENTROPY_CALIBRATION_2
Entropy calibration.
@ kLEGACY_CALIBRATION
Legacy calibration.
@ kENTROPY_CALIBRATION
Legacy entropy calibration.
@ kMINMAX_CALIBRATION
Minmax calibration.
LayerType
The type values of layer classes.
Definition: NvInfer.h:58
@ kGRID_SAMPLE
Grid sample layer.
@ kRAGGED_SOFTMAX
Ragged softmax layer.
@ kDECONVOLUTION
Deconvolution layer.
@ kASSERTION
Assertion layer.
@ kMATRIX_MULTIPLY
Matrix multiply layer.
@ kCONDITION
Condition layer.
@ kCUMULATIVE
Cumulative layer.
@ kCONDITIONAL_INPUT
Conditional Input layer.
@ kIDENTITY
Identity layer.
@ kNORMALIZATION
Normalization layer.
@ kQUANTIZE
Quantize layer.
@ kCONVOLUTION
Convolution layer.
@ kPARAMETRIC_RELU
Parametric ReLU layer.
@ kUNSQUEEZE
Unsqueeze Layer.
@ kCONCATENATION
Concatenation layer.
@ kREVERSE_SEQUENCE
Reverse sequence layer.
@ kRECURRENCE
Loop Recurrence layer.
@ kDEQUANTIZE
Dequantize layer.
@ kPLUGIN_V3
PluginV3 layer.
@ kITERATOR
Loop Iterator layer.
@ kTRIP_LIMIT
Loop Trip limit layer.
@ kDYNAMIC_QUANTIZE
Dynamic Quantize layer.
@ kUNARY
UnaryOp operation Layer.
@ kACTIVATION
Activation layer.
@ kELEMENTWISE
Elementwise layer.
@ kPLUGIN_V2
PluginV2 layer.
@ kLOOP_OUTPUT
Loop output layer.
@ kCONDITIONAL_OUTPUT
Conditional Output layer.
@ kCONSTANT
Constant layer.
@ kNON_ZERO
NonZero layer.
constexpr int32_t EnumMax< QuantizationFlag >() noexcept
Definition: NvInfer.h:8693
SampleMode
Controls how ISliceLayer and IGridSample handle out-of-bounds coordinates.
Definition: NvInfer.h:3220
@ 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:2509
@ kDEFAULT
Similar to ONNX Gather.
@ kELEMENT
Similar to ONNX GatherElements.
@ kND
Similar to ONNX GatherND.
uint32_t TensorFormats
It is capable of representing one or more TensorFormat by binary OR operations, e....
Definition: NvInfer.h:129
ProfilingVerbosity
List of verbosity levels of layer information exposed in NVTX annotations and in IEngineInspector.
Definition: NvInferRuntime.h:2846
NetworkDefinitionCreationFlag
List of immutable network properties expressed at network creation time. NetworkDefinitionCreationFla...
Definition: NvInfer.h:10316
@ kPREFER_JIT_PYTHON_PLUGINS
@ kPREFER_AOT_PYTHON_PLUGINS
ElementWiseOperation
Enumerates the binary operations that may be performed by an ElementWise layer.
Definition: NvInfer.h:2419
@ 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:8680
@ kCALIBRATE_BEFORE_FUSION
constexpr int32_t EnumMax< SampleMode >() noexcept
Definition: NvInfer.h:3236
InterpolationMode
Enumerates various modes of interpolation.
Definition: NvInfer.h:3922
@ kNEAREST
ND (0 < N <= 8) nearest neighbor resizing.
@ kCUBIC
Supports bicubic (2D) interpolation.
@ kLINEAR
Supports linear (1D), bilinear (2D), and trilinear (3D) interpolation.
BuilderFlag
List of valid modes that the builder can enable when creating an engine from a network definition.
Definition: NvInfer.h:8756
@ kWEIGHT_STREAMING
Enable weight streaming for the current engine.
@ kDEBUG
Enable debugging of layers via synchronizing after every layer.
@ kGPU_FALLBACK
Enable layers marked to execute on GPU if layer cannot execute on DLA.
@ kFP4
Enable plugins with FP4 input/output.
@ 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.
@ kERROR_ON_TIMING_CACHE_MISS
@ kINT8
Enable Int8 layer selection, with FP32 fallback with FP16 fallback if kFP16 also specified.
@ kEDITABLE_TIMING_CACHE
Enable editable timing cache.
@ kPREFER_PRECISION_CONSTRAINTS
@ kINT4
Enable plugins with INT4 input/output.
@ kSTRIP_PLAN
Strip the refittable weights from the engine plan file.
@ kMONITOR_MEMORY
Enable memory monitor during build time.
@ kDISABLE_TIMING_CACHE
Disable reuse of timing information across identical layers.
@ kDISABLE_COMPILATION_CACHE
@ kREFIT
Enable building a refittable engine.
@ kOBEY_PRECISION_CONSTRAINTS
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:3512
constexpr int32_t EnumMax< MemoryPoolType >() noexcept
Definition: NvInfer.h:9156
TopKOperation
Enumerates the operations that may be performed by a TopK layer.
Definition: NvInfer.h:3501
ReduceOperation
Enumerates the reduce operations that may be performed by a Reduce layer.
Definition: NvInfer.h:2857
constexpr int32_t EnumMax< LoopOutput >() noexcept
Definition: NvInfer.h:4382
constexpr int32_t EnumMax< NetworkDefinitionCreationFlag >() noexcept
Definition: NvInfer.h:10344
ScatterMode
Control form of IScatterLayer.
Definition: NvInfer.h:5806
MatrixOperation
Enumerates the operations that may be performed on a tensor by IMatrixMultiplyLayer before multiplica...
Definition: NvInfer.h:3627
@ kTRANSPOSE
Like kNONE, but transpose the matrix dimensions.
ResizeCoordinateTransformation
The resize coordinate transformation function.
Definition: NvInfer.h:3950
constexpr int32_t EnumMax< UnaryOperation >() noexcept
Definition: NvInfer.h:2796
LoopOutput
Enum that describes kinds of loop outputs.
Definition: NvInfer.h:4365
@ 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:6084
constexpr int32_t EnumMax< MatrixOperation >() noexcept
Definition: NvInfer.h:3655
PoolingType
The type of pooling to perform in a pooling layer.
Definition: NvInfer.h:1429
@ kAVERAGE
Average over elements. If the tensor is padded, the count includes the padding.
@ kMAX
Maximum over elements.
@ kMAX_AVERAGE_BLEND
Blending between max and average pooling: (1-blendFactor)*maxPool + blendFactor*avgPool.
v_1_0::IProgressMonitor IProgressMonitor
Definition: NvInfer.h:9375
constexpr int32_t EnumMax< FillOperation >() noexcept
Definition: NvInfer.h:5014
TensorLocation
The location for tensor data storage, device or host.
Definition: NvInferRuntime.h:204
OptProfileSelector
When setting or querying optimization profile parameters (such as shape tensor inputs or dynamic dime...
Definition: NvInferRuntime.h:2578
constexpr int32_t EnumMax< ScatterMode >() noexcept
Definition: NvInfer.h:5817
Represents a permutation of dimensions.
Definition: NvInfer.h:3029
Declaration of EnumMaxImpl struct to store maximum number of elements in an enumeration type.
Definition: NvInferRuntimeBase.h:118
The key to retrieve timing cache entries.
Definition: NvInfer.h:8937
Definition: NvInfer.h:8949
uint64_t tacticHash
Hash of the selected tactic.
Definition: NvInfer.h:8951
float timingMSec
Timing of this tactic in milliseconds. Negative numbers and NaN are invalid values.
Definition: NvInfer.h:8953