164 static constexpr int32_t kVALUE = 14;
202 mImpl->setName(name);
214 return mImpl->getName();
233 mImpl->setDimensions(dimensions);
247 return mImpl->getDimensions();
281 mImpl->setType(type);
296 return mImpl->getType();
313 return mImpl->setDynamicRange(min, max);
321 return mImpl->isNetworkInput();
329 return mImpl->isNetworkOutput();
346 mImpl->setBroadcastAcrossBatch(broadcastAcrossBatch);
360 return mImpl->getBroadcastAcrossBatch();
372 return mImpl->getLocation();
391 mImpl->setLocation(location);
403 return mImpl->dynamicRangeIsSet();
411 mImpl->resetDynamicRange();
421 return mImpl->getDynamicRangeMin();
431 return mImpl->getDynamicRangeMax();
453 mImpl->setAllowedFormats(formats);
466 return mImpl->getAllowedFormats();
497 return mImpl->isShapeTensor();
518 return mImpl->isExecutionTensor();
544 mImpl->setDimensionName(index, name);
559 return mImpl->getDimensionName(index);
584 return mLayer->getType();
598 mLayer->setName(name);
608 return mLayer->getName();
616 return mLayer->getNbInputs();
629 return mLayer->getInput(index);
637 return mLayer->getNbOutputs();
647 return mLayer->getOutput(index);
664 return mLayer->setInput(index, tensor);
695 mLayer->setPrecision(dataType);
707 return mLayer->getPrecision();
719 return mLayer->precisionIsSet();
729 mLayer->resetPrecision();
776 mLayer->setOutputType(index, dataType);
791 return mLayer->getOutputType(index);
805 return mLayer->outputTypeIsSet(index);
817 return mLayer->resetOutputType(index);
835 mLayer->setMetadata(metadata);
848 return mLayer->getMetadata();
853 apiv::VLayer* mLayer;
1030 static constexpr int32_t kVALUE = 4;
1058 mImpl->setNbOutputMaps(nbOutputMaps);
1068 return mImpl->getNbOutputMaps();
1088 mImpl->setNbGroups(nbGroups);
1098 return mImpl->getNbGroups();
1112 mImpl->setKernelWeights(weights);
1122 return mImpl->getKernelWeights();
1137 mImpl->setBiasWeights(weights);
1147 return mImpl->getBiasWeights();
1164 mImpl->setPrePadding(padding);
1174 return mImpl->getPrePadding();
1191 mImpl->setPostPadding(padding);
1201 return mImpl->getPostPadding();
1215 mImpl->setPaddingMode(paddingMode);
1227 return mImpl->getPaddingMode();
1240 mImpl->setKernelSizeNd(kernelSize);
1250 return mImpl->getKernelSizeNd();
1265 mImpl->setStrideNd(stride);
1275 return mImpl->getStrideNd();
1293 mImpl->setPaddingNd(padding);
1305 return mImpl->getPaddingNd();
1319 mImpl->setDilationNd(dilation);
1329 return mImpl->getDilationNd();
1378 mImpl->setActivationType(type);
1388 return mImpl->getActivationType();
1403 mImpl->setAlpha(alpha);
1417 mImpl->setBeta(beta);
1426 return mImpl->getAlpha();
1435 return mImpl->getBeta();
1465 static constexpr int32_t kVALUE = 3;
1492 mImpl->setPoolingType(type);
1502 return mImpl->getPoolingType();
1517 mImpl->setBlendFactor(blendFactor);
1530 return mImpl->getBlendFactor();
1544 mImpl->setAverageCountExcludesPadding(exclusive);
1555 return mImpl->getAverageCountExcludesPadding();
1573 mImpl->setPrePadding(padding);
1583 return mImpl->getPrePadding();
1601 mImpl->setPostPadding(padding);
1611 return mImpl->getPostPadding();
1624 mImpl->setPaddingMode(paddingMode);
1635 return mImpl->getPaddingMode();
1648 mImpl->setWindowSizeNd(windowSize);
1658 return mImpl->getWindowSizeNd();
1673 mImpl->setStrideNd(stride);
1683 return mImpl->getStrideNd();
1702 mImpl->setPaddingNd(padding);
1714 return mImpl->getPaddingNd();
1745 mImpl->setWindowSize(windowSize);
1755 return mImpl->getWindowSize();
1767 mImpl->setAlpha(alpha);
1777 return mImpl->getAlpha();
1789 mImpl->setBeta(beta);
1799 return mImpl->getBeta();
1821 return mImpl->getK();
1887 mImpl->setMode(mode);
1897 return mImpl->getMode();
1907 mImpl->setShift(shift);
1917 return mImpl->getShift();
1927 mImpl->setScale(scale);
1937 return mImpl->getScale();
1947 mImpl->setPower(power);
1957 return mImpl->getPower();
1972 return mImpl->getChannelAxis();
1993 mImpl->setChannelAxis(channelAxis);
2046 mImpl->setAxes(axes);
2056 return mImpl->getAxes();
2092 mImpl->setAxis(axis);
2102 return mImpl->getAxis();
2129 mImpl->setNbOutputMaps(nbOutputMaps);
2139 return mImpl->getNbOutputMaps();
2159 mImpl->setNbGroups(nbGroups);
2169 return mImpl->getNbGroups();
2183 mImpl->setKernelWeights(weights);
2193 return mImpl->getKernelWeights();
2208 mImpl->setBiasWeights(weights);
2218 return mImpl->getBiasWeights();
2235 mImpl->setPrePadding(padding);
2245 return mImpl->getPrePadding();
2262 mImpl->setPostPadding(padding);
2272 return mImpl->getPostPadding();
2286 mImpl->setPaddingMode(paddingMode);
2298 return mImpl->getPaddingMode();
2313 mImpl->setKernelSizeNd(kernelSize);
2323 return mImpl->getKernelSizeNd();
2340 mImpl->setStrideNd(stride);
2350 return mImpl->getStrideNd();
2368 mImpl->setPaddingNd(padding);
2380 return mImpl->getPaddingNd();
2406 mImpl->setDilationNd(dilation);
2416 return mImpl->getDilationNd();
2466 static constexpr int32_t kVALUE = 14;
2503 return mImpl->setOperation(op);
2515 return mImpl->getOperation();
2636 mImpl->setGatherAxis(axis);
2648 return mImpl->getGatherAxis();
2671 mImpl->setNbElementWiseDims(elementWiseDims);
2681 return mImpl->getNbElementWiseDims();
2691 mImpl->setMode(mode);
2701 return mImpl->getMode();
2730 return mImpl->getPlugin();
2757 return mImpl->getPlugin();
2840 mImpl->setOperation(op);
2850 return mImpl->getOperation();
2913 mImpl->setOperation(op);
2923 return mImpl->getOperation();
2933 mImpl->setReduceAxes(reduceAxes);
2943 return mImpl->getReduceAxes();
2953 mImpl->setKeepDimensions(keepDimensions);
2963 return mImpl->getKeepDimensions();
2997 mImpl->setPrePaddingNd(padding);
3009 return mImpl->getPrePaddingNd();
3023 mImpl->setPostPaddingNd(padding);
3035 return mImpl->getPostPaddingNd();
3085 mImpl->setFirstTranspose(permutation);
3097 return mImpl->getFirstTranspose();
3125 mImpl->setReshapeDimensions(dimensions);
3138 return mImpl->getReshapeDimensions();
3185 mImpl->setSecondTranspose(permutation);
3197 return mImpl->getSecondTranspose();
3213 return mImpl->setZeroIsPlaceholder(zeroIsPlaceholder);
3226 return mImpl->getZeroIsPlaceholder();
3337 mImpl->setStart(start);
3352 return mImpl->getStart();
3366 return mImpl->setSize(size);
3381 return mImpl->getSize();
3395 mImpl->setStride(stride);
3410 return mImpl->getStride();
3420 mImpl->setMode(mode);
3430 return mImpl->getMode();
3473 mImpl->setAxes(axes);
3488 return mImpl->getAxes();
3558 mImpl->setOperation(op);
3568 return mImpl->getOperation();
3596 return mImpl->getK();
3606 mImpl->setReduceAxes(reduceAxes);
3616 return mImpl->getReduceAxes();
3718 mImpl->setOperation(index, op);
3730 return mImpl->getOperation(index);
3855 mImpl->setToType(toType);
3866 return mImpl->getToType();
3895 mImpl->setWeights(weights);
3905 return mImpl->getWeights();
3917 mImpl->setDimensions(dimensions);
3929 return mImpl->getDimensions();
3975 static constexpr int32_t kVALUE = 3;
4029 static constexpr int32_t kVALUE = 3;
4059 static constexpr int32_t kVALUE = 2;
4095 static constexpr int32_t kVALUE = 4;
4158 return mImpl->setOutputDimensions(dimensions);
4168 return mImpl->getOutputDimensions();
4196 void setScales(
float const* scales, int32_t nbScales)
noexcept
4198 mImpl->setScales(scales, nbScales);
4215 int32_t
getScales(int32_t size,
float* scales)
const noexcept
4217 return mImpl->getScales(size, scales);
4229 mImpl->setResizeMode(interpolationMode);
4239 return mImpl->getResizeMode();
4274 mImpl->setCoordinateTransformation(coordTransform);
4284 return mImpl->getCoordinateTransformation();
4299 mImpl->setSelectorForSinglePixel(selector);
4309 return mImpl->getSelectorForSinglePixel();
4323 mImpl->setNearestRounding(value);
4333 return mImpl->getNearestRounding();
4355 mImpl->setCubicCoeff(A);
4365 return mImpl->getCubicCoeff();
4378 mImpl->setExcludeOutside(excludeFlag);
4388 return mImpl->getExcludeOutside();
4470 return mBoundary->getLoop();
4475 apiv::VLoopBoundaryLayer* mBoundary;
4493 return mBoundary->getConditional();
4498 apiv::VConditionalBoundaryLayer* mBoundary;
4582 return mImpl->setCondition(condition);
4600 return mImpl->addOutput(trueSubgraphOutput, falseSubgraphOutput);
4612 return mImpl->addInput(input);
4627 mImpl->setName(name);
4637 return mImpl->getName();
4707 return mImpl->getLoopOutput();
4724 mImpl->setAxis(axis);
4732 return mImpl->getAxis();
4781 return mImpl->getTripLimit();
4807 mImpl->setAxis(axis);
4815 return mImpl->getAxis();
4829 mImpl->setReverse(reverse);
4839 return mImpl->getReverse();
4868 return mImpl->addRecurrence(initialValue);
4889 return mImpl->addTripLimit(tensor, limit);
4902 return mImpl->addIterator(tensor, axis, reverse);
4915 return mImpl->addLoopOutput(tensor, outputKind, axis);
4930 mImpl->setName(name);
4940 return mImpl->getName();
4995 mImpl->setMessage(message);
5005 return mImpl->getMessage();
5107 mImpl->setDimensions(dimensions);
5122 return mImpl->getDimensions();
5132 mImpl->setOperation(op);
5142 return mImpl->getOperation();
5161 mImpl->setAlpha(alpha);
5176 return mImpl->getAlpha();
5195 mImpl->setBeta(beta);
5210 return mImpl->getBeta();
5271 mImpl->setAlphaInt64(alpha);
5286 return mImpl->getAlphaInt64();
5305 mImpl->setBetaInt64(beta);
5320 return mImpl->getBetaInt64();
5328 return mImpl->isAlphaBetaInt64();
5345 mImpl->setToType(toType);
5357 return mImpl->getToType();
5452 return mImpl->getAxis();
5463 mImpl->setAxis(axis);
5479 mImpl->setToType(toType);
5491 return mImpl->getToType();
5583 return mImpl->getAxis();
5594 mImpl->setAxis(axis);
5610 mImpl->setToType(toType);
5622 return mImpl->getToType();
5677 mImpl->setToType(toType);
5690 return mImpl->getToType();
5702 mImpl->setScaleType(scaleType);
5715 return mImpl->getScaleType();
5728 mImpl->setAxis(axis);
5738 return mImpl->getAxis();
5751 mImpl->setBlockSize(size);
5761 return mImpl->getBlockSize();
5819 return mImpl->setEquation(equation);
5829 return mImpl->getEquation();
5928 mImpl->setMode(mode);
5938 return mImpl->getMode();
5948 mImpl->setAxis(axis);
5956 return mImpl->getAxis();
6000 mImpl->setAxis(axis);
6008 return mImpl->getAxis();
6037 mImpl->setInterpolationMode(mode);
6049 return mImpl->getInterpolationMode();
6059 mImpl->setAlignCorners(alignCorners);
6071 return mImpl->getAlignCorners();
6083 return mImpl->setSampleMode(mode);
6095 return mImpl->getSampleMode();
6189 mImpl->setBoundingBoxFormat(fmt);
6201 return mImpl->getBoundingBoxFormat();
6215 mImpl->setTopKBoxLimit(limit);
6225 return mImpl->getTopKBoxLimit();
6278 mImpl->setBatchAxis(batchAxis);
6288 return mImpl->getBatchAxis();
6301 mImpl->setSequenceAxis(sequenceAxis);
6311 return mImpl->getSequenceAxis();
6349 return mImpl->setEpsilon(eps);
6359 return mImpl->getEpsilon();
6369 return mImpl->setAxes(axesMask);
6379 return mImpl->getAxes();
6400 return mImpl->setNbGroups(nbGroups);
6410 return mImpl->getNbGroups();
6436 return mImpl->setComputePrecision(type);
6446 return mImpl->getComputePrecision();
6541 static constexpr int32_t kVALUE = 1;
6587 return mImpl->setOperation(op);
6599 return mImpl->getOperation();
6611 mImpl->setExclusive(exclusive);
6623 return mImpl->getExclusive();
6635 mImpl->setReverse(reverse);
6647 return mImpl->getReverse();
6714 return mImpl->addInput(name, type, dimensions);
6728 mImpl->markOutput(tensor);
6746 return mImpl->markDebug(tensor);
6762 return mImpl->unmarkDebug(tensor);
6772 return mImpl->isDebugTensor(tensor);
6792 return mImpl->addActivation(input, type);
6811 return mImpl->addLRN(input, window, alpha, beta, k);
6837 return mImpl->addScale(input, mode, shift, scale, power);
6850 return mImpl->addSoftMax(input);
6867 return mImpl->addConcatenation(inputs, nbInputs);
6894 return mImpl->addElementWise(input1, input2, op);
6916 return mImpl->addUnary(input, operation);
6930 return mImpl->addShuffle(input);
6947 return mImpl->addOneHot(indices, values, depth, axis);
6959 return mImpl->getNbLayers();
6973 return mImpl->getLayer(index);
6985 return mImpl->getNbInputs();
7001 return mImpl->getInput(index);
7015 return mImpl->getNbOutputs();
7031 return mImpl->getOutput(index);
7058 return mImpl->addReduce(input, operation, reduceAxes, keepDimensions);
7090 return mImpl->addTopK(input, op, k, reduceAxes);
7106 return mImpl->addGather(data, indices, axis);
7122 return mImpl->addGatherV2(data, indices, mode);
7141 return mImpl->addRaggedSoftMax(input, bounds);
7163 return mImpl->addMatrixMultiply(input0, op0, input1, op1);
7177 return mImpl->addNonZero(input);
7201 return mImpl->addConstant(dimensions, weights);
7215 return mImpl->addIdentity(input);
7230 return mImpl->addCast(input, toType);
7245 mImpl->removeTensor(tensor);
7257 mImpl->unmarkOutput(tensor);
7278 return mImpl->addPluginV2(inputs, nbInputs, plugin);
7295 int32_t nbShapeInputs,
IPluginV3& plugin)
noexcept
7297 return mImpl->addPluginV3(inputs, nbInputs, shapeInputs, nbShapeInputs, plugin);
7316 return mImpl->addSlice(input, start, size, stride);
7340 mImpl->setName(name);
7354 return mImpl->getName();
7370 return mImpl->addShape(input);
7384 return mImpl->hasImplicitBatchDimension();
7394 return mImpl->getFlags();
7406 return mImpl->getFlag(networkDefinitionCreationFlag);
7423 return mImpl->markOutputForShapes(tensor);
7435 return mImpl->unmarkOutputForShapes(tensor);
7453 return mImpl->addParametricReLU(input, slope);
7476 return mImpl->addConvolutionNd(input, nbOutputMaps, kernelSize, kernelWeights, biasWeights);
7495 return mImpl->addPoolingNd(input, type, windowSize);
7518 return mImpl->addDeconvolutionNd(input, nbOutputMaps, kernelSize, kernelWeights, biasWeights);
7555 return mImpl->addScaleNd(input, mode, shift, scale, power, channelAxis);
7571 return mImpl->addResize(input);
7585 return mImpl->addLoop();
7600 return mImpl->addIfConditional();
7639 return mImpl->addSelect(condition, thenInput, elseInput);
7656 return mImpl->addAssertion(condition, message);
7681 return mImpl->addFill(dimensions, op);
7707 return mImpl->addFillV2(dimensions, op, outputType);
7723 return mImpl->addPaddingNd(input, prePadding, postPadding);
7747 return mImpl->setWeightsName(weights, name);
7766 mImpl->setErrorRecorder(recorder);
7781 return mImpl->getErrorRecorder();
7802 return mImpl->addDequantize(input, scale);
7824 return mImpl->addDequantizeV2(input, scale, outputType);
7844 return mImpl->addScatter(data, indices, updates, mode);
7865 return mImpl->addQuantize(input, scale);
7887 return mImpl->addQuantizeV2(input, scale, outputType);
7913 return mImpl->addDynamicQuantize(input, axis, blockSize, outputType, scaleType);
7928 return mImpl->addEinsum(inputs, nbInputs, equation);
7946 return mImpl->addGridSample(input, grid);
7964 return mImpl->addNMS(boxes, scores, maxOutputBoxesPerClass);
7981 return mImpl->addReverseSequence(input, sequenceLens);
8007 return mImpl->addNormalization(input, scale, bias, axesMask);
8029 return mImpl->addCumulative(input, axis, operation, exclusive, reverse);
8040 return mImpl->getBuilder();
8053 return mImpl->markWeightsRefittable(name);
8065 return mImpl->unmarkWeightsRefittable(name);
8078 return mImpl->areWeightsMarkedRefittable(name);
8097 return mImpl->addSqueeze(input, axes);
8118 return mImpl->addUnsqueeze(input, axes);
8190 virtual
bool getBatch(
void* bindings[],
char const* names[], int32_t nbBindings) noexcept = 0;
8206 virtual
void const* readCalibrationCache(std::
size_t& length) noexcept = 0;
8216 virtual
void writeCalibrationCache(
void const* ptr, std::
size_t length) noexcept = 0;
8382 virtual
double getRegressionCutoff() const noexcept = 0;
8396 virtual
void const* readHistogramCache(std::
size_t& length) noexcept = 0;
8406 virtual
void writeHistogramCache(
void const* ptr, std::
size_t length) noexcept = 0;
8449 return mImpl->getDataType();
8460 return mImpl->getStrides();
8470 return mImpl->getVectorizedDim();
8481 return mImpl->getComponentsPerElement();
8510 return mImpl->getImplementation();
8518 return mImpl->getTactic();
8546 return mImpl->getName();
8558 return mImpl->getDimensions(index, select);
8566 return mImpl->getNbInputs();
8574 return mImpl->getNbOutputs();
8603 return mImpl->getAlgorithmVariant();
8611 return mImpl->getTimingMSec();
8619 return mImpl->getWorkspaceSize();
8633 return mImpl->getAlgorithmIOInfoByIndex(index);
8668 int32_t nbChoices, int32_t* selection)
noexcept = 0;
8681 int32_t nbAlgorithms)
noexcept = 0;
8775 static constexpr int32_t kVALUE = 2;
8995 static constexpr uint64_t kINVALID_TACTIC_HASH = UINT64_MAX;
9028 return mImpl->serialize();
9052 return mImpl->combine(inputCache, ignoreMismatch);
9062 return mImpl->reset();
9079 int64_t
queryKeys(TimingCacheKey* keyBuffer, int64_t capacity)
const noexcept
9081 return mImpl->queryKeys(keyBuffer, capacity);
9096 TimingCacheValue
query(TimingCacheKey
const& key)
const noexcept
9098 return mImpl->query(key);
9118 bool update(TimingCacheKey
const& key, TimingCacheValue
const& value)
noexcept
9120 return mImpl->update(key, value);
9241 static constexpr int32_t kVALUE = 3;
9293 static constexpr int32_t kVALUE = 3;
9332 static constexpr int32_t kVALUE = 4;
9371 virtual void phaseStart(
char const* phaseName,
char const* parentPhase, int32_t nbSteps)
noexcept = 0;
9385 virtual bool stepComplete(
char const* phaseName, int32_t step)
noexcept = 0;
9444 virtual
void setAvgTimingIterations(int32_t avgTiming) noexcept
9446 mImpl->setAvgTimingIterations(avgTiming);
9458 return mImpl->getAvgTimingIterations();
9471 mImpl->setEngineCapability(capability);
9483 return mImpl->getEngineCapability();
9495 mImpl->setInt8Calibrator(calibrator);
9505 return mImpl->getInt8Calibrator();
9522 mImpl->setFlags(builderFlags);
9534 return mImpl->getFlags();
9546 mImpl->clearFlag(builderFlag);
9558 mImpl->setFlag(builderFlag);
9570 return mImpl->getFlag(builderFlag);
9587 mImpl->setDeviceType(layer, deviceType);
9597 return mImpl->getDeviceType(layer);
9609 return mImpl->isDeviceTypeSet(layer);
9619 mImpl->resetDeviceType(layer);
9629 return mImpl->canRunOnDLA(layer);
9645 mImpl->setDLACore(dlaCore);
9655 return mImpl->getDLACore();
9666 mImpl->setDefaultDeviceType(deviceType);
9676 return mImpl->getDefaultDeviceType();
9698 return mImpl->setProfileStream(stream);
9710 return mImpl->getProfileStream();
9727 return mImpl->addOptimizationProfile(profile);
9740 return mImpl->getNbOptimizationProfiles();
9752 mImpl->setProfilingVerbosity(verbosity);
9765 return mImpl->getProfilingVerbosity();
9777 mImpl->setAlgorithmSelector(selector);
9787 return mImpl->getAlgorithmSelector();
9805 return mImpl->setCalibrationProfile(profile);
9817 return mImpl->getCalibrationProfile();
9836 mImpl->setQuantizationFlags(flags);
9850 return mImpl->getQuantizationFlags();
9864 mImpl->clearQuantizationFlag(flag);
9878 mImpl->setQuantizationFlag(flag);
9892 return mImpl->getQuantizationFlag(flag);
9914 return mImpl->setTacticSources(tacticSources);
9929 return mImpl->getTacticSources();
9949 return mImpl->createTimingCache(blob, size);
9972 return mImpl->setTimingCache(cache, ignoreMismatch);
9982 return mImpl->getTimingCache();
10014 mImpl->setMemoryPoolLimit(pool, poolSize);
10033 return mImpl->getMemoryPoolLimit(pool);
10051 mImpl->setPreviewFeature(feature, enable);
10065 return mImpl->getPreviewFeature(feature);
10098 mImpl->setBuilderOptimizationLevel(level);
10110 return mImpl->getBuilderOptimizationLevel();
10127 mImpl->setHardwareCompatibilityLevel(hardwareCompatibilityLevel);
10140 return mImpl->getHardwareCompatibilityLevel();
10153 mImpl->setPluginsToSerialize(paths, nbPaths);
10166 return mImpl->getPluginToSerialize(index);
10176 return mImpl->getNbPluginsToSerialize();
10205 mImpl->setMaxAuxStreams(nbStreams);
10215 return mImpl->getMaxAuxStreams();
10231 return mImpl->setProgressMonitor(monitor);
10241 return mImpl->getProgressMonitor();
10257 mImpl->setRuntimePlatform(runtimePlatform);
10269 return mImpl->getRuntimePlatform();
10281 mImpl->setMaxNbTactics(maxNbTactics);
10293 return mImpl->getMaxNbTactics();
10309 return mImpl->setTilingOptimizationLevel(level);
10321 return mImpl->getTilingOptimizationLevel();
10337 return mImpl->setL2LimitForTiling(size);
10349 return mImpl->getL2LimitForTiling();
10426 return mImpl->platformHasFastFp16();
10436 return mImpl->platformHasFastInt8();
10448 return mImpl->getMaxDLABatchSize();
10456 return mImpl->getNbDLACores();
10474 mImpl->setGpuAllocator(allocator);
10488 return mImpl->createBuilderConfig();
10514 return mImpl->createNetworkV2(flags);
10529 return mImpl->createOptimizationProfile();
10548 mImpl->setErrorRecorder(recorder);
10563 return mImpl->getErrorRecorder();
10581 return mImpl->platformHasTf32();
10600 return mImpl->buildSerializedNetwork(network, config);
10620 return mImpl->buildEngineWithConfig(network, config);
10642 return mImpl->isNetworkSupported(network, config);
10652 return mImpl->getLogger();
10668 return mImpl->setMaxThreads(maxThreads);
10682 return mImpl->getMaxThreads();
10692 return mImpl->getPluginRegistry();
10705extern "C" TENSORRTAPI void* createInferBuilder_INTERNAL(
void* logger, int32_t version)
noexcept;
#define TENSORRTAPI
Definition: NvInferRuntimeBase.h:69
#define NV_TENSORRT_VERSION
Definition: NvInferRuntimeBase.h:101
#define TRT_DEPRECATED
Definition: NvInferRuntimeBase.h:42
#define TRT_DEPRECATED_ENUM
Definition: NvInferRuntimeBase.h:43
Definition: NvInferRuntimeBase.h:214
static constexpr int32_t MAX_DIMS
The maximum rank (number of dimensions) supported for a tensor.
Definition: NvInferRuntimeBase.h:217
An Activation layer in a network definition.
Definition: NvInfer.h:1367
void setBeta(float beta) noexcept
Set the beta parameter (must be finite).
Definition: NvInfer.h:1415
void setActivationType(ActivationType type) noexcept
Set the type of activation to be performed.
Definition: NvInfer.h:1376
ActivationType getActivationType() const noexcept
Get the type of activation to be performed.
Definition: NvInfer.h:1386
float getAlpha() const noexcept
Get the alpha parameter.
Definition: NvInfer.h:1424
virtual ~IActivationLayer() noexcept=default
float getBeta() const noexcept
Get the beta parameter.
Definition: NvInfer.h:1433
void setAlpha(float alpha) noexcept
Set the alpha parameter (must be finite).
Definition: NvInfer.h:1401
Describes the context and requirements, that could be fulfilled by one or more instances of IAlgorith...
Definition: NvInfer.h:8537
int32_t getNbOutputs() const noexcept
Return number of outputs of the algorithm.
Definition: NvInfer.h:8572
int32_t getNbInputs() const noexcept
Return number of inputs of the algorithm.
Definition: NvInfer.h:8564
char const * getName() const noexcept
Return name of the algorithm node.
Definition: NvInfer.h:8544
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:8556
Describes a variation of execution of a layer. An algorithm is represented by IAlgorithmVariant and t...
Definition: NvInfer.h:8596
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:8617
float getTimingMSec() const noexcept
The time in milliseconds to execute the algorithm.
Definition: NvInfer.h:8609
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:8631
virtual ~IAlgorithm() noexcept=default
IAlgorithmVariant const & getAlgorithmVariant() const noexcept
Returns the algorithm variant.
Definition: NvInfer.h:8601
Carries information about input or output of the algorithm. IAlgorithmIOInfo for all the input and ou...
Definition: NvInfer.h:8440
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:8468
Dims getStrides() const noexcept
Return strides of the input/output tensor of algorithm. For vectorized formats, strides are given in ...
Definition: NvInfer.h:8458
DataType getDataType() const noexcept
Return DataType of the input/output of algorithm.
Definition: NvInfer.h:8447
int64_t getComponentsPerElement() const noexcept
Return the number of components per element. This is always 1 for non-vectorized formats.
Definition: NvInfer.h:8479
provides a unique 128-bit identifier, which along with the input and output information denotes the v...
Definition: NvInfer.h:8503
virtual ~IAlgorithmVariant() noexcept=default
int64_t getTactic() const noexcept
Return tactic of the algorithm.
Definition: NvInfer.h:8516
int64_t getImplementation() const noexcept
Return implementation of the algorithm.
Definition: NvInfer.h:8508
An assertion layer in a network.
Definition: NvInfer.h:4983
void setMessage(char const *message) noexcept
Set the message to print if the assertion fails.
Definition: NvInfer.h:4993
char const * getMessage() const noexcept
Return the assertion message.
Definition: NvInfer.h:5003
virtual ~IAssertionLayer() noexcept=default
Holds properties for configuring a builder to produce an engine.
Definition: NvInfer.h:9432
void setMemoryPoolLimit(MemoryPoolType pool, std::size_t poolSize) noexcept
Set the memory size for the memory pool.
Definition: NvInfer.h:10012
TRT_DEPRECATED void setQuantizationFlags(QuantizationFlags flags) noexcept
Set the quantization flags.
Definition: NvInfer.h:9834
nvinfer1::ITimingCache * createTimingCache(void const *blob, std::size_t size) const noexcept
Create timing cache.
Definition: NvInfer.h:9947
void setPreviewFeature(PreviewFeature feature, bool enable) noexcept
Enable or disable a specific preview feature.
Definition: NvInfer.h:10049
TRT_DEPRECATED void setAlgorithmSelector(IAlgorithmSelector *selector) noexcept
Set Algorithm Selector.
Definition: NvInfer.h:9775
TRT_DEPRECATED void setInt8Calibrator(IInt8Calibrator *calibrator) noexcept
Set Int8 Calibration interface.
Definition: NvInfer.h:9493
bool getPreviewFeature(PreviewFeature feature) const noexcept
Get status of preview feature.
Definition: NvInfer.h:10063
int32_t getBuilderOptimizationLevel() noexcept
Get builder optimization level.
Definition: NvInfer.h:10108
bool setTacticSources(TacticSources tacticSources) noexcept
Set tactic sources.
Definition: NvInfer.h:9912
void setPluginsToSerialize(char const *const *paths, int32_t nbPaths) noexcept
Set the plugin libraries to be serialized with version-compatible engines.
Definition: NvInfer.h:10151
bool setTilingOptimizationLevel(TilingOptimizationLevel level) noexcept
Set the Tiling optimization level.
Definition: NvInfer.h:10307
bool setL2LimitForTiling(int64_t size) noexcept
Set the L2 cache usage limit for Tiling optimization.
Definition: NvInfer.h:10335
TRT_DEPRECATED IInt8Calibrator * getInt8Calibrator() const noexcept
Get Int8 Calibration interface.
Definition: NvInfer.h:9503
std::size_t getMemoryPoolLimit(MemoryPoolType pool) const noexcept
Get the memory size limit of the memory pool.
Definition: NvInfer.h:10031
int32_t getDLACore() const noexcept
Get the DLA core that the engine executes on.
Definition: NvInfer.h:9653
int32_t getNbPluginsToSerialize() const noexcept
Get the number of plugin library paths to be serialized with version-compatible engines.
Definition: NvInfer.h:10174
void setDeviceType(ILayer const *layer, DeviceType deviceType) noexcept
Set the device that this layer must execute on.
Definition: NvInfer.h:9585
void setEngineCapability(EngineCapability capability) noexcept
Configure the builder to target specified EngineCapability flow.
Definition: NvInfer.h:9469
int32_t getMaxAuxStreams() const noexcept
Get the maximum number of auxiliary streams that TRT is allowed to use.
Definition: NvInfer.h:10213
bool getFlag(BuilderFlag builderFlag) const noexcept
Returns true if the build mode flag is set.
Definition: NvInfer.h:9568
void setMaxNbTactics(int32_t maxNbTactics) noexcept
Set the maximum number of tactics to time when there is a choice of tactics.
Definition: NvInfer.h:10279
TRT_DEPRECATED void clearQuantizationFlag(QuantizationFlag flag) noexcept
clear a quantization flag.
Definition: NvInfer.h:9862
int64_t getL2LimitForTiling() const noexcept
Get the L2 cache usage limit for tiling optimization.
Definition: NvInfer.h:10347
void setProgressMonitor(IProgressMonitor *monitor) noexcept
Sets the progress monitor for building a network.
Definition: NvInfer.h:10229
void setProfilingVerbosity(ProfilingVerbosity verbosity) noexcept
Set verbosity level of layer information exposed in NVTX annotations and IEngineInspector.
Definition: NvInfer.h:9750
int32_t getNbOptimizationProfiles() const noexcept
Get number of optimization profiles.
Definition: NvInfer.h:9738
nvinfer1::ITimingCache const * getTimingCache() const noexcept
Get the pointer to the timing cache from current IBuilderConfig.
Definition: NvInfer.h:9980
void reset() noexcept
Resets the builder configuration to defaults.
Definition: NvInfer.h:9684
bool setTimingCache(ITimingCache const &cache, bool ignoreMismatch) noexcept
Attach a timing cache to IBuilderConfig.
Definition: NvInfer.h:9970
char const * getPluginToSerialize(int32_t index) const noexcept
Get the plugin library path to be serialized with version-compatible engines.
Definition: NvInfer.h:10164
EngineCapability getEngineCapability() const noexcept
Query EngineCapability flow configured for the builder.
Definition: NvInfer.h:9481
RuntimePlatform getRuntimePlatform() const noexcept
Get the target platform for runtime execution.
Definition: NvInfer.h:10267
DeviceType getDefaultDeviceType() const noexcept
Get the default DeviceType which was set by setDefaultDeviceType.
Definition: NvInfer.h:9674
void setRuntimePlatform(RuntimePlatform runtimePlatform) noexcept
Set the target platform for runtime execution.
Definition: NvInfer.h:10255
int32_t getMaxNbTactics() const noexcept
Query the maximum number of tactics timed when there is a choice.
Definition: NvInfer.h:10291
BuilderFlags getFlags() const noexcept
Get the build mode flags for this builder config. Defaults to 0.
Definition: NvInfer.h:9532
void setFlags(BuilderFlags builderFlags) noexcept
Set the build mode flags to turn on builder options for this network.
Definition: NvInfer.h:9520
TacticSources getTacticSources() const noexcept
Get tactic sources.
Definition: NvInfer.h:9927
void resetDeviceType(ILayer const *layer) noexcept
reset the DeviceType for this layer
Definition: NvInfer.h:9617
void setDLACore(int32_t dlaCore) noexcept
Sets the DLA core used by the network. Defaults to -1.
Definition: NvInfer.h:9643
HardwareCompatibilityLevel getHardwareCompatibilityLevel() const noexcept
Get the hardware compatibility level.
Definition: NvInfer.h:10138
TRT_DEPRECATED QuantizationFlags getQuantizationFlags() const noexcept
Get the quantization flags.
Definition: NvInfer.h:9848
void clearFlag(BuilderFlag builderFlag) noexcept
clear a single build mode flag.
Definition: NvInfer.h:9544
int32_t addOptimizationProfile(IOptimizationProfile const *profile) noexcept
Add an optimization profile.
Definition: NvInfer.h:9725
IProgressMonitor * getProgressMonitor() const noexcept
Definition: NvInfer.h:10239
apiv::VBuilderConfig * mImpl
Definition: NvInfer.h:10353
TRT_DEPRECATED IOptimizationProfile const * getCalibrationProfile() noexcept
Get the current calibration profile.
Definition: NvInfer.h:9815
int32_t getAvgTimingIterations() const noexcept
Query the number of averaging iterations.
Definition: NvInfer.h:9456
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:9664
void setFlag(BuilderFlag builderFlag) noexcept
Set a single build mode flag.
Definition: NvInfer.h:9556
TRT_DEPRECATED bool setCalibrationProfile(IOptimizationProfile const *profile) noexcept
Add a calibration profile.
Definition: NvInfer.h:9803
virtual ~IBuilderConfig() noexcept=default
DeviceType getDeviceType(ILayer const *layer) const noexcept
Get the device that this layer executes on.
Definition: NvInfer.h:9595
bool canRunOnDLA(ILayer const *layer) const noexcept
Checks if a layer can run on DLA.
Definition: NvInfer.h:9627
TRT_DEPRECATED bool getQuantizationFlag(QuantizationFlag flag) const noexcept
Returns true if the quantization flag is set.
Definition: NvInfer.h:9890
cudaStream_t getProfileStream() const noexcept
Get the CUDA stream that is used to profile this network.
Definition: NvInfer.h:9708
void setHardwareCompatibilityLevel(HardwareCompatibilityLevel hardwareCompatibilityLevel) noexcept
Set the hardware compatibility level.
Definition: NvInfer.h:10125
TilingOptimizationLevel getTilingOptimizationLevel() const noexcept
Get the Tiling optimization level.
Definition: NvInfer.h:10319
TRT_DEPRECATED void setQuantizationFlag(QuantizationFlag flag) noexcept
Set a single quantization flag.
Definition: NvInfer.h:9876
void setMaxAuxStreams(int32_t nbStreams) noexcept
Set the maximum number of auxiliary streams that TRT is allowed to use.
Definition: NvInfer.h:10203
ProfilingVerbosity getProfilingVerbosity() const noexcept
Get verbosity level of layer information exposed in NVTX annotations and IEngineInspector.
Definition: NvInfer.h:9763
bool isDeviceTypeSet(ILayer const *layer) const noexcept
whether the DeviceType has been explicitly set for this layer
Definition: NvInfer.h:9607
void setBuilderOptimizationLevel(int32_t level) noexcept
Set builder optimization level.
Definition: NvInfer.h:10096
void setProfileStream(const cudaStream_t stream) noexcept
Set the CUDA stream that is used to profile this network.
Definition: NvInfer.h:9696
TRT_DEPRECATED IAlgorithmSelector * getAlgorithmSelector() const noexcept
Get Algorithm Selector.
Definition: NvInfer.h:9785
Builds an engine from a network definition.
Definition: NvInfer.h:10415
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:10446
int32_t getNbDLACores() const noexcept
Return the number of DLA engines available to this builder.
Definition: NvInfer.h:10454
IErrorRecorder * getErrorRecorder() const noexcept
get the ErrorRecorder assigned to this interface.
Definition: NvInfer.h:10561
apiv::VBuilder * mImpl
Definition: NvInfer.h:10696
ILogger * getLogger() const noexcept
get the logger with which the builder was created
Definition: NvInfer.h:10650
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:10640
int32_t getMaxThreads() const noexcept
get the maximum number of threads that can be used by the builder.
Definition: NvInfer.h:10680
IPluginRegistry & getPluginRegistry() noexcept
get the local plugin registry that can be used by the builder.
Definition: NvInfer.h:10690
TRT_DEPRECATED bool platformHasFastInt8() const noexcept
Determine whether the platform has fast native int8.
Definition: NvInfer.h:10434
nvinfer1::IOptimizationProfile * createOptimizationProfile() noexcept
Create a new optimization profile.
Definition: NvInfer.h:10527
void setGpuAllocator(IGpuAllocator *allocator) noexcept
Set the GPU allocator.
Definition: NvInfer.h:10472
nvinfer1::INetworkDefinition * createNetworkV2(NetworkDefinitionCreationFlags flags) noexcept
Create a network definition object.
Definition: NvInfer.h:10512
nvinfer1::IBuilderConfig * createBuilderConfig() noexcept
Create a builder configuration object.
Definition: NvInfer.h:10486
void reset() noexcept
Resets the builder state to default values.
Definition: NvInfer.h:10569
bool setMaxThreads(int32_t maxThreads) noexcept
Set the maximum number of threads.
Definition: NvInfer.h:10666
void setErrorRecorder(IErrorRecorder *recorder) noexcept
Set the ErrorRecorder for this interface.
Definition: NvInfer.h:10546
nvinfer1::IHostMemory * buildSerializedNetwork(INetworkDefinition &network, IBuilderConfig &config) noexcept
Builds and serializes a network for the given INetworkDefinition and IBuilderConfig.
Definition: NvInfer.h:10598
virtual ~IBuilder() noexcept=default
TRT_DEPRECATED bool platformHasTf32() const noexcept
Determine whether the platform has TF32 support.
Definition: NvInfer.h:10579
nvinfer1::ICudaEngine * buildEngineWithConfig(INetworkDefinition &network, IBuilderConfig &config) noexcept
Builds a network for the given INetworkDefinition and IBuilderConfig.
Definition: NvInfer.h:10618
A cast layer in a network.
Definition: NvInfer.h:3844
virtual ~ICastLayer() noexcept=default
apiv::VCastLayer * mImpl
Definition: NvInfer.h:3870
DataType getToType() const noexcept
Return cast layer output type.
Definition: NvInfer.h:3864
void setToType(DataType toType) noexcept
Set cast layer output type.
Definition: NvInfer.h:3853
A concatenation layer in a network definition.
Definition: NvInfer.h:2077
void setAxis(int32_t axis) noexcept
Set the axis along which concatenation occurs.
Definition: NvInfer.h:2090
int32_t getAxis() const noexcept
Get the axis along which concatenation occurs.
Definition: NvInfer.h:2100
virtual ~IConcatenationLayer() noexcept=default
This layer represents a condition input to an IIfConditional.
Definition: NvInfer.h:4507
virtual ~IConditionLayer() noexcept=default
Layer that represents a constant value.
Definition: NvInfer.h:3883
void setWeights(Weights weights) noexcept
Set the weights for the layer.
Definition: NvInfer.h:3893
Weights getWeights() const noexcept
Get the weights for the layer.
Definition: NvInfer.h:3903
void setDimensions(Dims const &dimensions) noexcept
Set the dimensions for the layer.
Definition: NvInfer.h:3915
apiv::VConstantLayer * mImpl
Definition: NvInfer.h:3933
virtual ~IConstantLayer() noexcept=default
Dims getDimensions() const noexcept
Get the dimensions for the layer.
Definition: NvInfer.h:3927
A convolution layer in a network definition.
Definition: NvInfer.h:1047
Dims getPrePadding() const noexcept
Get the pre-padding.
Definition: NvInfer.h:1172
Weights getBiasWeights() const noexcept
Get the bias weights for the convolution.
Definition: NvInfer.h:1145
void setPaddingMode(PaddingMode paddingMode) noexcept
Set the padding mode.
Definition: NvInfer.h:1213
void setDilationNd(Dims const &dilation) noexcept
Set the multi-dimension dilation of the convolution.
Definition: NvInfer.h:1317
Dims getPaddingNd() const noexcept
Get the multi-dimension padding of the convolution.
Definition: NvInfer.h:1303
Dims getStrideNd() const noexcept
Get the multi-dimension stride of the convolution.
Definition: NvInfer.h:1273
Weights getKernelWeights() const noexcept
Get the kernel weights of the convolution.
Definition: NvInfer.h:1120
void setStrideNd(Dims const &stride) noexcept
Set the multi-dimension stride of the convolution.
Definition: NvInfer.h:1263
Dims getDilationNd() const noexcept
Get the multi-dimension dilation of the convolution.
Definition: NvInfer.h:1327
int64_t getNbOutputMaps() const noexcept
Get the number of output maps for the convolution.
Definition: NvInfer.h:1066
void setKernelWeights(Weights weights) noexcept
Set the kernel weights for the convolution.
Definition: NvInfer.h:1110
Dims getPostPadding() const noexcept
Get the post-padding.
Definition: NvInfer.h:1199
int64_t getNbGroups() const noexcept
Get the number of groups of the convolution.
Definition: NvInfer.h:1096
PaddingMode getPaddingMode() const noexcept
Get the padding mode.
Definition: NvInfer.h:1225
virtual ~IConvolutionLayer() noexcept=default
void setNbGroups(int64_t nbGroups) noexcept
Set the number of groups for a convolution.
Definition: NvInfer.h:1086
void setNbOutputMaps(int64_t nbOutputMaps) noexcept
Set the number of output maps for the convolution.
Definition: NvInfer.h:1056
void setBiasWeights(Weights weights) noexcept
Set the bias weights for the convolution.
Definition: NvInfer.h:1135
Dims getKernelSizeNd() const noexcept
Get the multi-dimension kernel size of the convolution.
Definition: NvInfer.h:1248
void setPaddingNd(Dims const &padding) noexcept
Set the multi-dimension padding of the convolution.
Definition: NvInfer.h:1291
void setPrePadding(Dims const &padding) noexcept
Set the multi-dimension pre-padding of the convolution.
Definition: NvInfer.h:1162
void setPostPadding(Dims const &padding) noexcept
Set the multi-dimension post-padding of the convolution.
Definition: NvInfer.h:1189
void setKernelSizeNd(Dims const &kernelSize) noexcept
Set the multi-dimension kernel size of the convolution.
Definition: NvInfer.h:1238
An engine for executing inference on a built network, with functionally unsafe features.
Definition: NvInferRuntime.h:3111
Layer that represents a cumulative operation across a tensor.
Definition: NvInfer.h:6574
bool setOperation(CumulativeOperation op) noexcept
Set the cumulative operation for the layer.
Definition: NvInfer.h:6585
void setReverse(bool reverse) noexcept
Specify whether the cumulative operation should be applied backward.
Definition: NvInfer.h:6633
apiv::VCumulativeLayer * mImpl
Definition: NvInfer.h:6651
bool getExclusive() const noexcept
Get whether it is exclusive accumulation or inclusive accumulation.
Definition: NvInfer.h:6621
virtual ~ICumulativeLayer() noexcept=default
bool getReverse() const noexcept
Get the boolean that specifies whether the cumulative operation should be applied backward.
Definition: NvInfer.h:6645
void setExclusive(bool exclusive) noexcept
Set whether it is an exclusive accumulation or inclusive accumulation.
Definition: NvInfer.h:6609
CumulativeOperation getOperation() const noexcept
Get the cumulative operation for the layer.
Definition: NvInfer.h:6597
A deconvolution layer in a network definition.
Definition: NvInfer.h:2118
void setBiasWeights(Weights weights) noexcept
Set the bias weights for the deconvolution.
Definition: NvInfer.h:2206
int64_t getNbGroups() const noexcept
Get the number of groups for a deconvolution.
Definition: NvInfer.h:2167
Weights getKernelWeights() const noexcept
Get the kernel weights for the deconvolution.
Definition: NvInfer.h:2191
void setPrePadding(Dims const &padding) noexcept
Set the multi-dimension pre-padding of the deconvolution.
Definition: NvInfer.h:2233
Dims getStrideNd() const noexcept
Get the multi-dimension stride of the deconvolution.
Definition: NvInfer.h:2348
Dims getDilationNd() const noexcept
Get the multi-dimension dilation of the deconvolution.
Definition: NvInfer.h:2414
Weights getBiasWeights() const noexcept
Get the bias weights for the deconvolution.
Definition: NvInfer.h:2216
void setKernelWeights(Weights weights) noexcept
Set the kernel weights for the deconvolution.
Definition: NvInfer.h:2181
int64_t getNbOutputMaps() const noexcept
Get the number of output feature maps for the deconvolution.
Definition: NvInfer.h:2137
void setStrideNd(Dims const &stride) noexcept
Set the multi-dimension stride of the deconvolution.
Definition: NvInfer.h:2338
Dims getPostPadding() const noexcept
Get the padding.
Definition: NvInfer.h:2270
Dims getKernelSizeNd() const noexcept
Get the multi-dimension kernel size of the deconvolution.
Definition: NvInfer.h:2321
void setPostPadding(Dims const &padding) noexcept
Set the multi-dimension post-padding of the deconvolution.
Definition: NvInfer.h:2260
void setKernelSizeNd(Dims const &kernelSize) noexcept
Set the multi-dimension kernel size of the deconvolution.
Definition: NvInfer.h:2311
virtual ~IDeconvolutionLayer() noexcept=default
void setPaddingNd(Dims const &padding) noexcept
Set the multi-dimension padding of the deconvolution.
Definition: NvInfer.h:2366
void setNbOutputMaps(int64_t nbOutputMaps) noexcept
Set the number of output feature maps for the deconvolution.
Definition: NvInfer.h:2127
Dims getPaddingNd() const noexcept
Get the multi-dimension padding of the deconvolution.
Definition: NvInfer.h:2378
void setDilationNd(Dims const &dilation) noexcept
Set the multi-dimension dilation of the deconvolution.
Definition: NvInfer.h:2404
void setPaddingMode(PaddingMode paddingMode) noexcept
Set the padding mode.
Definition: NvInfer.h:2284
void setNbGroups(int64_t nbGroups) noexcept
Set the number of groups for a deconvolution.
Definition: NvInfer.h:2157
Dims getPrePadding() const noexcept
Get the pre-padding.
Definition: NvInfer.h:2243
PaddingMode getPaddingMode() const noexcept
Get the padding mode.
Definition: NvInfer.h:2296
A Dequantize layer in a network definition.
Definition: NvInfer.h:5571
void setToType(DataType toType) noexcept
Set the Dequantize layer output type.
Definition: NvInfer.h:5608
virtual ~IDequantizeLayer() noexcept=default
int32_t getAxis() const noexcept
Get the quantization axis.
Definition: NvInfer.h:5581
DataType getToType() const noexcept
Return the Dequantize layer output type.
Definition: NvInfer.h:5620
void setAxis(int32_t axis) noexcept
Set the quantization axis.
Definition: NvInfer.h:5592
A network layer to perform dynamic quantization.
Definition: NvInfer.h:5648
int32_t getAxis() const noexcept
Get the axis along which blocking occurs.
Definition: NvInfer.h:5736
int32_t getBlockSize() const noexcept
Get the size of the quantization block.
Definition: NvInfer.h:5759
DataType getScaleType() const noexcept
Return the scale factors data type.
Definition: NvInfer.h:5713
void setScaleType(DataType scaleType) noexcept
Set the data type of the scale factors used to quantize the data.
Definition: NvInfer.h:5700
DataType getToType() const noexcept
Return DynamicQuantizeLayer's quantized output type.
Definition: NvInfer.h:5688
virtual ~IDynamicQuantizeLayer() noexcept=default
void setToType(DataType toType) noexcept
Set DynamicQuantizeLayer's quantized output type.
Definition: NvInfer.h:5675
void setAxis(int32_t axis) noexcept
Set the axis along which block quantization occurs.
Definition: NvInfer.h:5726
void setBlockSize(int32_t size) noexcept
Set the size of the quantization block.
Definition: NvInfer.h:5749
An Einsum layer in a network.
Definition: NvInfer.h:5806
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:5817
virtual ~IEinsumLayer() noexcept=default
char const * getEquation() const noexcept
Return the equation.
Definition: NvInfer.h:5827
A elementwise layer in a network definition.
Definition: NvInfer.h:2490
virtual ~IElementWiseLayer() noexcept=default
apiv::VElementWiseLayer * mImpl
Definition: NvInfer.h:2519
ElementWiseOperation getOperation() const noexcept
Get the binary operation for the layer.
Definition: NvInfer.h:2513
void setOperation(ElementWiseOperation op) noexcept
Set the binary operation for the layer.
Definition: NvInfer.h:2501
Generate a tensor according to a specified mode.
Definition: NvInfer.h:5094
bool isAlphaBetaInt64() const noexcept
Return true if alpha/beta have type int64, false if they have type double.
Definition: NvInfer.h:5326
FillOperation getOperation() const noexcept
Get the fill operation for the layer.
Definition: NvInfer.h:5140
void setOperation(FillOperation op) noexcept
Set the fill operation for the layer.
Definition: NvInfer.h:5130
DataType getToType() const noexcept
Get the fill layer output type.
Definition: NvInfer.h:5355
void setAlphaInt64(int64_t alpha) noexcept
Set the alpha parameter with int64 datatype.
Definition: NvInfer.h:5269
void setBetaInt64(int64_t beta) noexcept
Set the beta parameter with int64 datatype.
Definition: NvInfer.h:5303
void setBeta(double beta) noexcept
Set the beta parameter.
Definition: NvInfer.h:5193
int64_t getAlphaInt64() const noexcept
Get the value of alpha parameter with int64 datatype.
Definition: NvInfer.h:5284
int64_t getBetaInt64() const noexcept
Get the value of beta parameter with int64 datatype.
Definition: NvInfer.h:5318
double getAlpha() const noexcept
Get the value of alpha parameter.
Definition: NvInfer.h:5174
void setDimensions(Dims const &dimensions) noexcept
Set the output tensor's dimensions.
Definition: NvInfer.h:5105
void setAlpha(double alpha) noexcept
Set the alpha parameter.
Definition: NvInfer.h:5159
void setToType(DataType toType) noexcept
Set the fill layer output type.
Definition: NvInfer.h:5343
Dims getDimensions() const noexcept
Get the output tensor's dimensions.
Definition: NvInfer.h:5120
double getBeta() const noexcept
Get the value of beta parameter.
Definition: NvInfer.h:5208
virtual ~IFillLayer() noexcept=default
A Gather layer in a network definition. Supports several kinds of gathering.
Definition: NvInfer.h:2623
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:2634
void setNbElementWiseDims(int32_t elementWiseDims) noexcept
Set the number of leading dimensions of indices tensor to be handled elementwise.
Definition: NvInfer.h:2669
apiv::VGatherLayer * mImpl
Definition: NvInfer.h:2705
int32_t getNbElementWiseDims() const noexcept
Get the number of leading dimensions of indices tensor to be handled elementwise.
Definition: NvInfer.h:2679
void setMode(GatherMode mode) noexcept
Set the gather mode.
Definition: NvInfer.h:2689
int32_t getGatherAxis() const noexcept
Get the axis to gather on.
Definition: NvInfer.h:2646
GatherMode getMode() const noexcept
Get the gather mode.
Definition: NvInfer.h:2699
virtual ~IGatherLayer() noexcept=default
A GridSample layer in a network definition.
Definition: NvInfer.h:6028
void setInterpolationMode(InterpolationMode mode) noexcept
Set the grid sample interpolation mode.
Definition: NvInfer.h:6035
bool setSampleMode(SampleMode mode) noexcept
Set the sample mode.
Definition: NvInfer.h:6081
void setAlignCorners(bool alignCorners) noexcept
Set the align corners mode.
Definition: NvInfer.h:6057
apiv::VGridSampleLayer * mImpl
Definition: NvInfer.h:6099
SampleMode getSampleMode() const noexcept
Get the sample mode.
Definition: NvInfer.h:6093
InterpolationMode getInterpolationMode() const noexcept
Get the grid sample interpolation mode.
Definition: NvInfer.h:6047
bool getAlignCorners() const noexcept
Get the align corners mode.
Definition: NvInfer.h:6069
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:3831
apiv::VIdentityLayer * mImpl
Definition: NvInfer.h:3833
virtual ~IIdentityLayer() noexcept=default
This is a base class for Conditional boundary layers.
Definition: NvInfer.h:4486
IIfConditional * getConditional() const noexcept
Get a pointer to the IIfConditional associated with this boundary layer.
Definition: NvInfer.h:4491
virtual ~IIfConditionalBoundaryLayer() noexcept=default
Helper for constructing conditionally-executed subgraphs.
Definition: NvInfer.h:4569
IIfConditionalInputLayer * addInput(ITensor &input) noexcept
Add an If-conditional input.
Definition: NvInfer.h:4610
char const * getName() const noexcept
Return the name of the conditional.
Definition: NvInfer.h:4635
virtual ~IIfConditional() noexcept=default
IConditionLayer * setCondition(ITensor &condition) noexcept
Set the condition tensor for this If-Conditional construct.
Definition: NvInfer.h:4580
IIfConditionalOutputLayer * addOutput(ITensor &trueSubgraphOutput, ITensor &falseSubgraphOutput) noexcept
Add an If-conditional output.
Definition: NvInfer.h:4598
void setName(char const *name) noexcept
Set the name of the conditional.
Definition: NvInfer.h:4625
This layer represents an output of an IIfConditional.
Definition: NvInfer.h:4524
virtual ~IIfConditionalOutputLayer() noexcept=default
Application-implemented interface for calibration.
Definition: NvInfer.h:8165
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:4800
virtual ~IIteratorLayer() noexcept=default
void setReverse(bool reverse) noexcept
Set iteration order to be reverse.
Definition: NvInfer.h:4827
bool getReverse() const noexcept
Check if the iteration order is reverse.
Definition: NvInfer.h:4837
int32_t getAxis() const noexcept
Get axis being iterated over.
Definition: NvInfer.h:4813
void setAxis(int32_t axis) noexcept
Set axis to iterate over.
Definition: NvInfer.h:4805
A LRN layer in a network definition.
Definition: NvInfer.h:1732
int64_t getWindowSize() const noexcept
Get the LRN window size.
Definition: NvInfer.h:1753
float getAlpha() const noexcept
Get the LRN alpha value.
Definition: NvInfer.h:1775
void setWindowSize(int64_t windowSize) noexcept
Set the LRN window size.
Definition: NvInfer.h:1743
void setK(float k) noexcept
Set the LRN K value.
Definition: NvInfer.h:1809
void setAlpha(float alpha) noexcept
Set the LRN alpha value.
Definition: NvInfer.h:1765
void setBeta(float beta) noexcept
Set the LRN beta value.
Definition: NvInfer.h:1787
virtual ~ILRNLayer() noexcept=default
float getBeta() const noexcept
Get the LRN beta value.
Definition: NvInfer.h:1797
float getK() const noexcept
Get the LRN K value.
Definition: NvInfer.h:1819
Base class for all layer classes in a network definition.
Definition: NvInfer.h:575
bool precisionIsSet() const noexcept
whether the computational precision has been set for this layer
Definition: NvInfer.h:717
void setMetadata(char const *metadata) noexcept
Set the metadata for this layer.
Definition: NvInfer.h:833
void setPrecision(DataType dataType) noexcept
Set the preferred or required computational precision of this layer in a weakly-typed network.
Definition: NvInfer.h:693
void setName(char const *name) noexcept
Set the name of a layer.
Definition: NvInfer.h:596
void resetPrecision() noexcept
reset the computational precision for this layer
Definition: NvInfer.h:727
int32_t getNbInputs() const noexcept
Get the number of inputs of a layer.
Definition: NvInfer.h:614
char const * getMetadata() const noexcept
Get the metadata of the layer.
Definition: NvInfer.h:846
DataType getOutputType(int32_t index) const noexcept
get the output type of this layer
Definition: NvInfer.h:789
DataType getPrecision() const noexcept
get the computational precision of this layer
Definition: NvInfer.h:705
char const * getName() const noexcept
Return the name of a layer.
Definition: NvInfer.h:606
int32_t getNbOutputs() const noexcept
Get the number of outputs of a layer.
Definition: NvInfer.h:635
bool outputTypeIsSet(int32_t index) const noexcept
whether the output type has been set for this layer
Definition: NvInfer.h:803
ITensor * getOutput(int32_t index) const noexcept
Get the layer output corresponding to the given index.
Definition: NvInfer.h:645
void setInput(int32_t index, ITensor &tensor) noexcept
Replace an input of this layer with a specific tensor.
Definition: NvInfer.h:662
void resetOutputType(int32_t index) noexcept
reset the output type for this layer
Definition: NvInfer.h:815
ITensor * getInput(int32_t index) const noexcept
Get the layer input corresponding to the given index.
Definition: NvInfer.h:627
void setOutputType(int32_t index, DataType dataType) noexcept
Set the output type of this layer in a weakly-typed network.
Definition: NvInfer.h:774
LayerType getType() const noexcept
Return the type of a layer.
Definition: NvInfer.h:582
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:4463
virtual ~ILoopBoundaryLayer() noexcept=default
ILoop * getLoop() const noexcept
Get a pointer to ILoop associated with this boundary layer.
Definition: NvInfer.h:4468
Helper for creating a recurrent subgraph.
Definition: NvInfer.h:4858
void setName(char const *name) noexcept
Set the name of the loop.
Definition: NvInfer.h:4928
ITripLimitLayer * addTripLimit(ITensor &tensor, TripLimit limit) noexcept
Add a trip-count limiter, based on the given tensor.
Definition: NvInfer.h:4887
IIteratorLayer * addIterator(ITensor &tensor, int32_t axis=0, bool reverse=false) noexcept
Return layer that subscripts tensor by loop iteration.
Definition: NvInfer.h:4900
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:4913
virtual ~ILoop() noexcept=default
char const * getName() const noexcept
Return the name of the loop.
Definition: NvInfer.h:4938
IRecurrenceLayer * addRecurrence(ITensor &initialValue) noexcept
Create a recurrence layer for this loop with initialValue as its first input.
Definition: NvInfer.h:4866
An ILoopOutputLayer is the sole way to get output from a loop.
Definition: NvInfer.h:4700
virtual ~ILoopOutputLayer() noexcept=default
int32_t getAxis() const noexcept
Get axis being concatenated over.
Definition: NvInfer.h:4730
LoopOutput getLoopOutput() const noexcept
Get which kind a loop output has.
Definition: NvInfer.h:4705
void setAxis(int32_t axis) noexcept
Set where to insert the contenation axis. Ignored if getLoopOutput() is kLAST_VALUE.
Definition: NvInfer.h:4722
Layer that represents a Matrix Multiplication.
Definition: NvInfer.h:3706
apiv::VMatrixMultiplyLayer * mImpl
Definition: NvInfer.h:3734
virtual ~IMatrixMultiplyLayer() noexcept=default
MatrixOperation getOperation(int32_t index) const noexcept
Get the operation for an input tensor.
Definition: NvInfer.h:3728
void setOperation(int32_t index, MatrixOperation op) noexcept
Set the operation for an input tensor.
Definition: NvInfer.h:3716
A non-maximum suppression layer in a network definition.
Definition: NvInfer.h:6176
virtual ~INMSLayer() noexcept=default
void setTopKBoxLimit(int32_t limit) noexcept
Set the TopK box limit parameter for the layer.
Definition: NvInfer.h:6213
void setBoundingBoxFormat(BoundingBoxFormat fmt) noexcept
Set the bounding box format parameter for the layer.
Definition: NvInfer.h:6187
BoundingBoxFormat getBoundingBoxFormat() const noexcept
Get the bounding box format parameter for the layer.
Definition: NvInfer.h:6199
apiv::VNMSLayer * mImpl
Definition: NvInfer.h:6249
int32_t getTopKBoxLimit() const noexcept
Get the TopK box limit parameter for the layer.
Definition: NvInfer.h:6223
A network definition for input to the builder.
Definition: NvInfer.h:6673
IConcatenationLayer * addConcatenation(ITensor *const *inputs, int32_t nbInputs) noexcept
Add a concatenation layer to the network.
Definition: NvInfer.h:6865
IShuffleLayer * addShuffle(ITensor &input) noexcept
Add a shuffle layer to the network.
Definition: NvInfer.h:6928
INormalizationLayer * addNormalization(ITensor &input, ITensor &scale, ITensor &bias, uint32_t axesMask) noexcept
Add a normalization layer to the network.
Definition: NvInfer.h:8005
void setName(char const *name) noexcept
Sets the name of the network.
Definition: NvInfer.h:7338
bool markDebug(ITensor &tensor) noexcept
Mark a tensor as a debug tensor.
Definition: NvInfer.h:6744
ILRNLayer * addLRN(ITensor &input, int64_t window, float alpha, float beta, float k) noexcept
Add a LRN layer to the network.
Definition: NvInfer.h:6809
ITopKLayer * addTopK(ITensor &input, TopKOperation op, int32_t k, uint32_t reduceAxes) noexcept
Add a TopK layer to the network.
Definition: NvInfer.h:7088
ICumulativeLayer * addCumulative(ITensor &input, ITensor &axis, CumulativeOperation operation, bool exclusive, bool reverse) noexcept
Add a cumulative layer to the network.
Definition: NvInfer.h:8027
IAssertionLayer * addAssertion(ITensor &condition, char const *message) noexcept
Add an assertion layer to the network.
Definition: NvInfer.h:7654
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:7473
ICastLayer * addCast(ITensor &input, DataType toType) noexcept
Add a cast layer.
Definition: NvInfer.h:7228
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:7552
char const * getName() const noexcept
Returns the name associated with the network.
Definition: NvInfer.h:7352
IParametricReLULayer * addParametricReLU(ITensor &input, ITensor &slope) noexcept
Add a parametric ReLU layer to the network.
Definition: NvInfer.h:7451
ITensor * getOutput(int32_t index) const noexcept
Get the output tensor specified by the given index.
Definition: NvInfer.h:7029
ITensor * getInput(int32_t index) const noexcept
Get the input tensor specified by the given index.
Definition: NvInfer.h:6999
IDequantizeLayer * addDequantize(ITensor &input, ITensor &scale, DataType outputType) noexcept
Add a dequantization layer to the network.
Definition: NvInfer.h:7822
bool unmarkOutputForShapes(ITensor &tensor) noexcept
Undo markOutputForShapes.
Definition: NvInfer.h:7433
IFillLayer * addFill(Dims const &dimensions, FillOperation op, DataType outputType) noexcept
Add a fill layer to the network.
Definition: NvInfer.h:7705
ILoop * addLoop() noexcept
Add a loop to the network.
Definition: NvInfer.h:7583
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:7910
IActivationLayer * addActivation(ITensor &input, ActivationType type) noexcept
Add an activation layer to the network.
Definition: NvInfer.h:6790
TRT_DEPRECATED IFillLayer * addFill(Dims const &dimensions, FillOperation op) noexcept
Add a fill layer to the network.
Definition: NvInfer.h:7679
ISliceLayer * addSlice(ITensor &input, Dims const &start, Dims const &size, Dims const &stride) noexcept
Add a slice layer to the network.
Definition: NvInfer.h:7314
virtual ~INetworkDefinition() noexcept=default
TRT_DEPRECATED IQuantizeLayer * addQuantize(ITensor &input, ITensor &scale) noexcept
Add a quantization layer to the network.
Definition: NvInfer.h:7863
virtual IBuilder & getBuilder() const noexcept
Return the builder from which this INetworkDefinition was created.
Definition: NvInfer.h:8038
INMSLayer * addNMS(ITensor &boxes, ITensor &scores, ITensor &maxOutputBoxesPerClass) noexcept
Add a non-maximum suppression layer to the network.
Definition: NvInfer.h:7962
ILayer * getLayer(int32_t index) const noexcept
Get the layer specified by the given index.
Definition: NvInfer.h:6971
bool getFlag(NetworkDefinitionCreationFlag networkDefinitionCreationFlag) const noexcept
Returns true if the network definition creation flag is set.
Definition: NvInfer.h:7404
IIfConditional * addIfConditional() noexcept
Add an if-then-else to the network.
Definition: NvInfer.h:7598
IErrorRecorder * getErrorRecorder() const noexcept
get the ErrorRecorder assigned to this interface.
Definition: NvInfer.h:7779
ISqueezeLayer * addSqueeze(ITensor &input, ITensor &axes) noexcept
Add a squeeze layer to the network.
Definition: NvInfer.h:8095
IReverseSequenceLayer * addReverseSequence(ITensor &input, ITensor &sequenceLens) noexcept
Add a ReverseSequence layer to the network.
Definition: NvInfer.h:7979
int32_t getNbInputs() const noexcept
Get the number of inputs in the network.
Definition: NvInfer.h:6983
NetworkDefinitionCreationFlags getFlags() const noexcept
Get the network definition creation flags for this network definition object. Defaults to 0.
Definition: NvInfer.h:7392
IQuantizeLayer * addQuantize(ITensor &input, ITensor &scale, DataType outputType) noexcept
Add a quantization layer to the network.
Definition: NvInfer.h:7885
IReduceLayer * addReduce(ITensor &input, ReduceOperation operation, uint32_t reduceAxes, bool keepDimensions) noexcept
Add a reduce layer to the network.
Definition: NvInfer.h:7055
IUnaryLayer * addUnary(ITensor &input, UnaryOperation operation) noexcept
Add a unary layer to the network.
Definition: NvInfer.h:6914
IGridSampleLayer * addGridSample(ITensor &input, ITensor &grid) noexcept
Add a GridSample layer to the network.
Definition: NvInfer.h:7944
void removeTensor(ITensor &tensor) noexcept
remove a tensor from the network definition.
Definition: NvInfer.h:7243
bool areWeightsMarkedRefittable(char const *name) const noexcept
Whether the weight has been marked as refittable.
Definition: NvInfer.h:8076
ISelectLayer * addSelect(ITensor &condition, ITensor &thenInput, ITensor &elseInput) noexcept
Add a select layer to the network.
Definition: NvInfer.h:7637
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:7842
int32_t getNbLayers() const noexcept
Get the number of layers in the network.
Definition: NvInfer.h:6957
TRT_DEPRECATED bool hasImplicitBatchDimension() const noexcept
Query whether the network was created with an implicit batch dimension.
Definition: NvInfer.h:7382
apiv::VNetworkDefinition * mImpl
Definition: NvInfer.h:8122
bool markOutputForShapes(ITensor &tensor) noexcept
Enable tensor's value to be computed by IExecutionContext::getShapeBinding.
Definition: NvInfer.h:7421
IOneHotLayer * addOneHot(ITensor &indices, ITensor &values, ITensor &depth, int32_t axis) noexcept
Add a OneHot layer to the network.
Definition: NvInfer.h:6945
IScaleLayer * addScale(ITensor &input, ScaleMode mode, Weights shift, Weights scale, Weights power) noexcept
Add a Scale layer to the network.
Definition: NvInfer.h:6835
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:7294
void unmarkOutput(ITensor &tensor) noexcept
unmark a tensor as a network output.
Definition: NvInfer.h:7255
IIdentityLayer * addIdentity(ITensor &input) noexcept
Add an identity layer.
Definition: NvInfer.h:7213
IGatherLayer * addGatherV2(ITensor &data, ITensor &indices, GatherMode mode) noexcept
Add gather with specified mode, axis=0 and nbElementWiseDims=0.
Definition: NvInfer.h:7120
IElementWiseLayer * addElementWise(ITensor &input1, ITensor &input2, ElementWiseOperation op) noexcept
Add an elementwise layer to the network.
Definition: NvInfer.h:6892
IConstantLayer * addConstant(Dims const &dimensions, Weights weights) noexcept
Add a constant layer to the network.
Definition: NvInfer.h:7199
void setErrorRecorder(IErrorRecorder *recorder) noexcept
Set the ErrorRecorder for this interface.
Definition: NvInfer.h:7764
IPoolingLayer * addPoolingNd(ITensor &input, PoolingType type, Dims const &windowSize) noexcept
Add a multi-dimension pooling layer to the network.
Definition: NvInfer.h:7493
IRaggedSoftMaxLayer * addRaggedSoftMax(ITensor &input, ITensor &bounds) noexcept
Add a RaggedSoftMax layer to the network.
Definition: NvInfer.h:7139
IShapeLayer * addShape(ITensor &input) noexcept
Add a shape layer to the network.
Definition: NvInfer.h:7368
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:7104
bool unmarkWeightsRefittable(char const *name) noexcept
Unmark weights as refittable when the builder flag kREFIT_INDIVIDUAL is set.
Definition: NvInfer.h:8063
bool markWeightsRefittable(char const *name) noexcept
Mark weights as refittable when the builder flag kREFIT_INDIVIDUAL is set.
Definition: NvInfer.h:8051
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:7515
IResizeLayer * addResize(ITensor &input) noexcept
Add a resize layer to the network.
Definition: NvInfer.h:7569
IUnsqueezeLayer * addUnsqueeze(ITensor &input, ITensor &axes) noexcept
Add an unsqueeze layer to the network.
Definition: NvInfer.h:8116
IMatrixMultiplyLayer * addMatrixMultiply(ITensor &input0, MatrixOperation op0, ITensor &input1, MatrixOperation op1) noexcept
Add a MatrixMultiply layer to the network.
Definition: NvInfer.h:7160
ISoftMaxLayer * addSoftMax(ITensor &input) noexcept
Add a SoftMax layer to the network.
Definition: NvInfer.h:6848
bool isDebugTensor(nvinfer1::ITensor const &tensor) const noexcept
Check if a tensor is marked as debug tensor.
Definition: NvInfer.h:6770
bool unmarkDebug(ITensor &tensor) noexcept
Unmark a tensor as a debug tensor.
Definition: NvInfer.h:6760
IEinsumLayer * addEinsum(ITensor *const *inputs, int32_t nbInputs, char const *equation) noexcept
Add an Einsum layer to the network.
Definition: NvInfer.h:7926
void markOutput(ITensor &tensor) noexcept
Mark a tensor as a network output.
Definition: NvInfer.h:6726
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:7276
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:7721
INonZeroLayer * addNonZero(ITensor &input) noexcept
Add a nonzero layer to the network.
Definition: NvInfer.h:7175
TRT_DEPRECATED IDequantizeLayer * addDequantize(ITensor &input, ITensor &scale) noexcept
Add a dequantization layer to the network.
Definition: NvInfer.h:7800
int32_t getNbOutputs() const noexcept
Get the number of outputs in the network.
Definition: NvInfer.h:7013
bool setWeightsName(Weights weights, char const *name) noexcept
Associate a name with all current uses of the given weights.
Definition: NvInfer.h:7745
Forward declaration of IEngineInspector for use by other interfaces.
Definition: NvInferRuntime.h:51
Definition: NvInfer.h:3760
virtual ~INonZeroLayer() noexcept=default
A normalization layer in a network definition.
Definition: NvInfer.h:6338
float getEpsilon() const noexcept
Get the epsilon value used for the normalization calculation.
Definition: NvInfer.h:6357
uint32_t getAxes() const noexcept
Get the axes value used for the normalization calculation.
Definition: NvInfer.h:6377
virtual ~INormalizationLayer() noexcept=default
void setEpsilon(float eps) noexcept
Set the epsilon value used for the normalization calculation.
Definition: NvInfer.h:6347
DataType getComputePrecision() const noexcept
Get the compute precision of this layer.
Definition: NvInfer.h:6444
apiv::VNormalizationLayer * mImpl
Definition: NvInfer.h:6450
int64_t getNbGroups() const noexcept
Get the number of groups used to split the channels for the normalization calculation.
Definition: NvInfer.h:6408
void setAxes(uint32_t axesMask) noexcept
Set the reduction axes for the normalization calculation.
Definition: NvInfer.h:6367
void setComputePrecision(DataType type) noexcept
Set the compute precision of this layer.
Definition: NvInfer.h:6434
void setNbGroups(int64_t nbGroups) noexcept
Set the number of groups used to split the channels in the normalization calculation.
Definition: NvInfer.h:6398
A OneHot layer in a network definition.
Definition: NvInfer.h:5991
virtual ~IOneHotLayer() noexcept=default
apiv::VOneHotLayer * mImpl
Definition: NvInfer.h:6012
void setAxis(int32_t axis) noexcept
Set the axis parameter.
Definition: NvInfer.h:5998
int32_t getAxis() const noexcept
Get the value of the axis parameter.
Definition: NvInfer.h:6006
Optimization profile for dynamic input dimensions and shape tensors.
Definition: NvInferRuntime.h:2618
Layer that represents a padding operation.
Definition: NvInfer.h:2984
Dims getPostPaddingNd() const noexcept
Get the padding that is applied at the end of the tensor.
Definition: NvInfer.h:3033
void setPrePaddingNd(Dims const &padding) noexcept
Set the padding that is applied at the start of the tensor.
Definition: NvInfer.h:2995
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:3021
Dims getPrePaddingNd() const noexcept
Get the padding that is applied at the start of the tensor.
Definition: NvInfer.h:3007
apiv::VPaddingLayer * mImpl
Definition: NvInfer.h:3039
Layer that represents a parametric ReLU operation.
Definition: NvInfer.h:3947
apiv::VParametricReLULayer * mImpl
Definition: NvInfer.h:3949
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:139
Layer type for pluginV2.
Definition: NvInfer.h:2721
virtual ~IPluginV2Layer() noexcept=default
apiv::VPluginV2Layer * mImpl
Definition: NvInfer.h:2734
IPluginV2 & getPlugin() noexcept
Get the plugin for the layer.
Definition: NvInfer.h:2728
Layer type for V3 plugins.
Definition: NvInfer.h:2748
virtual ~IPluginV3Layer() noexcept=default
IPluginV3 & getPlugin() noexcept
Get the plugin for the layer.
Definition: NvInfer.h:2755
apiv::VPluginV3Layer * mImpl
Definition: NvInfer.h:2761
A Pooling layer in a network definition.
Definition: NvInfer.h:1481
PoolingType getPoolingType() const noexcept
Get the type of activation to be performed.
Definition: NvInfer.h:1500
PaddingMode getPaddingMode() const noexcept
Get the padding mode.
Definition: NvInfer.h:1633
Dims getPostPadding() const noexcept
Get the padding.
Definition: NvInfer.h:1609
bool getAverageCountExcludesPadding() const noexcept
Get whether average pooling uses as a denominator the overlap area between the window and the unpadde...
Definition: NvInfer.h:1553
Dims getPrePadding() const noexcept
Get the pre-padding.
Definition: NvInfer.h:1581
void setPoolingType(PoolingType type) noexcept
Set the type of activation to be performed.
Definition: NvInfer.h:1490
void setWindowSizeNd(Dims const &windowSize) noexcept
Set the multi-dimension window size for pooling.
Definition: NvInfer.h:1646
void setPaddingMode(PaddingMode paddingMode) noexcept
Set the padding mode.
Definition: NvInfer.h:1622
Dims getWindowSizeNd() const noexcept
Get the multi-dimension window size for pooling.
Definition: NvInfer.h:1656
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:1542
void setPaddingNd(Dims const &padding) noexcept
Set the multi-dimension padding for pooling.
Definition: NvInfer.h:1700
float getBlendFactor() const noexcept
Get the blending factor for the max_average_blend mode: max_average_blendPool = (1-blendFactor)*maxPo...
Definition: NvInfer.h:1528
void setStrideNd(Dims const &stride) noexcept
Set the multi-dimension stride for pooling.
Definition: NvInfer.h:1671
Dims getStrideNd() const noexcept
Get the multi-dimension stride for pooling.
Definition: NvInfer.h:1681
virtual ~IPoolingLayer() noexcept=default
Dims getPaddingNd() const noexcept
Get the multi-dimension padding for pooling.
Definition: NvInfer.h:1712
void setPostPadding(Dims const &padding) noexcept
Set the multi-dimension post-padding for pooling.
Definition: NvInfer.h:1599
void setPrePadding(Dims const &padding) noexcept
Set the multi-dimension pre-padding for pooling.
Definition: NvInfer.h:1571
void setBlendFactor(float blendFactor) noexcept
Set the blending factor for the max_average_blend mode: max_average_blendPool = (1-blendFactor)*maxPo...
Definition: NvInfer.h:1515
A Quantize layer in a network definition.
Definition: NvInfer.h:5440
void setToType(DataType toType) noexcept
Set the Quantize layer output type.
Definition: NvInfer.h:5477
void setAxis(int32_t axis) noexcept
Set the quantization axis.
Definition: NvInfer.h:5461
int32_t getAxis() const noexcept
Get the quantization axis.
Definition: NvInfer.h:5450
virtual ~IQuantizeLayer() noexcept=default
DataType getToType() const noexcept
Return the Quantize layer output type.
Definition: NvInfer.h:5489
A RaggedSoftmax layer in a network definition.
Definition: NvInfer.h:3781
apiv::VRaggedSoftMaxLayer * mImpl
Definition: NvInfer.h:3783
virtual ~IRaggedSoftMaxLayer() noexcept=default
A recurrence layer in a network definition.
Definition: NvInfer.h:4653
virtual ~IRecurrenceLayer() noexcept=default
Layer that represents a reduction across a non-bool tensor.
Definition: NvInfer.h:2904
void setKeepDimensions(bool keepDimensions) noexcept
Set the boolean that specifies whether or not to keep the reduced dimensions for the layer.
Definition: NvInfer.h:2951
void setOperation(ReduceOperation op) noexcept
Set the reduce operation for the layer.
Definition: NvInfer.h:2911
ReduceOperation getOperation() const noexcept
Get the reduce operation for the layer.
Definition: NvInfer.h:2921
virtual ~IReduceLayer() noexcept=default
uint32_t getReduceAxes() const noexcept
Get the axes over which to reduce for the layer.
Definition: NvInfer.h:2941
void setReduceAxes(uint32_t reduceAxes) noexcept
Set the axes over which to reduce.
Definition: NvInfer.h:2931
apiv::VReduceLayer * mImpl
Definition: NvInfer.h:2967
bool getKeepDimensions() const noexcept
Get the boolean that specifies whether or not to keep the reduced dimensions for the layer.
Definition: NvInfer.h:2961
A resize layer in a network definition.
Definition: NvInfer.h:4136
void setSelectorForSinglePixel(ResizeSelector selector) noexcept
Set coordinate selector function when resized to single pixel.
Definition: NvInfer.h:4297
void setNearestRounding(ResizeRoundMode value) noexcept
Set rounding mode for nearest neighbor resize.
Definition: NvInfer.h:4321
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:4215
void setOutputDimensions(Dims const &dimensions) noexcept
Set the output dimensions.
Definition: NvInfer.h:4156
void setCubicCoeff(float A) noexcept
Set the coefficient 'A' used in cubic interpolation.
Definition: NvInfer.h:4353
void setScales(float const *scales, int32_t nbScales) noexcept
Set the resize scales.
Definition: NvInfer.h:4196
float getCubicCoeff() const noexcept
Get the coefficient 'A' used in cubic interpolation.
Definition: NvInfer.h:4363
ResizeSelector getSelectorForSinglePixel() const noexcept
Get the coordinate selector function when resized to single pixel.
Definition: NvInfer.h:4307
InterpolationMode getResizeMode() const noexcept
Get resize mode for an input tensor.
Definition: NvInfer.h:4237
void setCoordinateTransformation(ResizeCoordinateTransformation coordTransform) noexcept
Set coordinate transformation function.
Definition: NvInfer.h:4272
void setExcludeOutside(bool excludeFlag) noexcept
Set the state for excluding outside pixels.
Definition: NvInfer.h:4376
void setResizeMode(InterpolationMode interpolationMode) noexcept
Set resize mode for an input tensor.
Definition: NvInfer.h:4227
Dims getOutputDimensions() const noexcept
Get the output dimensions.
Definition: NvInfer.h:4166
ResizeRoundMode getNearestRounding() const noexcept
Get rounding mode for nearest neighbor resize.
Definition: NvInfer.h:4331
bool getExcludeOutside() const noexcept
Get the state for excluding outside pixels.
Definition: NvInfer.h:4386
ResizeCoordinateTransformation getCoordinateTransformation() const noexcept
Get coordinate transformation function.
Definition: NvInfer.h:4282
A ReverseSequence layer in a network definition.
Definition: NvInfer.h:6266
void setSequenceAxis(int32_t sequenceAxis) noexcept
Set the sequence axis. Default is 0.
Definition: NvInfer.h:6299
int32_t getBatchAxis() const noexcept
Return the batch axis. Return 1 if no batch axis was set.
Definition: NvInfer.h:6286
apiv::VReverseSequenceLayer * mImpl
Definition: NvInfer.h:6315
int32_t getSequenceAxis() const noexcept
Return the sequence axis. Return 0 if no sequence axis was set.
Definition: NvInfer.h:6309
void setBatchAxis(int32_t batchAxis) noexcept
Set the batch axis. Default is 1.
Definition: NvInfer.h:6276
virtual ~IReverseSequenceLayer() noexcept=default
A Scale layer in a network definition.
Definition: NvInfer.h:1878
Weights getScale() const noexcept
Get the scale value.
Definition: NvInfer.h:1935
Weights getPower() const noexcept
Get the power value.
Definition: NvInfer.h:1955
void setScale(Weights scale) noexcept
Set the scale value.
Definition: NvInfer.h:1925
void setPower(Weights power) noexcept
Set the power value.
Definition: NvInfer.h:1945
ScaleMode getMode() const noexcept
Get the scale mode.
Definition: NvInfer.h:1895
void setShift(Weights shift) noexcept
Set the shift value.
Definition: NvInfer.h:1905
void setChannelAxis(int32_t channelAxis) noexcept
Set the channel axis.
Definition: NvInfer.h:1991
Weights getShift() const noexcept
Get the shift value.
Definition: NvInfer.h:1915
virtual ~IScaleLayer() noexcept=default
void setMode(ScaleMode mode) noexcept
Set the scale mode.
Definition: NvInfer.h:1885
int32_t getChannelAxis() const noexcept
Get the channel axis.
Definition: NvInfer.h:1970
A scatter layer in a network definition. Supports several kinds of scattering.
Definition: NvInfer.h:5919
void setMode(ScatterMode mode) noexcept
Set the scatter mode.
Definition: NvInfer.h:5926
apiv::VScatterLayer * mImpl
Definition: NvInfer.h:5960
void setAxis(int32_t axis) noexcept
Set the axis used by ScatterMode::kELEMENTS.
Definition: NvInfer.h:5946
int32_t getAxis() const noexcept
Get the axis.
Definition: NvInfer.h:5954
ScatterMode getMode() const noexcept
Get the scatter mode.
Definition: NvInfer.h:5936
virtual ~IScatterLayer() noexcept=default
Select elements from two data tensors based on a condition tensor.
Definition: NvInfer.h:4961
virtual ~ISelectLayer() noexcept=default
Layer type for getting shape of a tensor.
Definition: NvInfer.h:3509
virtual ~IShapeLayer() noexcept=default
apiv::VShapeLayer * mImpl
Definition: NvInfer.h:3511
Layer type for shuffling data.
Definition: NvInfer.h:3072
apiv::VShuffleLayer * mImpl
Definition: NvInfer.h:3230
void setFirstTranspose(Permutation permutation) noexcept
Set the permutation applied by the first transpose operation.
Definition: NvInfer.h:3083
void setSecondTranspose(Permutation permutation) noexcept
Set the permutation applied by the second transpose operation.
Definition: NvInfer.h:3183
Dims getReshapeDimensions() const noexcept
Get the reshaped dimensions.
Definition: NvInfer.h:3136
void setReshapeDimensions(Dims const &dimensions) noexcept
Set the reshaped dimensions.
Definition: NvInfer.h:3123
Permutation getFirstTranspose() const noexcept
Get the permutation applied by the first transpose operation.
Definition: NvInfer.h:3095
virtual ~IShuffleLayer() noexcept=default
Permutation getSecondTranspose() const noexcept
Get the permutation applied by the second transpose operation.
Definition: NvInfer.h:3195
bool getZeroIsPlaceholder() const noexcept
Get meaning of 0 in reshape dimensions.
Definition: NvInfer.h:3224
void setZeroIsPlaceholder(bool zeroIsPlaceholder) noexcept
Set meaning of 0 in reshape dimensions.
Definition: NvInfer.h:3211
Slices an input tensor into an output tensor based on the offset and strides.
Definition: NvInfer.h:3324
void setStride(Dims const &stride) noexcept
Set the stride for computing the output slice data.
Definition: NvInfer.h:3393
apiv::VSliceLayer * mImpl
Definition: NvInfer.h:3492
virtual ~ISliceLayer() noexcept=default
void setSize(Dims const &size) noexcept
Set the dimensions of the output slice.
Definition: NvInfer.h:3364
void setAxes(Dims const &axes) noexcept
Set the axes for this ISliceLayer.
Definition: NvInfer.h:3471
void setStart(Dims const &start) noexcept
Set the start offset that the slice layer uses to create the output slice.
Definition: NvInfer.h:3335
Dims getStart() const noexcept
Get the start offset for the slice layer.
Definition: NvInfer.h:3350
void setMode(SampleMode mode) noexcept
Set the slice mode.
Definition: NvInfer.h:3418
Dims getSize() const noexcept
Get dimensions of the output slice.
Definition: NvInfer.h:3379
SampleMode getMode() const noexcept
Get the slice mode.
Definition: NvInfer.h:3428
Dims getStride() const noexcept
Get the stride for the output slice.
Definition: NvInfer.h:3408
Dims getAxes() const noexcept
Get the axes for this ISliceLayer.
Definition: NvInfer.h:3486
A Softmax layer in a network definition.
Definition: NvInfer.h:2022
void setAxes(uint32_t axes) noexcept
Set the axis along which softmax is computed. Currently, only one axis can be set.
Definition: NvInfer.h:2044
uint32_t getAxes() const noexcept
Get the axis along which softmax occurs.
Definition: NvInfer.h:2054
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:6464
virtual ~ISqueezeLayer() noexcept=default
apiv::VSqueezeLayer * mImpl
Definition: NvInfer.h:6481
A tensor in a network definition.
Definition: NvInfer.h:183
void setAllowedFormats(TensorFormats formats) noexcept
Set allowed formats for an input or output tensor. By default all formats are allowed....
Definition: NvInfer.h:451
TensorLocation getLocation() const noexcept
Get the storage location of a tensor.
Definition: NvInfer.h:370
void setDimensions(Dims const &dimensions) noexcept
Set the dimensions of a tensor.
Definition: NvInfer.h:231
void resetDynamicRange() noexcept
Undo effect of setDynamicRange.
Definition: NvInfer.h:409
void setName(char const *name) noexcept
Set the tensor name.
Definition: NvInfer.h:200
bool isExecutionTensor() const noexcept
Whether the tensor is an execution tensor.
Definition: NvInfer.h:516
void setType(DataType type) noexcept
Set the data type of a tensor.
Definition: NvInfer.h:279
TRT_DEPRECATED bool dynamicRangeIsSet() const noexcept
Query whether dynamic range is set.
Definition: NvInfer.h:401
char const * getName() const noexcept
Get the tensor name.
Definition: NvInfer.h:212
bool isShapeTensor() const noexcept
Whether the tensor is a shape tensor.
Definition: NvInfer.h:495
float getDynamicRangeMax() const noexcept
Get maximum of dynamic range.
Definition: NvInfer.h:429
bool isNetworkInput() const noexcept
Whether the tensor is a network input.
Definition: NvInfer.h:319
TRT_DEPRECATED void setBroadcastAcrossBatch(bool broadcastAcrossBatch) noexcept
Set whether to enable broadcast of tensor across the implicit batch dimension.
Definition: NvInfer.h:344
TRT_DEPRECATED bool setDynamicRange(float min, float max) noexcept
Set dynamic range for the tensor.
Definition: NvInfer.h:311
TRT_DEPRECATED bool getBroadcastAcrossBatch() const noexcept
Check if tensor is broadcast across the implicit batch dimension.
Definition: NvInfer.h:358
bool isNetworkOutput() const noexcept
Whether the tensor is a network output.
Definition: NvInfer.h:327
DataType getType() const noexcept
Get the data type of a tensor.
Definition: NvInfer.h:294
apiv::VTensor * mImpl
Definition: NvInfer.h:563
float getDynamicRangeMin() const noexcept
Get minimum of dynamic range.
Definition: NvInfer.h:419
virtual ~ITensor() noexcept=default
void setDimensionName(int32_t index, char const *name) noexcept
Name a dimension of an input tensor.
Definition: NvInfer.h:542
char const * getDimensionName(int32_t index) const noexcept
Get the name of an input dimension.
Definition: NvInfer.h:557
TRT_DEPRECATED void setLocation(TensorLocation location) noexcept
Set the storage location of a tensor.
Definition: NvInfer.h:389
Dims getDimensions() const noexcept
Get the dimensions of a tensor.
Definition: NvInfer.h:245
TensorFormats getAllowedFormats() const noexcept
Get a bitmask of TensorFormat values that the tensor supports. For a shape tensor,...
Definition: NvInfer.h:464
Class to handle tactic timing info collected from builder.
Definition: NvInfer.h:9013
int64_t queryKeys(TimingCacheKey *keyBuffer, int64_t capacity) const noexcept
Query cache keys from Timing Cache.
Definition: NvInfer.h:9079
bool combine(ITimingCache const &inputCache, bool ignoreMismatch) noexcept
Combine input timing cache into local instance.
Definition: NvInfer.h:9050
TimingCacheValue query(TimingCacheKey const &key) const noexcept
Query value in a cache entry.
Definition: NvInfer.h:9096
virtual ~ITimingCache() noexcept=default
bool update(TimingCacheKey const &key, TimingCacheValue const &value) noexcept
Update values in a cache entry.
Definition: NvInfer.h:9118
apiv::VTimingCache * mImpl
Definition: NvInfer.h:9124
bool reset() noexcept
Empty the timing cache.
Definition: NvInfer.h:9060
Layer that represents a TopK reduction.
Definition: NvInfer.h:3549
void setK(int32_t k) noexcept
Set the static k value for the layer.
Definition: NvInfer.h:3580
void setReduceAxes(uint32_t reduceAxes) noexcept
Set which axes to reduce for the layer.
Definition: NvInfer.h:3604
TopKOperation getOperation() const noexcept
Get the operation for the layer.
Definition: NvInfer.h:3566
apiv::VTopKLayer * mImpl
Definition: NvInfer.h:3636
void setOperation(TopKOperation op) noexcept
Set the operation for the layer.
Definition: NvInfer.h:3556
int32_t getK() const noexcept
Get the k value for the layer.
Definition: NvInfer.h:3594
uint32_t getReduceAxes() const noexcept
Get the axes to reduce for the layer.
Definition: NvInfer.h:3614
virtual ~ITopKLayer() noexcept=default
A layer that represents a trip-count limiter.
Definition: NvInfer.h:4774
TripLimit getTripLimit() const noexcept
Get a trip limiter type.
Definition: NvInfer.h:4779
virtual ~ITripLimitLayer() noexcept=default
Layer that represents an unary operation.
Definition: NvInfer.h:2829
void setOperation(UnaryOperation op) noexcept
Set the unary operation for the layer.
Definition: NvInfer.h:2838
apiv::VUnaryLayer * mImpl
Definition: NvInfer.h:2854
UnaryOperation getOperation() const noexcept
Get the unary operation for the layer.
Definition: NvInfer.h:2848
virtual ~IUnaryLayer() noexcept=default
Layer that represents an unsqueeze operation, which reshapes the input tensor by inserting unit-lengt...
Definition: NvInfer.h:6493
virtual ~IUnsqueezeLayer() noexcept=default
apiv::VUnsqueezeLayer * mImpl
Definition: NvInfer.h:6510
An Interface class for version control.
Definition: NvInferRuntimeBase.h:274
Version information associated with a TRT interface.
Definition: NvInferRuntimeBase.h:239
An array of weights used as a layer parameter.
Definition: NvInferRuntime.h:124
Definition: NvInfer.h:8644
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:8649
virtual ~IAlgorithmSelector() noexcept=default
Definition: NvInferRuntimeBase.h:411
Definition: NvInferRuntime.h:1610
Definition: NvInfer.h:8271
~IInt8EntropyCalibrator2() noexcept override=default
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:8284
InterfaceInfo getInterfaceInfo() const noexcept override
Return version information associated with this interface. Applications must not override this method...
Definition: NvInfer.h:8276
Definition: NvInfer.h:8231
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:8244
InterfaceInfo getInterfaceInfo() const noexcept override
Return version information associated with this interface. Applications must not override this method...
Definition: NvInfer.h:8236
~IInt8EntropyCalibrator() noexcept override=default
Definition: NvInfer.h:8350
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:8363
InterfaceInfo getInterfaceInfo() const noexcept override
Return version information associated with this interface. Applications must not override this method...
Definition: NvInfer.h:8355
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:8311
~IInt8MinMaxCalibrator() noexcept override=default
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
Definition: NvInferPluginBase.h:206
Definition: NvInfer.h:9339
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:10719
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:2904
ResizeSelector
The coordinate selector when resize to single pixel output.
Definition: NvInfer.h:4041
@ 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:9135
ScaleMode
Controls how shift, scale and power are applied in a Scale layer.
Definition: NvInfer.h:1835
@ 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:8755
uint32_t QuantizationFlags
Represents one or more QuantizationFlag values using binary OR operations.
Definition: NvInfer.h:8707
HardwareCompatibilityLevel
Describes requirements of compatibility with GPU architectures other than that of the GPU on which th...
Definition: NvInfer.h:9254
@ kSAME_COMPUTE_CAPABILITY
CumulativeOperation
Enumerates the cumulative operations that may be performed by a Cumulative layer.
Definition: NvInfer.h:6526
BoundingBoxFormat
Representation of bounding box data used for the Boxes input tensor in INMSLayer.
Definition: NvInfer.h:6111
@ 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:8957
constexpr int32_t EnumMax< LayerType >() noexcept
Definition: NvInfer.h:118
constexpr int32_t EnumMax< CalibrationAlgoType >() noexcept
Definition: NvInfer.h:8146
UnaryOperation
Enumerates the unary operations that may be performed by a Unary layer.
Definition: NvInfer.h:2782
@ 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:2891
constexpr int32_t EnumMax< TripLimit >() noexcept
Definition: NvInfer.h:4442
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:5022
@ 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:4071
@ 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:1013
@ 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:4430
@ 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:10363
PreviewFeature
Define preview features.
Definition: NvInfer.h:9210
@ kRUNTIME_ACTIVATION_RESIZE_10_10
@ kALIASED_PLUGIN_IO_10_03
TilingOptimizationLevel
Define the optimization levels for Tiling.
Definition: NvInfer.h:9306
@ 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:2541
DataType
The type of weights and tensors.
Definition: NvInferRuntimeBase.h:143
uint32_t BuilderFlags
Represents one or more BuilderFlag values using binary OR operations, e.g., 1U << BuilderFlag::kFP16 ...
Definition: NvInfer.h:8785
DeviceType
The device that this layer/network will execute on.
Definition: NvInferRuntime.h:1304
constexpr int32_t EnumMax< ScaleMode >() noexcept
Definition: NvInfer.h:1847
CalibrationAlgoType
Version of calibration algorithm to use.
Definition: NvInfer.h:8133
@ 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:8732
SampleMode
Controls how ISliceLayer and IGridSample handle out-of-bounds coordinates.
Definition: NvInfer.h:3240
@ 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:2529
@ 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:2916
NetworkDefinitionCreationFlag
List of immutable network properties expressed at network creation time. NetworkDefinitionCreationFla...
Definition: NvInfer.h:10374
@ kPREFER_JIT_PYTHON_PLUGINS
@ kPREFER_AOT_PYTHON_PLUGINS
ElementWiseOperation
Enumerates the binary operations that may be performed by an ElementWise layer.
Definition: NvInfer.h:2439
@ 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:8719
@ kCALIBRATE_BEFORE_FUSION
constexpr int32_t EnumMax< SampleMode >() noexcept
Definition: NvInfer.h:3256
InterpolationMode
Enumerates various modes of interpolation.
Definition: NvInfer.h:3959
@ 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:8795
@ 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
constexpr int32_t EnumMax< TopKOperation >() noexcept
Definition: NvInfer.h:3532
constexpr int32_t EnumMax< MemoryPoolType >() noexcept
Definition: NvInfer.h:9196
TopKOperation
Enumerates the operations that may be performed by a TopK layer.
Definition: NvInfer.h:3521
ReduceOperation
Enumerates the reduce operations that may be performed by a Reduce layer.
Definition: NvInfer.h:2877
constexpr int32_t EnumMax< LoopOutput >() noexcept
Definition: NvInfer.h:4419
constexpr int32_t EnumMax< NetworkDefinitionCreationFlag >() noexcept
Definition: NvInfer.h:10402
ScatterMode
Control form of IScatterLayer.
Definition: NvInfer.h:5845
MatrixOperation
Enumerates the operations that may be performed on a tensor by IMatrixMultiplyLayer before multiplica...
Definition: NvInfer.h:3647
@ kTRANSPOSE
Like kNONE, but transpose the matrix dimensions.
ResizeCoordinateTransformation
The resize coordinate transformation function.
Definition: NvInfer.h:3987
constexpr int32_t EnumMax< UnaryOperation >() noexcept
Definition: NvInfer.h:2816
LoopOutput
Enum that describes kinds of loop outputs.
Definition: NvInfer.h:4402
@ 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:6124
constexpr int32_t EnumMax< MatrixOperation >() noexcept
Definition: NvInfer.h:3675
PoolingType
The type of pooling to perform in a pooling layer.
Definition: NvInfer.h:1449
@ 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:9422
constexpr int32_t EnumMax< FillOperation >() noexcept
Definition: NvInfer.h:5053
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:5856
Represents a permutation of dimensions.
Definition: NvInfer.h:3049
Declaration of EnumMaxImpl struct to store maximum number of elements in an enumeration type.
Definition: NvInferRuntimeBase.h:128
The key to retrieve timing cache entries.
Definition: NvInfer.h:8977
Definition: NvInfer.h:8989
uint64_t tacticHash
Hash of the selected tactic.
Definition: NvInfer.h:8991
float timingMSec
Timing of this tactic in milliseconds. Negative numbers and NaN are invalid values.
Definition: NvInfer.h:8993