167 static constexpr int32_t kVALUE = 14;
205 mImpl->setName(name);
217 return mImpl->getName();
236 mImpl->setDimensions(dimensions);
250 return mImpl->getDimensions();
286 mImpl->setType(type);
301 return mImpl->getType();
318 return mImpl->setDynamicRange(min, max);
326 return mImpl->isNetworkInput();
334 return mImpl->isNetworkOutput();
351 mImpl->setBroadcastAcrossBatch(broadcastAcrossBatch);
365 return mImpl->getBroadcastAcrossBatch();
377 return mImpl->getLocation();
396 mImpl->setLocation(location);
408 return mImpl->dynamicRangeIsSet();
416 mImpl->resetDynamicRange();
426 return mImpl->getDynamicRangeMin();
436 return mImpl->getDynamicRangeMax();
458 mImpl->setAllowedFormats(formats);
471 return mImpl->getAllowedFormats();
502 return mImpl->isShapeTensor();
523 return mImpl->isExecutionTensor();
549 mImpl->setDimensionName(index, name);
564 return mImpl->getDimensionName(index);
589 return mLayer->getType();
603 mLayer->setName(name);
613 return mLayer->getName();
621 return mLayer->getNbInputs();
634 return mLayer->getInput(index);
642 return mLayer->getNbOutputs();
652 return mLayer->getOutput(index);
669 return mLayer->setInput(index, tensor);
702 mLayer->setPrecision(dataType);
714 return mLayer->getPrecision();
728 return mLayer->precisionIsSet();
740 mLayer->resetPrecision();
790 mLayer->setOutputType(index, dataType);
805 return mLayer->getOutputType(index);
821 return mLayer->outputTypeIsSet(index);
835 return mLayer->resetOutputType(index);
853 mLayer->setMetadata(metadata);
866 return mLayer->getMetadata();
871 apiv::VLayer* mLayer;
1048 static constexpr int32_t kVALUE = 4;
1076 mImpl->setNbOutputMaps(nbOutputMaps);
1086 return mImpl->getNbOutputMaps();
1106 mImpl->setNbGroups(nbGroups);
1116 return mImpl->getNbGroups();
1130 mImpl->setKernelWeights(weights);
1140 return mImpl->getKernelWeights();
1155 mImpl->setBiasWeights(weights);
1165 return mImpl->getBiasWeights();
1182 mImpl->setPrePadding(padding);
1192 return mImpl->getPrePadding();
1209 mImpl->setPostPadding(padding);
1219 return mImpl->getPostPadding();
1233 mImpl->setPaddingMode(paddingMode);
1245 return mImpl->getPaddingMode();
1258 mImpl->setKernelSizeNd(kernelSize);
1268 return mImpl->getKernelSizeNd();
1283 mImpl->setStrideNd(stride);
1293 return mImpl->getStrideNd();
1311 mImpl->setPaddingNd(padding);
1323 return mImpl->getPaddingNd();
1337 mImpl->setDilationNd(dilation);
1347 return mImpl->getDilationNd();
1396 mImpl->setActivationType(type);
1406 return mImpl->getActivationType();
1421 mImpl->setAlpha(alpha);
1435 mImpl->setBeta(beta);
1444 return mImpl->getAlpha();
1453 return mImpl->getBeta();
1483 static constexpr int32_t kVALUE = 3;
1510 mImpl->setPoolingType(type);
1520 return mImpl->getPoolingType();
1535 mImpl->setBlendFactor(blendFactor);
1548 return mImpl->getBlendFactor();
1562 mImpl->setAverageCountExcludesPadding(exclusive);
1573 return mImpl->getAverageCountExcludesPadding();
1591 mImpl->setPrePadding(padding);
1601 return mImpl->getPrePadding();
1619 mImpl->setPostPadding(padding);
1629 return mImpl->getPostPadding();
1642 mImpl->setPaddingMode(paddingMode);
1653 return mImpl->getPaddingMode();
1666 mImpl->setWindowSizeNd(windowSize);
1676 return mImpl->getWindowSizeNd();
1691 mImpl->setStrideNd(stride);
1701 return mImpl->getStrideNd();
1720 mImpl->setPaddingNd(padding);
1732 return mImpl->getPaddingNd();
1763 mImpl->setWindowSize(windowSize);
1773 return mImpl->getWindowSize();
1785 mImpl->setAlpha(alpha);
1795 return mImpl->getAlpha();
1807 mImpl->setBeta(beta);
1817 return mImpl->getBeta();
1839 return mImpl->getK();
1905 mImpl->setMode(mode);
1915 return mImpl->getMode();
1925 mImpl->setShift(shift);
1935 return mImpl->getShift();
1945 mImpl->setScale(scale);
1955 return mImpl->getScale();
1965 mImpl->setPower(power);
1975 return mImpl->getPower();
1990 return mImpl->getChannelAxis();
2011 mImpl->setChannelAxis(channelAxis);
2064 mImpl->setAxes(axes);
2074 return mImpl->getAxes();
2110 mImpl->setAxis(axis);
2120 return mImpl->getAxis();
2147 mImpl->setNbOutputMaps(nbOutputMaps);
2157 return mImpl->getNbOutputMaps();
2177 mImpl->setNbGroups(nbGroups);
2187 return mImpl->getNbGroups();
2201 mImpl->setKernelWeights(weights);
2211 return mImpl->getKernelWeights();
2226 mImpl->setBiasWeights(weights);
2236 return mImpl->getBiasWeights();
2253 mImpl->setPrePadding(padding);
2263 return mImpl->getPrePadding();
2280 mImpl->setPostPadding(padding);
2290 return mImpl->getPostPadding();
2304 mImpl->setPaddingMode(paddingMode);
2316 return mImpl->getPaddingMode();
2331 mImpl->setKernelSizeNd(kernelSize);
2341 return mImpl->getKernelSizeNd();
2358 mImpl->setStrideNd(stride);
2368 return mImpl->getStrideNd();
2386 mImpl->setPaddingNd(padding);
2398 return mImpl->getPaddingNd();
2424 mImpl->setDilationNd(dilation);
2434 return mImpl->getDilationNd();
2482 static constexpr int32_t kVALUE = 14;
2519 return mImpl->setOperation(op);
2531 return mImpl->getOperation();
2652 mImpl->setGatherAxis(axis);
2664 return mImpl->getGatherAxis();
2687 mImpl->setNbElementWiseDims(elementWiseDims);
2697 return mImpl->getNbElementWiseDims();
2707 mImpl->setMode(mode);
2717 return mImpl->getMode();
2746 return mImpl->getPlugin();
2773 return mImpl->getPlugin();
2856 mImpl->setOperation(op);
2866 return mImpl->getOperation();
2929 mImpl->setOperation(op);
2939 return mImpl->getOperation();
2949 mImpl->setReduceAxes(reduceAxes);
2959 return mImpl->getReduceAxes();
2969 mImpl->setKeepDimensions(keepDimensions);
2979 return mImpl->getKeepDimensions();
3013 mImpl->setPrePaddingNd(padding);
3025 return mImpl->getPrePaddingNd();
3039 mImpl->setPostPaddingNd(padding);
3051 return mImpl->getPostPaddingNd();
3101 mImpl->setFirstTranspose(permutation);
3113 return mImpl->getFirstTranspose();
3141 mImpl->setReshapeDimensions(dimensions);
3154 return mImpl->getReshapeDimensions();
3201 mImpl->setSecondTranspose(permutation);
3213 return mImpl->getSecondTranspose();
3229 return mImpl->setZeroIsPlaceholder(zeroIsPlaceholder);
3242 return mImpl->getZeroIsPlaceholder();
3353 mImpl->setStart(start);
3368 return mImpl->getStart();
3382 return mImpl->setSize(size);
3397 return mImpl->getSize();
3411 mImpl->setStride(stride);
3426 return mImpl->getStride();
3436 mImpl->setMode(mode);
3446 return mImpl->getMode();
3489 mImpl->setAxes(axes);
3504 return mImpl->getAxes();
3574 mImpl->setOperation(op);
3584 return mImpl->getOperation();
3612 return mImpl->getK();
3622 mImpl->setReduceAxes(reduceAxes);
3632 return mImpl->getReduceAxes();
3663 return mImpl->setIndicesType(type);
3675 return mImpl->getIndicesType();
3761 mImpl->setOperation(index, op);
3773 return mImpl->getOperation(index);
3817 return mImpl->setIndicesType(type);
3829 return mImpl->getIndicesType();
3926 mImpl->setToType(toType);
3937 return mImpl->getToType();
3966 mImpl->setWeights(weights);
3976 return mImpl->getWeights();
3988 mImpl->setDimensions(dimensions);
4000 return mImpl->getDimensions();
4046 static constexpr int32_t kVALUE = 3;
4100 static constexpr int32_t kVALUE = 3;
4130 static constexpr int32_t kVALUE = 2;
4166 static constexpr int32_t kVALUE = 4;
4229 return mImpl->setOutputDimensions(dimensions);
4239 return mImpl->getOutputDimensions();
4267 void setScales(
float const* scales, int32_t nbScales)
noexcept
4269 mImpl->setScales(scales, nbScales);
4286 int32_t
getScales(int32_t size,
float* scales)
const noexcept
4288 return mImpl->getScales(size, scales);
4300 mImpl->setResizeMode(interpolationMode);
4310 return mImpl->getResizeMode();
4345 mImpl->setCoordinateTransformation(coordTransform);
4355 return mImpl->getCoordinateTransformation();
4370 mImpl->setSelectorForSinglePixel(selector);
4380 return mImpl->getSelectorForSinglePixel();
4394 mImpl->setNearestRounding(value);
4404 return mImpl->getNearestRounding();
4426 mImpl->setCubicCoeff(A);
4436 return mImpl->getCubicCoeff();
4449 mImpl->setExcludeOutside(excludeFlag);
4459 return mImpl->getExcludeOutside();
4541 return mBoundary->getLoop();
4546 apiv::VLoopBoundaryLayer* mBoundary;
4564 return mBoundary->getConditional();
4569 apiv::VConditionalBoundaryLayer* mBoundary;
4653 return mImpl->setCondition(condition);
4671 return mImpl->addOutput(trueSubgraphOutput, falseSubgraphOutput);
4683 return mImpl->addInput(input);
4698 mImpl->setName(name);
4708 return mImpl->getName();
4778 return mImpl->getLoopOutput();
4795 mImpl->setAxis(axis);
4803 return mImpl->getAxis();
4852 return mImpl->getTripLimit();
4878 mImpl->setAxis(axis);
4886 return mImpl->getAxis();
4900 mImpl->setReverse(reverse);
4910 return mImpl->getReverse();
4939 return mImpl->addRecurrence(initialValue);
4960 return mImpl->addTripLimit(tensor, limit);
4973 return mImpl->addIterator(tensor, axis, reverse);
4986 return mImpl->addLoopOutput(tensor, outputKind, axis);
5001 mImpl->setName(name);
5011 return mImpl->getName();
5066 mImpl->setMessage(message);
5076 return mImpl->getMessage();
5178 mImpl->setDimensions(dimensions);
5193 return mImpl->getDimensions();
5203 mImpl->setOperation(op);
5213 return mImpl->getOperation();
5232 mImpl->setAlpha(alpha);
5247 return mImpl->getAlpha();
5266 mImpl->setBeta(beta);
5281 return mImpl->getBeta();
5342 mImpl->setAlphaInt64(alpha);
5357 return mImpl->getAlphaInt64();
5376 mImpl->setBetaInt64(beta);
5391 return mImpl->getBetaInt64();
5399 return mImpl->isAlphaBetaInt64();
5416 mImpl->setToType(toType);
5428 return mImpl->getToType();
5523 return mImpl->getAxis();
5534 mImpl->setAxis(axis);
5550 mImpl->setToType(toType);
5562 return mImpl->getToType();
5651 return mImpl->getAxis();
5662 mImpl->setAxis(axis);
5678 mImpl->setToType(toType);
5690 return mImpl->getToType();
5745 mImpl->setToType(toType);
5758 return mImpl->getToType();
5771 mImpl->setScaleType(scaleType);
5784 return mImpl->getScaleType();
5797 mImpl->setAxis(axis);
5807 return mImpl->getAxis();
5820 mImpl->setBlockSize(size);
5830 return mImpl->getBlockSize();
5886 return mImpl->setEquation(equation);
5896 return mImpl->getEquation();
5995 mImpl->setMode(mode);
6005 return mImpl->getMode();
6015 mImpl->setAxis(axis);
6023 return mImpl->getAxis();
6067 mImpl->setAxis(axis);
6075 return mImpl->getAxis();
6104 mImpl->setInterpolationMode(mode);
6116 return mImpl->getInterpolationMode();
6126 mImpl->setAlignCorners(alignCorners);
6138 return mImpl->getAlignCorners();
6150 return mImpl->setSampleMode(mode);
6162 return mImpl->getSampleMode();
6260 mImpl->setBoundingBoxFormat(fmt);
6272 return mImpl->getBoundingBoxFormat();
6286 mImpl->setTopKBoxLimit(limit);
6296 return mImpl->getTopKBoxLimit();
6331 return mImpl->setIndicesType(type);
6343 return mImpl->getIndicesType();
6376 mImpl->setBatchAxis(batchAxis);
6386 return mImpl->getBatchAxis();
6399 mImpl->setSequenceAxis(sequenceAxis);
6409 return mImpl->getSequenceAxis();
6447 return mImpl->setEpsilon(eps);
6457 return mImpl->getEpsilon();
6467 return mImpl->setAxes(axesMask);
6477 return mImpl->getAxes();
6498 return mImpl->setNbGroups(nbGroups);
6508 return mImpl->getNbGroups();
6534 return mImpl->setComputePrecision(type);
6544 return mImpl->getComputePrecision();
6639 static constexpr int32_t kVALUE = 1;
6685 return mImpl->setOperation(op);
6697 return mImpl->getOperation();
6709 mImpl->setExclusive(exclusive);
6721 return mImpl->getExclusive();
6733 mImpl->setReverse(reverse);
6745 return mImpl->getReverse();
6775 static constexpr int32_t kVALUE = 2;
6797 return mBoundary->getAttention();
6802 apiv::VAttentionBoundaryLayer* mBoundary;
6915 return mImpl->setNormalizationOperation(op);
6927 return mImpl->getNormalizationOperation();
6944 return mImpl->setMask(mask);
6956 return mImpl->getMask();
6969 return mImpl->setCausal(isCausal);
6981 return mImpl->getCausal();
6993 return mImpl->setDecomposable(decomposable);
7006 return mImpl->getDecomposable();
7025 return mImpl->setInput(index, input);
7034 return mImpl->getNbInputs();
7046 return mImpl->getInput(index);
7054 return mImpl->getNbOutputs();
7066 return mImpl->getOutput(index);
7083 return mImpl->setName(name);
7095 return mImpl->getName();
7111 return mImpl->setNormalizationQuantizeScale(tensor);
7122 return mImpl->getNormalizationQuantizeScale();
7135 return mImpl->setNormalizationQuantizeToType(type);
7147 return mImpl->getNormalizationQuantizeToType();
7215 return mImpl->addInput(name, type, dimensions);
7229 mImpl->markOutput(tensor);
7247 return mImpl->markDebug(tensor);
7263 return mImpl->unmarkDebug(tensor);
7273 return mImpl->isDebugTensor(tensor);
7295 return mImpl->markUnfusedTensorsAsDebugTensors();
7309 return mImpl->unmarkUnfusedTensorsAsDebugTensors();
7329 return mImpl->addActivation(input, type);
7348 return mImpl->addLRN(input, window, alpha, beta, k);
7374 return mImpl->addScale(input, mode, shift, scale, power);
7387 return mImpl->addSoftMax(input);
7404 return mImpl->addConcatenation(inputs, nbInputs);
7431 return mImpl->addElementWise(input1, input2, op);
7453 return mImpl->addUnary(input, operation);
7467 return mImpl->addShuffle(input);
7484 return mImpl->addOneHot(indices, values, depth, axis);
7496 return mImpl->getNbLayers();
7510 return mImpl->getLayer(index);
7522 return mImpl->getNbInputs();
7538 return mImpl->getInput(index);
7552 return mImpl->getNbOutputs();
7568 return mImpl->getOutput(index);
7595 return mImpl->addReduce(input, operation, reduceAxes, keepDimensions);
7631 return mImpl->addTopK(input, op, k, reduceAxes);
7665 return mImpl->addTopKV2(input, op, k, reduceAxes, indicesType);
7681 return mImpl->addGather(data, indices, axis);
7697 return mImpl->addGatherV2(data, indices, mode);
7716 return mImpl->addRaggedSoftMax(input, bounds);
7738 return mImpl->addMatrixMultiply(input0, op0, input1, op1);
7756 return mImpl->addNonZero(input);
7772 return mImpl->addNonZeroV2(input, indicesType);
7796 return mImpl->addConstant(dimensions, weights);
7810 return mImpl->addIdentity(input);
7825 return mImpl->addCast(input, toType);
7840 mImpl->removeTensor(tensor);
7852 mImpl->unmarkOutput(tensor);
7873 return mImpl->addPluginV2(inputs, nbInputs, plugin);
7890 int32_t nbShapeInputs,
IPluginV3& plugin)
noexcept
7892 return mImpl->addPluginV3(inputs, nbInputs, shapeInputs, nbShapeInputs, plugin);
7911 return mImpl->addSlice(input, start, size, stride);
7935 mImpl->setName(name);
7949 return mImpl->getName();
7965 return mImpl->addShape(input);
7979 return mImpl->hasImplicitBatchDimension();
7989 return mImpl->getFlags();
8001 return mImpl->getFlag(networkDefinitionCreationFlag);
8018 return mImpl->markOutputForShapes(tensor);
8030 return mImpl->unmarkOutputForShapes(tensor);
8048 return mImpl->addParametricReLU(input, slope);
8071 return mImpl->addConvolutionNd(input, nbOutputMaps, kernelSize, kernelWeights, biasWeights);
8090 return mImpl->addPoolingNd(input, type, windowSize);
8113 return mImpl->addDeconvolutionNd(input, nbOutputMaps, kernelSize, kernelWeights, biasWeights);
8150 return mImpl->addScaleNd(input, mode, shift, scale, power, channelAxis);
8166 return mImpl->addResize(input);
8180 return mImpl->addLoop();
8195 return mImpl->addIfConditional();
8234 return mImpl->addSelect(condition, thenInput, elseInput);
8251 return mImpl->addAssertion(condition, message);
8276 return mImpl->addFill(dimensions, op);
8302 return mImpl->addFillV2(dimensions, op, outputType);
8318 return mImpl->addPaddingNd(input, prePadding, postPadding);
8342 return mImpl->setWeightsName(weights, name);
8361 mImpl->setErrorRecorder(recorder);
8376 return mImpl->getErrorRecorder();
8397 return mImpl->addDequantize(input, scale);
8420 return mImpl->addDequantizeV2(input, scale, outputType);
8440 return mImpl->addScatter(data, indices, updates, mode);
8461 return mImpl->addQuantize(input, scale);
8485 return mImpl->addQuantizeV2(input, scale, outputType);
8513 return mImpl->addDynamicQuantize(input, axis, blockSize, outputType, scaleType);
8528 return mImpl->addEinsum(inputs, nbInputs, equation);
8546 return mImpl->addGridSample(input, grid);
8568 return mImpl->addNMS(boxes, scores, maxOutputBoxesPerClass);
8588 return mImpl->addNMSV2(boxes, scores, maxOutputBoxesPerClass, indicesType);
8605 return mImpl->addReverseSequence(input, sequenceLens);
8631 return mImpl->addNormalization(input, scale, bias, axesMask);
8653 return mImpl->addCumulative(input, axis, operation, exclusive, reverse);
8681 return mImpl->addAttention(query, key, value, normOp, causal);
8692 return mImpl->getBuilder();
8705 return mImpl->markWeightsRefittable(name);
8717 return mImpl->unmarkWeightsRefittable(name);
8730 return mImpl->areWeightsMarkedRefittable(name);
8749 return mImpl->addSqueeze(input, axes);
8770 return mImpl->addUnsqueeze(input, axes);
8842 virtual
bool getBatch(
void* bindings[],
char const* names[], int32_t nbBindings) noexcept = 0;
8858 virtual
void const* readCalibrationCache(std::
size_t& length) noexcept = 0;
8868 virtual
void writeCalibrationCache(
void const* ptr, std::
size_t length) noexcept = 0;
9034 virtual
double getRegressionCutoff() const noexcept = 0;
9048 virtual
void const* readHistogramCache(std::
size_t& length) noexcept = 0;
9058 virtual
void writeHistogramCache(
void const* ptr, std::
size_t length) noexcept = 0;
9101 return mImpl->getDataType();
9112 return mImpl->getStrides();
9122 return mImpl->getVectorizedDim();
9133 return mImpl->getComponentsPerElement();
9162 return mImpl->getImplementation();
9170 return mImpl->getTactic();
9198 return mImpl->getName();
9210 return mImpl->getDimensions(index, select);
9218 return mImpl->getNbInputs();
9226 return mImpl->getNbOutputs();
9255 return mImpl->getAlgorithmVariant();
9263 return mImpl->getTimingMSec();
9271 return mImpl->getWorkspaceSize();
9285 return mImpl->getAlgorithmIOInfoByIndex(index);
9320 int32_t nbChoices, int32_t* selection)
noexcept = 0;
9333 int32_t nbAlgorithms)
noexcept = 0;
9429 static constexpr int32_t kVALUE = 2;
9624#if ENABLE_FEATURE_DISABLE_RUNTIME_ALLOCATION
9631 kREQUIRE_USER_ALLOCATION = 29,
9644#if ENABLE_FEATURE_DISABLE_RUNTIME_ALLOCATION
9684 static constexpr uint64_t kINVALID_TACTIC_HASH = UINT64_MAX;
9717 return mImpl->serialize();
9741 return mImpl->combine(inputCache, ignoreMismatch);
9751 return mImpl->reset();
9768 int64_t
queryKeys(TimingCacheKey* keyBuffer, int64_t capacity)
const noexcept
9770 return mImpl->queryKeys(keyBuffer, capacity);
9785 TimingCacheValue
query(TimingCacheKey
const& key)
const noexcept
9787 return mImpl->query(key);
9807 bool update(TimingCacheKey
const& key, TimingCacheValue
const& value)
noexcept
9809 return mImpl->update(key, value);
9930 static constexpr int32_t kVALUE = 3;
9982 static constexpr int32_t kVALUE = 3;
10022 static constexpr int32_t kVALUE = 4;
10061 virtual void phaseStart(
char const* phaseName,
char const* parentPhase, int32_t nbSteps)
noexcept = 0;
10134 virtual
void setAvgTimingIterations(int32_t avgTiming) noexcept
10136 mImpl->setAvgTimingIterations(avgTiming);
10148 return mImpl->getAvgTimingIterations();
10161 mImpl->setEngineCapability(capability);
10173 return mImpl->getEngineCapability();
10185 mImpl->setInt8Calibrator(calibrator);
10195 return mImpl->getInt8Calibrator();
10212 mImpl->setFlags(builderFlags);
10224 return mImpl->getFlags();
10236 mImpl->clearFlag(builderFlag);
10248 mImpl->setFlag(builderFlag);
10260 return mImpl->getFlag(builderFlag);
10277 mImpl->setDeviceType(layer, deviceType);
10287 return mImpl->getDeviceType(layer);
10299 return mImpl->isDeviceTypeSet(layer);
10309 mImpl->resetDeviceType(layer);
10319 return mImpl->canRunOnDLA(layer);
10335 mImpl->setDLACore(dlaCore);
10345 return mImpl->getDLACore();
10356 mImpl->setDefaultDeviceType(deviceType);
10366 return mImpl->getDefaultDeviceType();
10388 return mImpl->setProfileStream(stream);
10400 return mImpl->getProfileStream();
10417 return mImpl->addOptimizationProfile(profile);
10430 return mImpl->getNbOptimizationProfiles();
10442 mImpl->setProfilingVerbosity(verbosity);
10455 return mImpl->getProfilingVerbosity();
10467 mImpl->setAlgorithmSelector(selector);
10477 return mImpl->getAlgorithmSelector();
10495 return mImpl->setCalibrationProfile(profile);
10507 return mImpl->getCalibrationProfile();
10526 mImpl->setQuantizationFlags(flags);
10540 return mImpl->getQuantizationFlags();
10554 mImpl->clearQuantizationFlag(flag);
10568 mImpl->setQuantizationFlag(flag);
10582 return mImpl->getQuantizationFlag(flag);
10604 return mImpl->setTacticSources(tacticSources);
10619 return mImpl->getTacticSources();
10639 return mImpl->createTimingCache(blob, size);
10662 return mImpl->setTimingCache(cache, ignoreMismatch);
10672 return mImpl->getTimingCache();
10704 mImpl->setMemoryPoolLimit(pool, poolSize);
10723 return mImpl->getMemoryPoolLimit(pool);
10741 mImpl->setPreviewFeature(feature, enable);
10755 return mImpl->getPreviewFeature(feature);
10788 mImpl->setBuilderOptimizationLevel(level);
10800 return mImpl->getBuilderOptimizationLevel();
10817 mImpl->setHardwareCompatibilityLevel(hardwareCompatibilityLevel);
10830 return mImpl->getHardwareCompatibilityLevel();
10843 mImpl->setPluginsToSerialize(paths, nbPaths);
10856 return mImpl->getPluginToSerialize(index);
10866 return mImpl->getNbPluginsToSerialize();
10895 mImpl->setMaxAuxStreams(nbStreams);
10905 return mImpl->getMaxAuxStreams();
10921 return mImpl->setProgressMonitor(monitor);
10931 return mImpl->getProgressMonitor();
10947 mImpl->setRuntimePlatform(runtimePlatform);
10959 return mImpl->getRuntimePlatform();
10971 mImpl->setMaxNbTactics(maxNbTactics);
10983 return mImpl->getMaxNbTactics();
10999 return mImpl->setTilingOptimizationLevel(level);
11011 return mImpl->getTilingOptimizationLevel();
11027 return mImpl->setL2LimitForTiling(size);
11039 return mImpl->getL2LimitForTiling();
11053 return mImpl->setRemoteAutoTuningConfig(config);
11063 return mImpl->getRemoteAutoTuningConfig();
11140 return mImpl->platformHasFastFp16();
11150 return mImpl->platformHasFastInt8();
11162 return mImpl->getMaxDLABatchSize();
11170 return mImpl->getNbDLACores();
11188 mImpl->setGpuAllocator(allocator);
11202 return mImpl->createBuilderConfig();
11228 return mImpl->createNetworkV2(flags);
11243 return mImpl->createOptimizationProfile();
11262 mImpl->setErrorRecorder(recorder);
11277 return mImpl->getErrorRecorder();
11295 return mImpl->platformHasTf32();
11314 return mImpl->buildSerializedNetwork(network, config);
11336 return mImpl->buildSerializedNetworkToStream(network, config, writer);
11360 return mImpl->buildSerializedNetworkWithKernelText(network, config, kernelText);
11380 return mImpl->buildEngineWithConfig(network, config);
11402 return mImpl->isNetworkSupported(network, config);
11412 return mImpl->getLogger();
11428 return mImpl->setMaxThreads(maxThreads);
11442 return mImpl->getMaxThreads();
11452 return mImpl->getPluginRegistry();
11465extern "C" TENSORRTAPI void* createInferBuilder_INTERNAL(
void* logger, int32_t version)
noexcept;
#define TRT_DEPRECATED_API
Definition: NvInferRuntimeBase.h:44
#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:218
static constexpr int32_t MAX_DIMS
The maximum rank (number of dimensions) supported for a tensor.
Definition: NvInferRuntimeBase.h:221
An Activation layer in a network definition.
Definition: NvInfer.h:1385
void setBeta(float beta) noexcept
Set the beta parameter (must be finite).
Definition: NvInfer.h:1433
void setActivationType(ActivationType type) noexcept
Set the type of activation to be performed.
Definition: NvInfer.h:1394
ActivationType getActivationType() const noexcept
Get the type of activation to be performed.
Definition: NvInfer.h:1404
float getAlpha() const noexcept
Get the alpha parameter.
Definition: NvInfer.h:1442
virtual ~IActivationLayer() noexcept=default
float getBeta() const noexcept
Get the beta parameter.
Definition: NvInfer.h:1451
void setAlpha(float alpha) noexcept
Set the alpha parameter (must be finite).
Definition: NvInfer.h:1419
Describes the context and requirements, that could be fulfilled by one or more instances of IAlgorith...
Definition: NvInfer.h:9189
int32_t getNbOutputs() const noexcept
Return number of outputs of the algorithm.
Definition: NvInfer.h:9224
int32_t getNbInputs() const noexcept
Return number of inputs of the algorithm.
Definition: NvInfer.h:9216
char const * getName() const noexcept
Return name of the algorithm node.
Definition: NvInfer.h:9196
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:9208
Describes a variation of execution of a layer. An algorithm is represented by IAlgorithmVariant and t...
Definition: NvInfer.h:9248
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:9269
float getTimingMSec() const noexcept
The time in milliseconds to execute the algorithm.
Definition: NvInfer.h:9261
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:9283
virtual ~IAlgorithm() noexcept=default
IAlgorithmVariant const & getAlgorithmVariant() const noexcept
Returns the algorithm variant.
Definition: NvInfer.h:9253
Carries information about input or output of the algorithm. IAlgorithmIOInfo for all the input and ou...
Definition: NvInfer.h:9092
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:9120
Dims getStrides() const noexcept
Return strides of the input/output tensor of algorithm. For vectorized formats, strides are given in ...
Definition: NvInfer.h:9110
DataType getDataType() const noexcept
Return DataType of the input/output of algorithm.
Definition: NvInfer.h:9099
int64_t getComponentsPerElement() const noexcept
Return the number of components per element. This is always 1 for non-vectorized formats.
Definition: NvInfer.h:9131
provides a unique 128-bit identifier, which along with the input and output information denotes the v...
Definition: NvInfer.h:9155
virtual ~IAlgorithmVariant() noexcept=default
int64_t getTactic() const noexcept
Return tactic of the algorithm.
Definition: NvInfer.h:9168
int64_t getImplementation() const noexcept
Return implementation of the algorithm.
Definition: NvInfer.h:9160
An assertion layer in a network.
Definition: NvInfer.h:5054
void setMessage(char const *message) noexcept
Set the message to print if the assertion fails.
Definition: NvInfer.h:5064
char const * getMessage() const noexcept
Return the assertion message.
Definition: NvInfer.h:5074
virtual ~IAssertionLayer() noexcept=default
This is a base class for Attention boundary layers.
Definition: NvInfer.h:6790
IAttention * getAttention() const noexcept
Get a pointer to the IAttention associated with this boundary layer.
Definition: NvInfer.h:6795
virtual ~IAttentionBoundaryLayer() noexcept=default
Helper for constructing an attention that consumes query, key and value tensors.
Definition: NvInfer.h:6904
ITensor * getMask() noexcept
Get the optional mask in attention.
Definition: NvInfer.h:6954
bool setDecomposable(bool decomposable) noexcept
Set whether the attention can be decomposed to use multiple kernels if no fused kernel support found.
Definition: NvInfer.h:6991
bool setName(char const *name) noexcept
Set the name of the attention.
Definition: NvInfer.h:7081
bool getDecomposable() const noexcept
Get whether the attention can be decomposed to use multiple kernels if no fused kernel support found.
Definition: NvInfer.h:7004
ITensor * getInput(int32_t index) const noexcept
Get the IAttention input corresponding to the given index.
Definition: NvInfer.h:7044
ITensor * getOutput(int32_t index) const noexcept
Get the IAttention output corresponding to the given index. IAttention has only one output.
Definition: NvInfer.h:7064
int32_t getNbOutputs() const noexcept
Get the number of outputs of a layer. IAttention has one output.
Definition: NvInfer.h:7052
int32_t getNbInputs() const noexcept
Get the number of inputs of IAttention. IAttention has three inputs.
Definition: NvInfer.h:7032
bool setCausal(bool isCausal) noexcept
Set whether the attention will run a causal inference. Cannot be used together with setMask().
Definition: NvInfer.h:6967
bool setNormalizationOperation(AttentionNormalizationOp op) noexcept
Set the normalization operation for the attention.
Definition: NvInfer.h:6913
char const * getName() const noexcept
Return the name of the attention.
Definition: NvInfer.h:7093
bool setNormalizationQuantizeToType(DataType type) noexcept
Set the datatype the attention normalization is quantized to.
Definition: NvInfer.h:7133
AttentionNormalizationOp getNormalizationOperation() const noexcept
Get the normalization operation for the attention.
Definition: NvInfer.h:6925
bool setNormalizationQuantizeScale(ITensor &tensor) noexcept
Set the quantization scale for the attention normalization output.
Definition: NvInfer.h:7109
DataType getNormalizationQuantizeToType() const noexcept
Get the datatype the attention normalization is quantized to.
Definition: NvInfer.h:7145
ITensor * getNormalizationQuantizeScale() const noexcept
Get the quantization scale for the attention normalization output.
Definition: NvInfer.h:7120
bool setInput(int32_t index, ITensor &input) noexcept
Append or replace an input of this layer with a specific tensor.
Definition: NvInfer.h:7023
bool setMask(ITensor &mask) noexcept
Set whether a mask will be used for the normalization operation.
Definition: NvInfer.h:6942
bool getCausal() const noexcept
Get whether the attention will run a causal inference.
Definition: NvInfer.h:6979
apiv::VAttention * mImpl
Definition: NvInfer.h:7152
virtual ~IAttention() noexcept=default
This layer represents an output of an IAttention.
Definition: NvInfer.h:6851
virtual ~IAttentionOutputLayer() noexcept=default
Holds properties for configuring a builder to produce an engine.
Definition: NvInfer.h:10122
void setMemoryPoolLimit(MemoryPoolType pool, std::size_t poolSize) noexcept
Set the memory size for the memory pool.
Definition: NvInfer.h:10702
TRT_DEPRECATED void setQuantizationFlags(QuantizationFlags flags) noexcept
Set the quantization flags.
Definition: NvInfer.h:10524
nvinfer1::ITimingCache * createTimingCache(void const *blob, std::size_t size) const noexcept
Create timing cache.
Definition: NvInfer.h:10637
void setPreviewFeature(PreviewFeature feature, bool enable) noexcept
Enable or disable a specific preview feature.
Definition: NvInfer.h:10739
TRT_DEPRECATED void setAlgorithmSelector(IAlgorithmSelector *selector) noexcept
Set Algorithm Selector.
Definition: NvInfer.h:10465
TRT_DEPRECATED void setInt8Calibrator(IInt8Calibrator *calibrator) noexcept
Set Int8 Calibration interface.
Definition: NvInfer.h:10183
bool getPreviewFeature(PreviewFeature feature) const noexcept
Get status of preview feature.
Definition: NvInfer.h:10753
int32_t getBuilderOptimizationLevel() noexcept
Get builder optimization level.
Definition: NvInfer.h:10798
bool setTacticSources(TacticSources tacticSources) noexcept
Set tactic sources.
Definition: NvInfer.h:10602
void setPluginsToSerialize(char const *const *paths, int32_t nbPaths) noexcept
Set the plugin libraries to be serialized with version-compatible engines.
Definition: NvInfer.h:10841
bool setTilingOptimizationLevel(TilingOptimizationLevel level) noexcept
Set the Tiling optimization level.
Definition: NvInfer.h:10997
bool setL2LimitForTiling(int64_t size) noexcept
Set the L2 cache usage limit for Tiling optimization.
Definition: NvInfer.h:11025
TRT_DEPRECATED IInt8Calibrator * getInt8Calibrator() const noexcept
Get Int8 Calibration interface.
Definition: NvInfer.h:10193
std::size_t getMemoryPoolLimit(MemoryPoolType pool) const noexcept
Get the memory size limit of the memory pool.
Definition: NvInfer.h:10721
int32_t getDLACore() const noexcept
Get the DLA core that the engine executes on.
Definition: NvInfer.h:10343
int32_t getNbPluginsToSerialize() const noexcept
Get the number of plugin library paths to be serialized with version-compatible engines.
Definition: NvInfer.h:10864
void setDeviceType(ILayer const *layer, DeviceType deviceType) noexcept
Set the device that this layer must execute on.
Definition: NvInfer.h:10275
void setEngineCapability(EngineCapability capability) noexcept
Configure the builder to target specified EngineCapability flow.
Definition: NvInfer.h:10159
int32_t getMaxAuxStreams() const noexcept
Get the maximum number of auxiliary streams that TRT is allowed to use.
Definition: NvInfer.h:10903
bool getFlag(BuilderFlag builderFlag) const noexcept
Returns true if the build mode flag is set.
Definition: NvInfer.h:10258
void setMaxNbTactics(int32_t maxNbTactics) noexcept
Set the maximum number of tactics to time when there is a choice of tactics.
Definition: NvInfer.h:10969
TRT_DEPRECATED void clearQuantizationFlag(QuantizationFlag flag) noexcept
clear a quantization flag.
Definition: NvInfer.h:10552
int64_t getL2LimitForTiling() const noexcept
Get the L2 cache usage limit for tiling optimization.
Definition: NvInfer.h:11037
bool setRemoteAutoTuningConfig(char const *config) noexcept
Set a config string for remote auto tuning.
Definition: NvInfer.h:11051
void setProgressMonitor(IProgressMonitor *monitor) noexcept
Sets the progress monitor for building a network.
Definition: NvInfer.h:10919
void setProfilingVerbosity(ProfilingVerbosity verbosity) noexcept
Set verbosity level of layer information exposed in NVTX annotations and IEngineInspector.
Definition: NvInfer.h:10440
int32_t getNbOptimizationProfiles() const noexcept
Get number of optimization profiles.
Definition: NvInfer.h:10428
nvinfer1::ITimingCache const * getTimingCache() const noexcept
Get the pointer to the timing cache from current IBuilderConfig.
Definition: NvInfer.h:10670
void reset() noexcept
Resets the builder configuration to defaults.
Definition: NvInfer.h:10374
bool setTimingCache(ITimingCache const &cache, bool ignoreMismatch) noexcept
Attach a timing cache to IBuilderConfig.
Definition: NvInfer.h:10660
char const * getPluginToSerialize(int32_t index) const noexcept
Get the plugin library path to be serialized with version-compatible engines.
Definition: NvInfer.h:10854
EngineCapability getEngineCapability() const noexcept
Query EngineCapability flow configured for the builder.
Definition: NvInfer.h:10171
RuntimePlatform getRuntimePlatform() const noexcept
Get the target platform for runtime execution.
Definition: NvInfer.h:10957
DeviceType getDefaultDeviceType() const noexcept
Get the default DeviceType which was set by setDefaultDeviceType.
Definition: NvInfer.h:10364
void setRuntimePlatform(RuntimePlatform runtimePlatform) noexcept
Set the target platform for runtime execution.
Definition: NvInfer.h:10945
int32_t getMaxNbTactics() const noexcept
Query the maximum number of tactics timed when there is a choice.
Definition: NvInfer.h:10981
BuilderFlags getFlags() const noexcept
Get the build mode flags for this builder config. Defaults to 0.
Definition: NvInfer.h:10222
void setFlags(BuilderFlags builderFlags) noexcept
Set the build mode flags to turn on builder options for this network.
Definition: NvInfer.h:10210
TacticSources getTacticSources() const noexcept
Get tactic sources.
Definition: NvInfer.h:10617
void resetDeviceType(ILayer const *layer) noexcept
reset the DeviceType for this layer
Definition: NvInfer.h:10307
void setDLACore(int32_t dlaCore) noexcept
Sets the DLA core used by the network. Defaults to -1.
Definition: NvInfer.h:10333
HardwareCompatibilityLevel getHardwareCompatibilityLevel() const noexcept
Get the hardware compatibility level.
Definition: NvInfer.h:10828
char const * getRemoteAutoTuningConfig() const noexcept
Get a config string for remote auto tuning.
Definition: NvInfer.h:11061
TRT_DEPRECATED QuantizationFlags getQuantizationFlags() const noexcept
Get the quantization flags.
Definition: NvInfer.h:10538
void clearFlag(BuilderFlag builderFlag) noexcept
clear a single build mode flag.
Definition: NvInfer.h:10234
int32_t addOptimizationProfile(IOptimizationProfile const *profile) noexcept
Add an optimization profile.
Definition: NvInfer.h:10415
IProgressMonitor * getProgressMonitor() const noexcept
Definition: NvInfer.h:10929
apiv::VBuilderConfig * mImpl
Definition: NvInfer.h:11067
TRT_DEPRECATED IOptimizationProfile const * getCalibrationProfile() noexcept
Get the current calibration profile.
Definition: NvInfer.h:10505
int32_t getAvgTimingIterations() const noexcept
Query the number of averaging iterations.
Definition: NvInfer.h:10146
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:10354
void setFlag(BuilderFlag builderFlag) noexcept
Set a single build mode flag.
Definition: NvInfer.h:10246
TRT_DEPRECATED bool setCalibrationProfile(IOptimizationProfile const *profile) noexcept
Add a calibration profile.
Definition: NvInfer.h:10493
virtual ~IBuilderConfig() noexcept=default
DeviceType getDeviceType(ILayer const *layer) const noexcept
Get the device that this layer executes on.
Definition: NvInfer.h:10285
bool canRunOnDLA(ILayer const *layer) const noexcept
Checks if a layer can run on DLA.
Definition: NvInfer.h:10317
TRT_DEPRECATED bool getQuantizationFlag(QuantizationFlag flag) const noexcept
Returns true if the quantization flag is set.
Definition: NvInfer.h:10580
cudaStream_t getProfileStream() const noexcept
Get the CUDA stream that is used to profile this network.
Definition: NvInfer.h:10398
void setHardwareCompatibilityLevel(HardwareCompatibilityLevel hardwareCompatibilityLevel) noexcept
Set the hardware compatibility level.
Definition: NvInfer.h:10815
TilingOptimizationLevel getTilingOptimizationLevel() const noexcept
Get the Tiling optimization level.
Definition: NvInfer.h:11009
TRT_DEPRECATED void setQuantizationFlag(QuantizationFlag flag) noexcept
Set a single quantization flag.
Definition: NvInfer.h:10566
void setMaxAuxStreams(int32_t nbStreams) noexcept
Set the maximum number of auxiliary streams that TRT is allowed to use.
Definition: NvInfer.h:10893
ProfilingVerbosity getProfilingVerbosity() const noexcept
Get verbosity level of layer information exposed in NVTX annotations and IEngineInspector.
Definition: NvInfer.h:10453
bool isDeviceTypeSet(ILayer const *layer) const noexcept
whether the DeviceType has been explicitly set for this layer
Definition: NvInfer.h:10297
void setBuilderOptimizationLevel(int32_t level) noexcept
Set builder optimization level.
Definition: NvInfer.h:10786
void setProfileStream(const cudaStream_t stream) noexcept
Set the CUDA stream that is used to profile this network.
Definition: NvInfer.h:10386
TRT_DEPRECATED IAlgorithmSelector * getAlgorithmSelector() const noexcept
Get Algorithm Selector.
Definition: NvInfer.h:10475
Builds an engine from a network definition.
Definition: NvInfer.h:11129
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:11160
int32_t getNbDLACores() const noexcept
Return the number of DLA engines available to this builder.
Definition: NvInfer.h:11168
IErrorRecorder * getErrorRecorder() const noexcept
get the ErrorRecorder assigned to this interface.
Definition: NvInfer.h:11275
apiv::VBuilder * mImpl
Definition: NvInfer.h:11456
ILogger * getLogger() const noexcept
get the logger with which the builder was created
Definition: NvInfer.h:11410
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:11400
int32_t getMaxThreads() const noexcept
get the maximum number of threads that can be used by the builder.
Definition: NvInfer.h:11440
IPluginRegistry & getPluginRegistry() noexcept
get the local plugin registry that can be used by the builder.
Definition: NvInfer.h:11450
TRT_DEPRECATED bool platformHasFastInt8() const noexcept
Determine whether the platform has fast native int8.
Definition: NvInfer.h:11148
nvinfer1::IOptimizationProfile * createOptimizationProfile() noexcept
Create a new optimization profile.
Definition: NvInfer.h:11241
void setGpuAllocator(IGpuAllocator *allocator) noexcept
Set the GPU allocator.
Definition: NvInfer.h:11186
nvinfer1::INetworkDefinition * createNetworkV2(NetworkDefinitionCreationFlags flags) noexcept
Create a network definition object.
Definition: NvInfer.h:11226
nvinfer1::IBuilderConfig * createBuilderConfig() noexcept
Create a builder configuration object.
Definition: NvInfer.h:11200
void reset() noexcept
Resets the builder state to default values.
Definition: NvInfer.h:11283
bool setMaxThreads(int32_t maxThreads) noexcept
Set the maximum number of threads.
Definition: NvInfer.h:11426
void setErrorRecorder(IErrorRecorder *recorder) noexcept
Set the ErrorRecorder for this interface.
Definition: NvInfer.h:11260
nvinfer1::IHostMemory * buildSerializedNetwork(INetworkDefinition &network, IBuilderConfig &config) noexcept
Builds and serializes a network for the given INetworkDefinition and IBuilderConfig.
Definition: NvInfer.h:11312
virtual ~IBuilder() noexcept=default
TRT_DEPRECATED bool platformHasTf32() const noexcept
Determine whether the platform has TF32 support.
Definition: NvInfer.h:11293
bool buildSerializedNetworkToStream(INetworkDefinition &network, IBuilderConfig &config, IStreamWriter &writer) noexcept
Builds and serializes a network into stream for the given INetworkDefinition and IBuilderConfig.
Definition: NvInfer.h:11333
nvinfer1::ICudaEngine * buildEngineWithConfig(INetworkDefinition &network, IBuilderConfig &config) noexcept
Builds a network for the given INetworkDefinition and IBuilderConfig.
Definition: NvInfer.h:11378
nvinfer1::IHostMemory * buildSerializedNetwork(INetworkDefinition &network, IBuilderConfig &config, IHostMemory *&kernelText) noexcept
Extended form of buildSerializedNetwork that optionally permits getting the kernelText.
Definition: NvInfer.h:11357
A cast layer in a network.
Definition: NvInfer.h:3915
virtual ~ICastLayer() noexcept=default
apiv::VCastLayer * mImpl
Definition: NvInfer.h:3941
DataType getToType() const noexcept
Return cast layer output type.
Definition: NvInfer.h:3935
void setToType(DataType toType) noexcept
Set cast layer output type.
Definition: NvInfer.h:3924
A concatenation layer in a network definition.
Definition: NvInfer.h:2095
void setAxis(int32_t axis) noexcept
Set the axis along which concatenation occurs.
Definition: NvInfer.h:2108
int32_t getAxis() const noexcept
Get the axis along which concatenation occurs.
Definition: NvInfer.h:2118
virtual ~IConcatenationLayer() noexcept=default
This layer represents a condition input to an IIfConditional.
Definition: NvInfer.h:4578
virtual ~IConditionLayer() noexcept=default
Layer that represents a constant value.
Definition: NvInfer.h:3954
void setWeights(Weights weights) noexcept
Set the weights for the layer.
Definition: NvInfer.h:3964
Weights getWeights() const noexcept
Get the weights for the layer.
Definition: NvInfer.h:3974
void setDimensions(Dims const &dimensions) noexcept
Set the dimensions for the layer.
Definition: NvInfer.h:3986
apiv::VConstantLayer * mImpl
Definition: NvInfer.h:4004
virtual ~IConstantLayer() noexcept=default
Dims getDimensions() const noexcept
Get the dimensions for the layer.
Definition: NvInfer.h:3998
A convolution layer in a network definition.
Definition: NvInfer.h:1065
Dims getPrePadding() const noexcept
Get the pre-padding.
Definition: NvInfer.h:1190
Weights getBiasWeights() const noexcept
Get the bias weights for the convolution.
Definition: NvInfer.h:1163
void setPaddingMode(PaddingMode paddingMode) noexcept
Set the padding mode.
Definition: NvInfer.h:1231
void setDilationNd(Dims const &dilation) noexcept
Set the multi-dimension dilation of the convolution.
Definition: NvInfer.h:1335
Dims getPaddingNd() const noexcept
Get the multi-dimension padding of the convolution.
Definition: NvInfer.h:1321
Dims getStrideNd() const noexcept
Get the multi-dimension stride of the convolution.
Definition: NvInfer.h:1291
Weights getKernelWeights() const noexcept
Get the kernel weights of the convolution.
Definition: NvInfer.h:1138
void setStrideNd(Dims const &stride) noexcept
Set the multi-dimension stride of the convolution.
Definition: NvInfer.h:1281
Dims getDilationNd() const noexcept
Get the multi-dimension dilation of the convolution.
Definition: NvInfer.h:1345
int64_t getNbOutputMaps() const noexcept
Get the number of output maps for the convolution.
Definition: NvInfer.h:1084
void setKernelWeights(Weights weights) noexcept
Set the kernel weights for the convolution.
Definition: NvInfer.h:1128
Dims getPostPadding() const noexcept
Get the post-padding.
Definition: NvInfer.h:1217
int64_t getNbGroups() const noexcept
Get the number of groups of the convolution.
Definition: NvInfer.h:1114
PaddingMode getPaddingMode() const noexcept
Get the padding mode.
Definition: NvInfer.h:1243
virtual ~IConvolutionLayer() noexcept=default
void setNbGroups(int64_t nbGroups) noexcept
Set the number of groups for a convolution.
Definition: NvInfer.h:1104
void setNbOutputMaps(int64_t nbOutputMaps) noexcept
Set the number of output maps for the convolution.
Definition: NvInfer.h:1074
void setBiasWeights(Weights weights) noexcept
Set the bias weights for the convolution.
Definition: NvInfer.h:1153
Dims getKernelSizeNd() const noexcept
Get the multi-dimension kernel size of the convolution.
Definition: NvInfer.h:1266
void setPaddingNd(Dims const &padding) noexcept
Set the multi-dimension padding of the convolution.
Definition: NvInfer.h:1309
void setPrePadding(Dims const &padding) noexcept
Set the multi-dimension pre-padding of the convolution.
Definition: NvInfer.h:1180
void setPostPadding(Dims const &padding) noexcept
Set the multi-dimension post-padding of the convolution.
Definition: NvInfer.h:1207
void setKernelSizeNd(Dims const &kernelSize) noexcept
Set the multi-dimension kernel size of the convolution.
Definition: NvInfer.h:1256
An engine for executing inference on a built network, with functionally unsafe features.
Definition: NvInferRuntime.h:3197
Layer that represents a cumulative operation across a tensor.
Definition: NvInfer.h:6672
bool setOperation(CumulativeOperation op) noexcept
Set the cumulative operation for the layer.
Definition: NvInfer.h:6683
void setReverse(bool reverse) noexcept
Specify whether the cumulative operation should be applied backward.
Definition: NvInfer.h:6731
apiv::VCumulativeLayer * mImpl
Definition: NvInfer.h:6749
bool getExclusive() const noexcept
Get whether it is exclusive accumulation or inclusive accumulation.
Definition: NvInfer.h:6719
virtual ~ICumulativeLayer() noexcept=default
bool getReverse() const noexcept
Get the boolean that specifies whether the cumulative operation should be applied backward.
Definition: NvInfer.h:6743
void setExclusive(bool exclusive) noexcept
Set whether it is an exclusive accumulation or inclusive accumulation.
Definition: NvInfer.h:6707
CumulativeOperation getOperation() const noexcept
Get the cumulative operation for the layer.
Definition: NvInfer.h:6695
A deconvolution layer in a network definition.
Definition: NvInfer.h:2136
void setBiasWeights(Weights weights) noexcept
Set the bias weights for the deconvolution.
Definition: NvInfer.h:2224
int64_t getNbGroups() const noexcept
Get the number of groups for a deconvolution.
Definition: NvInfer.h:2185
Weights getKernelWeights() const noexcept
Get the kernel weights for the deconvolution.
Definition: NvInfer.h:2209
void setPrePadding(Dims const &padding) noexcept
Set the multi-dimension pre-padding of the deconvolution.
Definition: NvInfer.h:2251
Dims getStrideNd() const noexcept
Get the multi-dimension stride of the deconvolution.
Definition: NvInfer.h:2366
Dims getDilationNd() const noexcept
Get the multi-dimension dilation of the deconvolution.
Definition: NvInfer.h:2432
Weights getBiasWeights() const noexcept
Get the bias weights for the deconvolution.
Definition: NvInfer.h:2234
void setKernelWeights(Weights weights) noexcept
Set the kernel weights for the deconvolution.
Definition: NvInfer.h:2199
int64_t getNbOutputMaps() const noexcept
Get the number of output feature maps for the deconvolution.
Definition: NvInfer.h:2155
void setStrideNd(Dims const &stride) noexcept
Set the multi-dimension stride of the deconvolution.
Definition: NvInfer.h:2356
Dims getPostPadding() const noexcept
Get the padding.
Definition: NvInfer.h:2288
Dims getKernelSizeNd() const noexcept
Get the multi-dimension kernel size of the deconvolution.
Definition: NvInfer.h:2339
void setPostPadding(Dims const &padding) noexcept
Set the multi-dimension post-padding of the deconvolution.
Definition: NvInfer.h:2278
void setKernelSizeNd(Dims const &kernelSize) noexcept
Set the multi-dimension kernel size of the deconvolution.
Definition: NvInfer.h:2329
virtual ~IDeconvolutionLayer() noexcept=default
void setPaddingNd(Dims const &padding) noexcept
Set the multi-dimension padding of the deconvolution.
Definition: NvInfer.h:2384
void setNbOutputMaps(int64_t nbOutputMaps) noexcept
Set the number of output feature maps for the deconvolution.
Definition: NvInfer.h:2145
Dims getPaddingNd() const noexcept
Get the multi-dimension padding of the deconvolution.
Definition: NvInfer.h:2396
void setDilationNd(Dims const &dilation) noexcept
Set the multi-dimension dilation of the deconvolution.
Definition: NvInfer.h:2422
void setPaddingMode(PaddingMode paddingMode) noexcept
Set the padding mode.
Definition: NvInfer.h:2302
void setNbGroups(int64_t nbGroups) noexcept
Set the number of groups for a deconvolution.
Definition: NvInfer.h:2175
Dims getPrePadding() const noexcept
Get the pre-padding.
Definition: NvInfer.h:2261
PaddingMode getPaddingMode() const noexcept
Get the padding mode.
Definition: NvInfer.h:2314
A Dequantize layer in a network definition.
Definition: NvInfer.h:5639
void setToType(DataType toType) noexcept
Set the Dequantize layer output type.
Definition: NvInfer.h:5676
virtual ~IDequantizeLayer() noexcept=default
int32_t getAxis() const noexcept
Get the quantization axis.
Definition: NvInfer.h:5649
DataType getToType() const noexcept
Return the Dequantize layer output type.
Definition: NvInfer.h:5688
void setAxis(int32_t axis) noexcept
Set the quantization axis.
Definition: NvInfer.h:5660
A network layer to perform dynamic quantization.
Definition: NvInfer.h:5716
int32_t getAxis() const noexcept
Get the axis along which blocking occurs.
Definition: NvInfer.h:5805
int32_t getBlockSize() const noexcept
Get the size of the quantization block.
Definition: NvInfer.h:5828
DataType getScaleType() const noexcept
Return the scale factors data type.
Definition: NvInfer.h:5782
void setScaleType(DataType scaleType) noexcept
Set the data type of the scale factors used to quantize the data.
Definition: NvInfer.h:5769
DataType getToType() const noexcept
Return DynamicQuantizeLayer's quantized output type.
Definition: NvInfer.h:5756
virtual ~IDynamicQuantizeLayer() noexcept=default
void setToType(DataType toType) noexcept
Set DynamicQuantizeLayer's quantized output type.
Definition: NvInfer.h:5743
void setAxis(int32_t axis) noexcept
Set the axis along which block quantization occurs.
Definition: NvInfer.h:5795
void setBlockSize(int32_t size) noexcept
Set the size of the quantization block.
Definition: NvInfer.h:5818
An Einsum layer in a network.
Definition: NvInfer.h:5873
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:5884
virtual ~IEinsumLayer() noexcept=default
char const * getEquation() const noexcept
Return the equation.
Definition: NvInfer.h:5894
A elementwise layer in a network definition.
Definition: NvInfer.h:2506
virtual ~IElementWiseLayer() noexcept=default
apiv::VElementWiseLayer * mImpl
Definition: NvInfer.h:2535
ElementWiseOperation getOperation() const noexcept
Get the binary operation for the layer.
Definition: NvInfer.h:2529
void setOperation(ElementWiseOperation op) noexcept
Set the binary operation for the layer.
Definition: NvInfer.h:2517
Generate a tensor according to a specified mode.
Definition: NvInfer.h:5165
bool isAlphaBetaInt64() const noexcept
Return true if alpha/beta have type int64, false if they have type double.
Definition: NvInfer.h:5397
FillOperation getOperation() const noexcept
Get the fill operation for the layer.
Definition: NvInfer.h:5211
void setOperation(FillOperation op) noexcept
Set the fill operation for the layer.
Definition: NvInfer.h:5201
DataType getToType() const noexcept
Get the fill layer output type.
Definition: NvInfer.h:5426
void setAlphaInt64(int64_t alpha) noexcept
Set the alpha parameter with int64 datatype.
Definition: NvInfer.h:5340
void setBetaInt64(int64_t beta) noexcept
Set the beta parameter with int64 datatype.
Definition: NvInfer.h:5374
void setBeta(double beta) noexcept
Set the beta parameter.
Definition: NvInfer.h:5264
int64_t getAlphaInt64() const noexcept
Get the value of alpha parameter with int64 datatype.
Definition: NvInfer.h:5355
int64_t getBetaInt64() const noexcept
Get the value of beta parameter with int64 datatype.
Definition: NvInfer.h:5389
double getAlpha() const noexcept
Get the value of alpha parameter.
Definition: NvInfer.h:5245
void setDimensions(Dims const &dimensions) noexcept
Set the output tensor's dimensions.
Definition: NvInfer.h:5176
void setAlpha(double alpha) noexcept
Set the alpha parameter.
Definition: NvInfer.h:5230
void setToType(DataType toType) noexcept
Set the fill layer output type.
Definition: NvInfer.h:5414
Dims getDimensions() const noexcept
Get the output tensor's dimensions.
Definition: NvInfer.h:5191
double getBeta() const noexcept
Get the value of beta parameter.
Definition: NvInfer.h:5279
virtual ~IFillLayer() noexcept=default
A Gather layer in a network definition. Supports several kinds of gathering.
Definition: NvInfer.h:2639
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:2650
void setNbElementWiseDims(int32_t elementWiseDims) noexcept
Set the number of leading dimensions of indices tensor to be handled elementwise.
Definition: NvInfer.h:2685
apiv::VGatherLayer * mImpl
Definition: NvInfer.h:2721
int32_t getNbElementWiseDims() const noexcept
Get the number of leading dimensions of indices tensor to be handled elementwise.
Definition: NvInfer.h:2695
void setMode(GatherMode mode) noexcept
Set the gather mode.
Definition: NvInfer.h:2705
int32_t getGatherAxis() const noexcept
Get the axis to gather on.
Definition: NvInfer.h:2662
GatherMode getMode() const noexcept
Get the gather mode.
Definition: NvInfer.h:2715
virtual ~IGatherLayer() noexcept=default
A GridSample layer in a network definition.
Definition: NvInfer.h:6095
void setInterpolationMode(InterpolationMode mode) noexcept
Set the grid sample interpolation mode.
Definition: NvInfer.h:6102
bool setSampleMode(SampleMode mode) noexcept
Set the sample mode.
Definition: NvInfer.h:6148
void setAlignCorners(bool alignCorners) noexcept
Set the align corners mode.
Definition: NvInfer.h:6124
apiv::VGridSampleLayer * mImpl
Definition: NvInfer.h:6166
SampleMode getSampleMode() const noexcept
Get the sample mode.
Definition: NvInfer.h:6160
InterpolationMode getInterpolationMode() const noexcept
Get the grid sample interpolation mode.
Definition: NvInfer.h:6114
bool getAlignCorners() const noexcept
Get the align corners mode.
Definition: NvInfer.h:6136
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:3902
apiv::VIdentityLayer * mImpl
Definition: NvInfer.h:3904
virtual ~IIdentityLayer() noexcept=default
This is a base class for Conditional boundary layers.
Definition: NvInfer.h:4557
IIfConditional * getConditional() const noexcept
Get a pointer to the IIfConditional associated with this boundary layer.
Definition: NvInfer.h:4562
virtual ~IIfConditionalBoundaryLayer() noexcept=default
Helper for constructing conditionally-executed subgraphs.
Definition: NvInfer.h:4640
IIfConditionalInputLayer * addInput(ITensor &input) noexcept
Add an If-conditional input.
Definition: NvInfer.h:4681
char const * getName() const noexcept
Return the name of the conditional.
Definition: NvInfer.h:4706
virtual ~IIfConditional() noexcept=default
IConditionLayer * setCondition(ITensor &condition) noexcept
Set the condition tensor for this If-Conditional construct.
Definition: NvInfer.h:4651
IIfConditionalOutputLayer * addOutput(ITensor &trueSubgraphOutput, ITensor &falseSubgraphOutput) noexcept
Add an If-conditional output.
Definition: NvInfer.h:4669
void setName(char const *name) noexcept
Set the name of the conditional.
Definition: NvInfer.h:4696
This layer represents an output of an IIfConditional.
Definition: NvInfer.h:4595
virtual ~IIfConditionalOutputLayer() noexcept=default
Application-implemented interface for calibration.
Definition: NvInfer.h:8817
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:4871
virtual ~IIteratorLayer() noexcept=default
void setReverse(bool reverse) noexcept
Set iteration order to be reverse.
Definition: NvInfer.h:4898
bool getReverse() const noexcept
Check if the iteration order is reverse.
Definition: NvInfer.h:4908
int32_t getAxis() const noexcept
Get axis being iterated over.
Definition: NvInfer.h:4884
void setAxis(int32_t axis) noexcept
Set axis to iterate over.
Definition: NvInfer.h:4876
A LRN layer in a network definition.
Definition: NvInfer.h:1750
int64_t getWindowSize() const noexcept
Get the LRN window size.
Definition: NvInfer.h:1771
float getAlpha() const noexcept
Get the LRN alpha value.
Definition: NvInfer.h:1793
void setWindowSize(int64_t windowSize) noexcept
Set the LRN window size.
Definition: NvInfer.h:1761
void setK(float k) noexcept
Set the LRN K value.
Definition: NvInfer.h:1827
void setAlpha(float alpha) noexcept
Set the LRN alpha value.
Definition: NvInfer.h:1783
void setBeta(float beta) noexcept
Set the LRN beta value.
Definition: NvInfer.h:1805
virtual ~ILRNLayer() noexcept=default
float getBeta() const noexcept
Get the LRN beta value.
Definition: NvInfer.h:1815
float getK() const noexcept
Get the LRN K value.
Definition: NvInfer.h:1837
Base class for all layer classes in a network definition.
Definition: NvInfer.h:580
TRT_DEPRECATED void setPrecision(DataType dataType) noexcept
Set the preferred or required computational precision of this layer in a weakly-typed network.
Definition: NvInfer.h:700
TRT_DEPRECATED void setOutputType(int32_t index, DataType dataType) noexcept
Set the output type of this layer in a weakly-typed network.
Definition: NvInfer.h:788
TRT_DEPRECATED bool precisionIsSet() const noexcept
whether the computational precision has been set for this layer
Definition: NvInfer.h:726
void setMetadata(char const *metadata) noexcept
Set the metadata for this layer.
Definition: NvInfer.h:851
TRT_DEPRECATED void resetOutputType(int32_t index) noexcept
reset the output type for this layer
Definition: NvInfer.h:833
void setName(char const *name) noexcept
Set the name of a layer.
Definition: NvInfer.h:601
int32_t getNbInputs() const noexcept
Get the number of inputs of a layer.
Definition: NvInfer.h:619
char const * getMetadata() const noexcept
Get the metadata of the layer.
Definition: NvInfer.h:864
DataType getOutputType(int32_t index) const noexcept
get the output type of this layer
Definition: NvInfer.h:803
DataType getPrecision() const noexcept
get the computational precision of this layer
Definition: NvInfer.h:712
TRT_DEPRECATED bool outputTypeIsSet(int32_t index) const noexcept
whether the output type has been set for this layer
Definition: NvInfer.h:819
char const * getName() const noexcept
Return the name of a layer.
Definition: NvInfer.h:611
int32_t getNbOutputs() const noexcept
Get the number of outputs of a layer.
Definition: NvInfer.h:640
ITensor * getOutput(int32_t index) const noexcept
Get the layer output corresponding to the given index.
Definition: NvInfer.h:650
void setInput(int32_t index, ITensor &tensor) noexcept
Replace an input of this layer with a specific tensor.
Definition: NvInfer.h:667
ITensor * getInput(int32_t index) const noexcept
Get the layer input corresponding to the given index.
Definition: NvInfer.h:632
LayerType getType() const noexcept
Return the type of a layer.
Definition: NvInfer.h:587
TRT_DEPRECATED void resetPrecision() noexcept
reset the computational precision for this layer
Definition: NvInfer.h:738
virtual ~ILayer() noexcept=default
Application-implemented logging interface for the builder, refitter and runtime.
Definition: NvInferRuntime.h:1588
This is a base class for Loop boundary layers.
Definition: NvInfer.h:4534
virtual ~ILoopBoundaryLayer() noexcept=default
ILoop * getLoop() const noexcept
Get a pointer to ILoop associated with this boundary layer.
Definition: NvInfer.h:4539
Helper for creating a recurrent subgraph.
Definition: NvInfer.h:4929
void setName(char const *name) noexcept
Set the name of the loop.
Definition: NvInfer.h:4999
ITripLimitLayer * addTripLimit(ITensor &tensor, TripLimit limit) noexcept
Add a trip-count limiter, based on the given tensor.
Definition: NvInfer.h:4958
IIteratorLayer * addIterator(ITensor &tensor, int32_t axis=0, bool reverse=false) noexcept
Return layer that subscripts tensor by loop iteration.
Definition: NvInfer.h:4971
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:4984
virtual ~ILoop() noexcept=default
char const * getName() const noexcept
Return the name of the loop.
Definition: NvInfer.h:5009
IRecurrenceLayer * addRecurrence(ITensor &initialValue) noexcept
Create a recurrence layer for this loop with initialValue as its first input.
Definition: NvInfer.h:4937
An ILoopOutputLayer is the sole way to get output from a loop.
Definition: NvInfer.h:4771
virtual ~ILoopOutputLayer() noexcept=default
int32_t getAxis() const noexcept
Get axis being concatenated over.
Definition: NvInfer.h:4801
LoopOutput getLoopOutput() const noexcept
Get which kind a loop output has.
Definition: NvInfer.h:4776
void setAxis(int32_t axis) noexcept
Set where to insert the contenation axis. Ignored if getLoopOutput() is kLAST_VALUE.
Definition: NvInfer.h:4793
Layer that represents a Matrix Multiplication.
Definition: NvInfer.h:3749
apiv::VMatrixMultiplyLayer * mImpl
Definition: NvInfer.h:3777
virtual ~IMatrixMultiplyLayer() noexcept=default
MatrixOperation getOperation(int32_t index) const noexcept
Get the operation for an input tensor.
Definition: NvInfer.h:3771
void setOperation(int32_t index, MatrixOperation op) noexcept
Set the operation for an input tensor.
Definition: NvInfer.h:3759
A non-maximum suppression layer in a network definition.
Definition: NvInfer.h:6247
virtual ~INMSLayer() noexcept=default
void setTopKBoxLimit(int32_t limit) noexcept
Set the TopK box limit parameter for the layer.
Definition: NvInfer.h:6284
void setBoundingBoxFormat(BoundingBoxFormat fmt) noexcept
Set the bounding box format parameter for the layer.
Definition: NvInfer.h:6258
BoundingBoxFormat getBoundingBoxFormat() const noexcept
Get the bounding box format parameter for the layer.
Definition: NvInfer.h:6270
bool setIndicesType(DataType type) noexcept
Set the indices type for the layer.
Definition: NvInfer.h:6329
apiv::VNMSLayer * mImpl
Definition: NvInfer.h:6347
int32_t getTopKBoxLimit() const noexcept
Get the TopK box limit parameter for the layer.
Definition: NvInfer.h:6294
DataType getIndicesType() const noexcept
Return the NMS layer indices type.
Definition: NvInfer.h:6341
A network definition for input to the builder.
Definition: NvInfer.h:7174
IConcatenationLayer * addConcatenation(ITensor *const *inputs, int32_t nbInputs) noexcept
Add a concatenation layer to the network.
Definition: NvInfer.h:7402
IShuffleLayer * addShuffle(ITensor &input) noexcept
Add a shuffle layer to the network.
Definition: NvInfer.h:7465
INormalizationLayer * addNormalization(ITensor &input, ITensor &scale, ITensor &bias, uint32_t axesMask) noexcept
Add a normalization layer to the network.
Definition: NvInfer.h:8629
void setName(char const *name) noexcept
Sets the name of the network.
Definition: NvInfer.h:7933
ITopKLayer * addTopK(ITensor &input, TopKOperation op, int32_t k, uint32_t reduceAxes, DataType indicesType) noexcept
Add a TopK layer to the network.
Definition: NvInfer.h:7663
bool markDebug(ITensor &tensor) noexcept
Mark a tensor as a debug tensor.
Definition: NvInfer.h:7245
ILRNLayer * addLRN(ITensor &input, int64_t window, float alpha, float beta, float k) noexcept
Add a LRN layer to the network.
Definition: NvInfer.h:7346
ICumulativeLayer * addCumulative(ITensor &input, ITensor &axis, CumulativeOperation operation, bool exclusive, bool reverse) noexcept
Add a cumulative layer to the network.
Definition: NvInfer.h:8651
IAssertionLayer * addAssertion(ITensor &condition, char const *message) noexcept
Add an assertion layer to the network.
Definition: NvInfer.h:8249
TRT_DEPRECATED INonZeroLayer * addNonZero(ITensor &input) noexcept
Add a nonzero layer to the network.
Definition: NvInfer.h:7754
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:8068
ICastLayer * addCast(ITensor &input, DataType toType) noexcept
Add a cast layer.
Definition: NvInfer.h:7823
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:8147
char const * getName() const noexcept
Returns the name associated with the network.
Definition: NvInfer.h:7947
IParametricReLULayer * addParametricReLU(ITensor &input, ITensor &slope) noexcept
Add a parametric ReLU layer to the network.
Definition: NvInfer.h:8046
ITensor * getOutput(int32_t index) const noexcept
Get the output tensor specified by the given index.
Definition: NvInfer.h:7566
ITensor * getInput(int32_t index) const noexcept
Get the input tensor specified by the given index.
Definition: NvInfer.h:7536
TRT_DEPRECATED ITopKLayer * addTopK(ITensor &input, TopKOperation op, int32_t k, uint32_t reduceAxes) noexcept
Add a TopK layer to the network.
Definition: NvInfer.h:7629
IDequantizeLayer * addDequantize(ITensor &input, ITensor &scale, DataType outputType) noexcept
Add a dequantization layer to the network.
Definition: NvInfer.h:8418
bool unmarkOutputForShapes(ITensor &tensor) noexcept
Undo markOutputForShapes.
Definition: NvInfer.h:8028
IFillLayer * addFill(Dims const &dimensions, FillOperation op, DataType outputType) noexcept
Add a fill layer to the network.
Definition: NvInfer.h:8300
ILoop * addLoop() noexcept
Add a loop to the network.
Definition: NvInfer.h:8178
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:8510
bool markUnfusedTensorsAsDebugTensors() noexcept
Mark unfused tensors as debug tensors.
Definition: NvInfer.h:7293
IActivationLayer * addActivation(ITensor &input, ActivationType type) noexcept
Add an activation layer to the network.
Definition: NvInfer.h:7327
TRT_DEPRECATED IFillLayer * addFill(Dims const &dimensions, FillOperation op) noexcept
Add a fill layer to the network.
Definition: NvInfer.h:8274
ISliceLayer * addSlice(ITensor &input, Dims const &start, Dims const &size, Dims const &stride) noexcept
Add a slice layer to the network.
Definition: NvInfer.h:7909
virtual ~INetworkDefinition() noexcept=default
TRT_DEPRECATED IQuantizeLayer * addQuantize(ITensor &input, ITensor &scale) noexcept
Add a quantization layer to the network.
Definition: NvInfer.h:8459
virtual IBuilder & getBuilder() const noexcept
Return the builder from which this INetworkDefinition was created.
Definition: NvInfer.h:8690
ILayer * getLayer(int32_t index) const noexcept
Get the layer specified by the given index.
Definition: NvInfer.h:7508
bool isDebugTensor(ITensor const &tensor) const noexcept
Check if a tensor is marked as debug tensor.
Definition: NvInfer.h:7271
bool getFlag(NetworkDefinitionCreationFlag networkDefinitionCreationFlag) const noexcept
Returns true if the network definition creation flag is set.
Definition: NvInfer.h:7999
IIfConditional * addIfConditional() noexcept
Add an if-then-else to the network.
Definition: NvInfer.h:8193
IErrorRecorder * getErrorRecorder() const noexcept
get the ErrorRecorder assigned to this interface.
Definition: NvInfer.h:8374
ISqueezeLayer * addSqueeze(ITensor &input, ITensor &axes) noexcept
Add a squeeze layer to the network.
Definition: NvInfer.h:8747
TRT_DEPRECATED INMSLayer * addNMS(ITensor &boxes, ITensor &scores, ITensor &maxOutputBoxesPerClass) noexcept
Add a non-maximum suppression layer to the network.
Definition: NvInfer.h:8566
IReverseSequenceLayer * addReverseSequence(ITensor &input, ITensor &sequenceLens) noexcept
Add a ReverseSequence layer to the network.
Definition: NvInfer.h:8603
int32_t getNbInputs() const noexcept
Get the number of inputs in the network.
Definition: NvInfer.h:7520
NetworkDefinitionCreationFlags getFlags() const noexcept
Get the network definition creation flags for this network definition object. Defaults to 0.
Definition: NvInfer.h:7987
IQuantizeLayer * addQuantize(ITensor &input, ITensor &scale, DataType outputType) noexcept
Add a quantization layer to the network.
Definition: NvInfer.h:8483
IReduceLayer * addReduce(ITensor &input, ReduceOperation operation, uint32_t reduceAxes, bool keepDimensions) noexcept
Add a reduce layer to the network.
Definition: NvInfer.h:7592
IUnaryLayer * addUnary(ITensor &input, UnaryOperation operation) noexcept
Add a unary layer to the network.
Definition: NvInfer.h:7451
IGridSampleLayer * addGridSample(ITensor &input, ITensor &grid) noexcept
Add a GridSample layer to the network.
Definition: NvInfer.h:8544
void removeTensor(ITensor &tensor) noexcept
remove a tensor from the network definition.
Definition: NvInfer.h:7838
bool areWeightsMarkedRefittable(char const *name) const noexcept
Whether the weight has been marked as refittable.
Definition: NvInfer.h:8728
ISelectLayer * addSelect(ITensor &condition, ITensor &thenInput, ITensor &elseInput) noexcept
Add a select layer to the network.
Definition: NvInfer.h:8232
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:8438
int32_t getNbLayers() const noexcept
Get the number of layers in the network.
Definition: NvInfer.h:7494
TRT_DEPRECATED bool hasImplicitBatchDimension() const noexcept
Query whether the network was created with an implicit batch dimension.
Definition: NvInfer.h:7977
apiv::VNetworkDefinition * mImpl
Definition: NvInfer.h:8774
bool markOutputForShapes(ITensor &tensor) noexcept
Enable tensor's value to be computed by IExecutionContext::getShapeBinding.
Definition: NvInfer.h:8016
IOneHotLayer * addOneHot(ITensor &indices, ITensor &values, ITensor &depth, int32_t axis) noexcept
Add a OneHot layer to the network.
Definition: NvInfer.h:7482
IScaleLayer * addScale(ITensor &input, ScaleMode mode, Weights shift, Weights scale, Weights power) noexcept
Add a Scale layer to the network.
Definition: NvInfer.h:7372
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:7889
void unmarkOutput(ITensor &tensor) noexcept
unmark a tensor as a network output.
Definition: NvInfer.h:7850
IIdentityLayer * addIdentity(ITensor &input) noexcept
Add an identity layer.
Definition: NvInfer.h:7808
IGatherLayer * addGatherV2(ITensor &data, ITensor &indices, GatherMode mode) noexcept
Add gather with specified mode, axis=0 and nbElementWiseDims=0.
Definition: NvInfer.h:7695
INonZeroLayer * addNonZero(ITensor &input, DataType indicesType) noexcept
Add a nonzero layer to the network.
Definition: NvInfer.h:7770
IElementWiseLayer * addElementWise(ITensor &input1, ITensor &input2, ElementWiseOperation op) noexcept
Add an elementwise layer to the network.
Definition: NvInfer.h:7429
IConstantLayer * addConstant(Dims const &dimensions, Weights weights) noexcept
Add a constant layer to the network.
Definition: NvInfer.h:7794
void setErrorRecorder(IErrorRecorder *recorder) noexcept
Set the ErrorRecorder for this interface.
Definition: NvInfer.h:8359
IPoolingLayer * addPoolingNd(ITensor &input, PoolingType type, Dims const &windowSize) noexcept
Add a multi-dimension pooling layer to the network.
Definition: NvInfer.h:8088
INMSLayer * addNMS(ITensor &boxes, ITensor &scores, ITensor &maxOutputBoxesPerClass, DataType indicesType) noexcept
Add a non-maximum suppression layer to the network.
Definition: NvInfer.h:8586
IRaggedSoftMaxLayer * addRaggedSoftMax(ITensor &input, ITensor &bounds) noexcept
Add a RaggedSoftMax layer to the network.
Definition: NvInfer.h:7714
IShapeLayer * addShape(ITensor &input) noexcept
Add a shape layer to the network.
Definition: NvInfer.h:7963
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:7679
IAttention * addAttention(ITensor &query, ITensor &key, ITensor &value, AttentionNormalizationOp normOp, bool causal) noexcept
Add an attention to the network.
Definition: NvInfer.h:8678
bool unmarkWeightsRefittable(char const *name) noexcept
Unmark weights as refittable when the builder flag kREFIT_INDIVIDUAL is set.
Definition: NvInfer.h:8715
bool markWeightsRefittable(char const *name) noexcept
Mark weights as refittable when the builder flag kREFIT_INDIVIDUAL is set.
Definition: NvInfer.h:8703
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:8110
IResizeLayer * addResize(ITensor &input) noexcept
Add a resize layer to the network.
Definition: NvInfer.h:8164
IUnsqueezeLayer * addUnsqueeze(ITensor &input, ITensor &axes) noexcept
Add an unsqueeze layer to the network.
Definition: NvInfer.h:8768
IMatrixMultiplyLayer * addMatrixMultiply(ITensor &input0, MatrixOperation op0, ITensor &input1, MatrixOperation op1) noexcept
Add a MatrixMultiply layer to the network.
Definition: NvInfer.h:7735
ISoftMaxLayer * addSoftMax(ITensor &input) noexcept
Add a SoftMax layer to the network.
Definition: NvInfer.h:7385
bool unmarkDebug(ITensor &tensor) noexcept
Unmark a tensor as a debug tensor.
Definition: NvInfer.h:7261
IEinsumLayer * addEinsum(ITensor *const *inputs, int32_t nbInputs, char const *equation) noexcept
Add an Einsum layer to the network.
Definition: NvInfer.h:8526
void markOutput(ITensor &tensor) noexcept
Mark a tensor as a network output.
Definition: NvInfer.h:7227
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:7871
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:8316
TRT_DEPRECATED IDequantizeLayer * addDequantize(ITensor &input, ITensor &scale) noexcept
Add a dequantization layer to the network.
Definition: NvInfer.h:8395
int32_t getNbOutputs() const noexcept
Get the number of outputs in the network.
Definition: NvInfer.h:7550
bool setWeightsName(Weights weights, char const *name) noexcept
Associate a name with all current uses of the given weights.
Definition: NvInfer.h:8340
bool unmarkUnfusedTensorsAsDebugTensors() noexcept
Undo the marking of unfused tensors as debug tensors.
Definition: NvInfer.h:7307
Forward declaration of IEngineInspector for use by other interfaces.
Definition: NvInferRuntime.h:51
Definition: NvInfer.h:3803
DataType getIndicesType() const noexcept
Return the NonZero layer indices type.
Definition: NvInfer.h:3827
bool setIndicesType(DataType type) noexcept
Set the indices type for the layer.
Definition: NvInfer.h:3815
virtual ~INonZeroLayer() noexcept=default
A normalization layer in a network definition.
Definition: NvInfer.h:6436
float getEpsilon() const noexcept
Get the epsilon value used for the normalization calculation.
Definition: NvInfer.h:6455
uint32_t getAxes() const noexcept
Get the axes value used for the normalization calculation.
Definition: NvInfer.h:6475
virtual ~INormalizationLayer() noexcept=default
void setEpsilon(float eps) noexcept
Set the epsilon value used for the normalization calculation.
Definition: NvInfer.h:6445
DataType getComputePrecision() const noexcept
Get the compute precision of this layer.
Definition: NvInfer.h:6542
apiv::VNormalizationLayer * mImpl
Definition: NvInfer.h:6548
int64_t getNbGroups() const noexcept
Get the number of groups used to split the channels for the normalization calculation.
Definition: NvInfer.h:6506
void setAxes(uint32_t axesMask) noexcept
Set the reduction axes for the normalization calculation.
Definition: NvInfer.h:6465
void setComputePrecision(DataType type) noexcept
Set the compute precision of this layer.
Definition: NvInfer.h:6532
void setNbGroups(int64_t nbGroups) noexcept
Set the number of groups used to split the channels in the normalization calculation.
Definition: NvInfer.h:6496
A OneHot layer in a network definition.
Definition: NvInfer.h:6058
virtual ~IOneHotLayer() noexcept=default
apiv::VOneHotLayer * mImpl
Definition: NvInfer.h:6079
void setAxis(int32_t axis) noexcept
Set the axis parameter.
Definition: NvInfer.h:6065
int32_t getAxis() const noexcept
Get the value of the axis parameter.
Definition: NvInfer.h:6073
Optimization profile for dynamic input dimensions and shape tensors.
Definition: NvInferRuntime.h:2675
Layer that represents a padding operation.
Definition: NvInfer.h:3000
Dims getPostPaddingNd() const noexcept
Get the padding that is applied at the end of the tensor.
Definition: NvInfer.h:3049
void setPrePaddingNd(Dims const &padding) noexcept
Set the padding that is applied at the start of the tensor.
Definition: NvInfer.h:3011
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:3037
Dims getPrePaddingNd() const noexcept
Get the padding that is applied at the start of the tensor.
Definition: NvInfer.h:3023
apiv::VPaddingLayer * mImpl
Definition: NvInfer.h:3055
Layer that represents a parametric ReLU operation.
Definition: NvInfer.h:4018
apiv::VParametricReLULayer * mImpl
Definition: NvInfer.h:4020
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:2737
virtual ~IPluginV2Layer() noexcept=default
apiv::VPluginV2Layer * mImpl
Definition: NvInfer.h:2750
IPluginV2 & getPlugin() noexcept
Get the plugin for the layer.
Definition: NvInfer.h:2744
Layer type for V3 plugins.
Definition: NvInfer.h:2764
virtual ~IPluginV3Layer() noexcept=default
IPluginV3 & getPlugin() noexcept
Get the plugin for the layer.
Definition: NvInfer.h:2771
apiv::VPluginV3Layer * mImpl
Definition: NvInfer.h:2777
A Pooling layer in a network definition.
Definition: NvInfer.h:1499
PoolingType getPoolingType() const noexcept
Get the type of activation to be performed.
Definition: NvInfer.h:1518
PaddingMode getPaddingMode() const noexcept
Get the padding mode.
Definition: NvInfer.h:1651
Dims getPostPadding() const noexcept
Get the padding.
Definition: NvInfer.h:1627
bool getAverageCountExcludesPadding() const noexcept
Get whether average pooling uses as a denominator the overlap area between the window and the unpadde...
Definition: NvInfer.h:1571
Dims getPrePadding() const noexcept
Get the pre-padding.
Definition: NvInfer.h:1599
void setPoolingType(PoolingType type) noexcept
Set the type of activation to be performed.
Definition: NvInfer.h:1508
void setWindowSizeNd(Dims const &windowSize) noexcept
Set the multi-dimension window size for pooling.
Definition: NvInfer.h:1664
void setPaddingMode(PaddingMode paddingMode) noexcept
Set the padding mode.
Definition: NvInfer.h:1640
Dims getWindowSizeNd() const noexcept
Get the multi-dimension window size for pooling.
Definition: NvInfer.h:1674
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:1560
void setPaddingNd(Dims const &padding) noexcept
Set the multi-dimension padding for pooling.
Definition: NvInfer.h:1718
float getBlendFactor() const noexcept
Get the blending factor for the max_average_blend mode: max_average_blendPool = (1-blendFactor)*maxPo...
Definition: NvInfer.h:1546
void setStrideNd(Dims const &stride) noexcept
Set the multi-dimension stride for pooling.
Definition: NvInfer.h:1689
Dims getStrideNd() const noexcept
Get the multi-dimension stride for pooling.
Definition: NvInfer.h:1699
virtual ~IPoolingLayer() noexcept=default
Dims getPaddingNd() const noexcept
Get the multi-dimension padding for pooling.
Definition: NvInfer.h:1730
void setPostPadding(Dims const &padding) noexcept
Set the multi-dimension post-padding for pooling.
Definition: NvInfer.h:1617
void setPrePadding(Dims const &padding) noexcept
Set the multi-dimension pre-padding for pooling.
Definition: NvInfer.h:1589
void setBlendFactor(float blendFactor) noexcept
Set the blending factor for the max_average_blend mode: max_average_blendPool = (1-blendFactor)*maxPo...
Definition: NvInfer.h:1533
A Quantize layer in a network definition.
Definition: NvInfer.h:5511
void setToType(DataType toType) noexcept
Set the Quantize layer output type.
Definition: NvInfer.h:5548
void setAxis(int32_t axis) noexcept
Set the quantization axis.
Definition: NvInfer.h:5532
int32_t getAxis() const noexcept
Get the quantization axis.
Definition: NvInfer.h:5521
virtual ~IQuantizeLayer() noexcept=default
DataType getToType() const noexcept
Return the Quantize layer output type.
Definition: NvInfer.h:5560
A RaggedSoftmax layer in a network definition.
Definition: NvInfer.h:3852
apiv::VRaggedSoftMaxLayer * mImpl
Definition: NvInfer.h:3854
virtual ~IRaggedSoftMaxLayer() noexcept=default
A recurrence layer in a network definition.
Definition: NvInfer.h:4724
virtual ~IRecurrenceLayer() noexcept=default
Layer that represents a reduction across a non-bool tensor.
Definition: NvInfer.h:2920
void setKeepDimensions(bool keepDimensions) noexcept
Set the boolean that specifies whether or not to keep the reduced dimensions for the layer.
Definition: NvInfer.h:2967
void setOperation(ReduceOperation op) noexcept
Set the reduce operation for the layer.
Definition: NvInfer.h:2927
ReduceOperation getOperation() const noexcept
Get the reduce operation for the layer.
Definition: NvInfer.h:2937
virtual ~IReduceLayer() noexcept=default
uint32_t getReduceAxes() const noexcept
Get the axes over which to reduce for the layer.
Definition: NvInfer.h:2957
void setReduceAxes(uint32_t reduceAxes) noexcept
Set the axes over which to reduce.
Definition: NvInfer.h:2947
apiv::VReduceLayer * mImpl
Definition: NvInfer.h:2983
bool getKeepDimensions() const noexcept
Get the boolean that specifies whether or not to keep the reduced dimensions for the layer.
Definition: NvInfer.h:2977
A resize layer in a network definition.
Definition: NvInfer.h:4207
void setSelectorForSinglePixel(ResizeSelector selector) noexcept
Set coordinate selector function when resized to single pixel.
Definition: NvInfer.h:4368
void setNearestRounding(ResizeRoundMode value) noexcept
Set rounding mode for nearest neighbor resize.
Definition: NvInfer.h:4392
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:4286
void setOutputDimensions(Dims const &dimensions) noexcept
Set the output dimensions.
Definition: NvInfer.h:4227
void setCubicCoeff(float A) noexcept
Set the coefficient 'A' used in cubic interpolation.
Definition: NvInfer.h:4424
void setScales(float const *scales, int32_t nbScales) noexcept
Set the resize scales.
Definition: NvInfer.h:4267
float getCubicCoeff() const noexcept
Get the coefficient 'A' used in cubic interpolation.
Definition: NvInfer.h:4434
ResizeSelector getSelectorForSinglePixel() const noexcept
Get the coordinate selector function when resized to single pixel.
Definition: NvInfer.h:4378
InterpolationMode getResizeMode() const noexcept
Get resize mode for an input tensor.
Definition: NvInfer.h:4308
void setCoordinateTransformation(ResizeCoordinateTransformation coordTransform) noexcept
Set coordinate transformation function.
Definition: NvInfer.h:4343
void setExcludeOutside(bool excludeFlag) noexcept
Set the state for excluding outside pixels.
Definition: NvInfer.h:4447
void setResizeMode(InterpolationMode interpolationMode) noexcept
Set resize mode for an input tensor.
Definition: NvInfer.h:4298
Dims getOutputDimensions() const noexcept
Get the output dimensions.
Definition: NvInfer.h:4237
ResizeRoundMode getNearestRounding() const noexcept
Get rounding mode for nearest neighbor resize.
Definition: NvInfer.h:4402
bool getExcludeOutside() const noexcept
Get the state for excluding outside pixels.
Definition: NvInfer.h:4457
ResizeCoordinateTransformation getCoordinateTransformation() const noexcept
Get coordinate transformation function.
Definition: NvInfer.h:4353
A ReverseSequence layer in a network definition.
Definition: NvInfer.h:6364
void setSequenceAxis(int32_t sequenceAxis) noexcept
Set the sequence axis. Default is 0.
Definition: NvInfer.h:6397
int32_t getBatchAxis() const noexcept
Return the batch axis. Return 1 if no batch axis was set.
Definition: NvInfer.h:6384
apiv::VReverseSequenceLayer * mImpl
Definition: NvInfer.h:6413
int32_t getSequenceAxis() const noexcept
Return the sequence axis. Return 0 if no sequence axis was set.
Definition: NvInfer.h:6407
void setBatchAxis(int32_t batchAxis) noexcept
Set the batch axis. Default is 1.
Definition: NvInfer.h:6374
virtual ~IReverseSequenceLayer() noexcept=default
A Scale layer in a network definition.
Definition: NvInfer.h:1896
Weights getScale() const noexcept
Get the scale value.
Definition: NvInfer.h:1953
Weights getPower() const noexcept
Get the power value.
Definition: NvInfer.h:1973
void setScale(Weights scale) noexcept
Set the scale value.
Definition: NvInfer.h:1943
void setPower(Weights power) noexcept
Set the power value.
Definition: NvInfer.h:1963
ScaleMode getMode() const noexcept
Get the scale mode.
Definition: NvInfer.h:1913
void setShift(Weights shift) noexcept
Set the shift value.
Definition: NvInfer.h:1923
void setChannelAxis(int32_t channelAxis) noexcept
Set the channel axis.
Definition: NvInfer.h:2009
Weights getShift() const noexcept
Get the shift value.
Definition: NvInfer.h:1933
virtual ~IScaleLayer() noexcept=default
void setMode(ScaleMode mode) noexcept
Set the scale mode.
Definition: NvInfer.h:1903
int32_t getChannelAxis() const noexcept
Get the channel axis.
Definition: NvInfer.h:1988
A scatter layer in a network definition. Supports several kinds of scattering.
Definition: NvInfer.h:5986
void setMode(ScatterMode mode) noexcept
Set the scatter mode.
Definition: NvInfer.h:5993
apiv::VScatterLayer * mImpl
Definition: NvInfer.h:6027
void setAxis(int32_t axis) noexcept
Set the axis used by ScatterMode::kELEMENTS.
Definition: NvInfer.h:6013
int32_t getAxis() const noexcept
Get the axis.
Definition: NvInfer.h:6021
ScatterMode getMode() const noexcept
Get the scatter mode.
Definition: NvInfer.h:6003
virtual ~IScatterLayer() noexcept=default
Select elements from two data tensors based on a condition tensor.
Definition: NvInfer.h:5032
virtual ~ISelectLayer() noexcept=default
Layer type for getting shape of a tensor.
Definition: NvInfer.h:3525
virtual ~IShapeLayer() noexcept=default
apiv::VShapeLayer * mImpl
Definition: NvInfer.h:3527
Layer type for shuffling data.
Definition: NvInfer.h:3088
apiv::VShuffleLayer * mImpl
Definition: NvInfer.h:3246
void setFirstTranspose(Permutation permutation) noexcept
Set the permutation applied by the first transpose operation.
Definition: NvInfer.h:3099
void setSecondTranspose(Permutation permutation) noexcept
Set the permutation applied by the second transpose operation.
Definition: NvInfer.h:3199
Dims getReshapeDimensions() const noexcept
Get the reshaped dimensions.
Definition: NvInfer.h:3152
void setReshapeDimensions(Dims const &dimensions) noexcept
Set the reshaped dimensions.
Definition: NvInfer.h:3139
Permutation getFirstTranspose() const noexcept
Get the permutation applied by the first transpose operation.
Definition: NvInfer.h:3111
virtual ~IShuffleLayer() noexcept=default
Permutation getSecondTranspose() const noexcept
Get the permutation applied by the second transpose operation.
Definition: NvInfer.h:3211
bool getZeroIsPlaceholder() const noexcept
Get meaning of 0 in reshape dimensions.
Definition: NvInfer.h:3240
void setZeroIsPlaceholder(bool zeroIsPlaceholder) noexcept
Set meaning of 0 in reshape dimensions.
Definition: NvInfer.h:3227
Slices an input tensor into an output tensor based on the offset and strides.
Definition: NvInfer.h:3340
void setStride(Dims const &stride) noexcept
Set the stride for computing the output slice data.
Definition: NvInfer.h:3409
apiv::VSliceLayer * mImpl
Definition: NvInfer.h:3508
virtual ~ISliceLayer() noexcept=default
void setSize(Dims const &size) noexcept
Set the dimensions of the output slice.
Definition: NvInfer.h:3380
void setAxes(Dims const &axes) noexcept
Set the axes for this ISliceLayer.
Definition: NvInfer.h:3487
void setStart(Dims const &start) noexcept
Set the start offset that the slice layer uses to create the output slice.
Definition: NvInfer.h:3351
Dims getStart() const noexcept
Get the start offset for the slice layer.
Definition: NvInfer.h:3366
void setMode(SampleMode mode) noexcept
Set the slice mode.
Definition: NvInfer.h:3434
Dims getSize() const noexcept
Get dimensions of the output slice.
Definition: NvInfer.h:3395
SampleMode getMode() const noexcept
Get the slice mode.
Definition: NvInfer.h:3444
Dims getStride() const noexcept
Get the stride for the output slice.
Definition: NvInfer.h:3424
Dims getAxes() const noexcept
Get the axes for this ISliceLayer.
Definition: NvInfer.h:3502
A Softmax layer in a network definition.
Definition: NvInfer.h:2040
void setAxes(uint32_t axes) noexcept
Set the axis along which softmax is computed. Currently, only one axis can be set.
Definition: NvInfer.h:2062
uint32_t getAxes() const noexcept
Get the axis along which softmax occurs.
Definition: NvInfer.h:2072
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:6562
virtual ~ISqueezeLayer() noexcept=default
apiv::VSqueezeLayer * mImpl
Definition: NvInfer.h:6579
A tensor in a network definition.
Definition: NvInfer.h:186
void setAllowedFormats(TensorFormats formats) noexcept
Set allowed formats for an input or output tensor. By default all formats are allowed....
Definition: NvInfer.h:456
TensorLocation getLocation() const noexcept
Get the storage location of a tensor.
Definition: NvInfer.h:375
void setDimensions(Dims const &dimensions) noexcept
Set the dimensions of a tensor.
Definition: NvInfer.h:234
void resetDynamicRange() noexcept
Undo effect of setDynamicRange.
Definition: NvInfer.h:414
void setName(char const *name) noexcept
Set the tensor name.
Definition: NvInfer.h:203
bool isExecutionTensor() const noexcept
Whether the tensor is an execution tensor.
Definition: NvInfer.h:521
TRT_DEPRECATED bool dynamicRangeIsSet() const noexcept
Query whether dynamic range is set.
Definition: NvInfer.h:406
char const * getName() const noexcept
Get the tensor name.
Definition: NvInfer.h:215
bool isShapeTensor() const noexcept
Whether the tensor is a shape tensor.
Definition: NvInfer.h:500
float getDynamicRangeMax() const noexcept
Get maximum of dynamic range.
Definition: NvInfer.h:434
bool isNetworkInput() const noexcept
Whether the tensor is a network input.
Definition: NvInfer.h:324
TRT_DEPRECATED void setBroadcastAcrossBatch(bool broadcastAcrossBatch) noexcept
Set whether to enable broadcast of tensor across the implicit batch dimension.
Definition: NvInfer.h:349
TRT_DEPRECATED bool setDynamicRange(float min, float max) noexcept
Set dynamic range for the tensor.
Definition: NvInfer.h:316
TRT_DEPRECATED void setType(DataType type) noexcept
Set the data type of a tensor.
Definition: NvInfer.h:284
TRT_DEPRECATED bool getBroadcastAcrossBatch() const noexcept
Check if tensor is broadcast across the implicit batch dimension.
Definition: NvInfer.h:363
bool isNetworkOutput() const noexcept
Whether the tensor is a network output.
Definition: NvInfer.h:332
DataType getType() const noexcept
Get the data type of a tensor.
Definition: NvInfer.h:299
apiv::VTensor * mImpl
Definition: NvInfer.h:568
float getDynamicRangeMin() const noexcept
Get minimum of dynamic range.
Definition: NvInfer.h:424
virtual ~ITensor() noexcept=default
void setDimensionName(int32_t index, char const *name) noexcept
Name a dimension of an input tensor.
Definition: NvInfer.h:547
char const * getDimensionName(int32_t index) const noexcept
Get the name of an input dimension.
Definition: NvInfer.h:562
TRT_DEPRECATED void setLocation(TensorLocation location) noexcept
Set the storage location of a tensor.
Definition: NvInfer.h:394
Dims getDimensions() const noexcept
Get the dimensions of a tensor.
Definition: NvInfer.h:248
TensorFormats getAllowedFormats() const noexcept
Get a bitmask of TensorFormat values that the tensor supports. For a shape tensor,...
Definition: NvInfer.h:469
Class to handle tactic timing info collected from builder.
Definition: NvInfer.h:9702
int64_t queryKeys(TimingCacheKey *keyBuffer, int64_t capacity) const noexcept
Query cache keys from Timing Cache.
Definition: NvInfer.h:9768
bool combine(ITimingCache const &inputCache, bool ignoreMismatch) noexcept
Combine input timing cache into local instance.
Definition: NvInfer.h:9739
TimingCacheValue query(TimingCacheKey const &key) const noexcept
Query value in a cache entry.
Definition: NvInfer.h:9785
virtual ~ITimingCache() noexcept=default
bool update(TimingCacheKey const &key, TimingCacheValue const &value) noexcept
Update values in a cache entry.
Definition: NvInfer.h:9807
apiv::VTimingCache * mImpl
Definition: NvInfer.h:9813
bool reset() noexcept
Empty the timing cache.
Definition: NvInfer.h:9749
Layer that represents a TopK reduction.
Definition: NvInfer.h:3565
void setK(int32_t k) noexcept
Set the static k value for the layer.
Definition: NvInfer.h:3596
void setReduceAxes(uint32_t reduceAxes) noexcept
Set which axes to reduce for the layer.
Definition: NvInfer.h:3620
TopKOperation getOperation() const noexcept
Get the operation for the layer.
Definition: NvInfer.h:3582
apiv::VTopKLayer * mImpl
Definition: NvInfer.h:3679
void setOperation(TopKOperation op) noexcept
Set the operation for the layer.
Definition: NvInfer.h:3572
bool setIndicesType(DataType type) noexcept
Set the indices type for the layer.
Definition: NvInfer.h:3661
int32_t getK() const noexcept
Get the k value for the layer.
Definition: NvInfer.h:3610
uint32_t getReduceAxes() const noexcept
Get the axes to reduce for the layer.
Definition: NvInfer.h:3630
virtual ~ITopKLayer() noexcept=default
DataType getIndicesType() const noexcept
Return the TopK layer indices type.
Definition: NvInfer.h:3673
A layer that represents a trip-count limiter.
Definition: NvInfer.h:4845
TripLimit getTripLimit() const noexcept
Get a trip limiter type.
Definition: NvInfer.h:4850
virtual ~ITripLimitLayer() noexcept=default
Layer that represents an unary operation.
Definition: NvInfer.h:2845
void setOperation(UnaryOperation op) noexcept
Set the unary operation for the layer.
Definition: NvInfer.h:2854
apiv::VUnaryLayer * mImpl
Definition: NvInfer.h:2870
UnaryOperation getOperation() const noexcept
Get the unary operation for the layer.
Definition: NvInfer.h:2864
virtual ~IUnaryLayer() noexcept=default
Layer that represents an unsqueeze operation, which reshapes the input tensor by inserting unit-lengt...
Definition: NvInfer.h:6591
virtual ~IUnsqueezeLayer() noexcept=default
apiv::VUnsqueezeLayer * mImpl
Definition: NvInfer.h:6608
An Interface class for version control.
Definition: NvInferRuntimeBase.h:278
Version information associated with a TRT interface.
Definition: NvInferRuntimeBase.h:243
An array of weights used as a layer parameter.
Definition: NvInferRuntime.h:124
Definition: NvInfer.h:9296
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:9301
virtual ~IAlgorithmSelector() noexcept=default
Definition: NvInferRuntimeBase.h:415
Definition: NvInferRuntime.h:1656
Definition: NvInfer.h:8923
~IInt8EntropyCalibrator2() noexcept override=default
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:8936
InterfaceInfo getInterfaceInfo() const noexcept override
Return version information associated with this interface. Applications must not override this method...
Definition: NvInfer.h:8928
Definition: NvInfer.h:8883
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:8896
InterfaceInfo getInterfaceInfo() const noexcept override
Return version information associated with this interface. Applications must not override this method...
Definition: NvInfer.h:8888
~IInt8EntropyCalibrator() noexcept override=default
Definition: NvInfer.h:9002
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:9015
InterfaceInfo getInterfaceInfo() const noexcept override
Return version information associated with this interface. Applications must not override this method...
Definition: NvInfer.h:9007
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:8963
~IInt8MinMaxCalibrator() noexcept override=default
CalibrationAlgoType getAlgorithm() noexcept override
Definition: NvInfer.h:8976
InterfaceInfo getInterfaceInfo() const noexcept override
Return version information associated with this interface. Applications must not override this method...
Definition: NvInfer.h:8968
Definition: NvInferPluginBase.h:206
Definition: NvInfer.h:10029
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.
Definition: NvInferRuntime.h:666
IBuilder * createInferBuilder(ILogger &logger) noexcept
Create an instance of an IBuilder class.
Definition: NvInfer.h:11479
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:2961
ResizeSelector
The coordinate selector when resize to single pixel output.
Definition: NvInfer.h:4112
@ 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
MemoryPoolType
The type for memory pools used by TensorRT.
Definition: NvInfer.h:9824
ScaleMode
Controls how shift, scale and power are applied in a Scale layer.
Definition: NvInfer.h:1853
@ 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:9407
uint32_t QuantizationFlags
Represents one or more QuantizationFlag values using binary OR operations.
Definition: NvInfer.h:9359
HardwareCompatibilityLevel
Describes requirements of compatibility with GPU architectures other than that of the GPU on which th...
Definition: NvInfer.h:9943
@ kSAME_COMPUTE_CAPABILITY
CumulativeOperation
Enumerates the cumulative operations that may be performed by a Cumulative layer.
Definition: NvInfer.h:6624
BoundingBoxFormat
Representation of bounding box data used for the Boxes input tensor in INMSLayer.
Definition: NvInfer.h:6178
@ 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:9642
constexpr int32_t EnumMax< LayerType >() noexcept
Definition: NvInfer.h:120
constexpr int32_t EnumMax< CalibrationAlgoType >() noexcept
Definition: NvInfer.h:8798
UnaryOperation
Enumerates the unary operations that may be performed by a Unary layer.
Definition: NvInfer.h:2798
@ 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:2907
constexpr int32_t EnumMax< TripLimit >() noexcept
Definition: NvInfer.h:4513
ActivationType
Enumerates the types of activation to perform in an activation layer.
Definition: NvInfer.h:139
@ 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:5093
@ 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:4142
@ kHALF_DOWN
Round half down.
PaddingMode
Enumerates the modes of padding to perform in convolution, deconvolution and pooling layer,...
Definition: NvInfer.h:1031
@ 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:4501
@ 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:11077
PreviewFeature
Define preview features.
Definition: NvInfer.h:9899
@ kRUNTIME_ACTIVATION_RESIZE_10_10
@ kALIASED_PLUGIN_IO_10_03
TilingOptimizationLevel
Define the optimization levels for Tiling.
Definition: NvInfer.h:9996
@ 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:2557
DataType
The type of weights and tensors. The datatypes other than kBOOL, kINT32, and kINT64 are "activation d...
Definition: NvInferRuntimeBase.h:145
uint32_t BuilderFlags
Represents one or more BuilderFlag values using binary OR operations, e.g., 1U << BuilderFlag::kFP16 ...
Definition: NvInfer.h:9439
DeviceType
The device that this layer/network will execute on.
Definition: NvInferRuntime.h:1350
constexpr int32_t EnumMax< ScaleMode >() noexcept
Definition: NvInfer.h:1865
CalibrationAlgoType
Version of calibration algorithm to use.
Definition: NvInfer.h:8785
@ 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.
@ kATTENTION_INPUT
Attention Input.
@ 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.
@ kATTENTION_OUTPUT
Attention Output.
@ 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:9384
SampleMode
Controls how ISliceLayer and IGridSample handle out-of-bounds coordinates.
Definition: NvInfer.h:3256
@ 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:2545
@ 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:131
ProfilingVerbosity
List of verbosity levels of layer information exposed in NVTX annotations and in IEngineInspector.
Definition: NvInferRuntime.h:2973
NetworkDefinitionCreationFlag
List of immutable network properties expressed at network creation time. NetworkDefinitionCreationFla...
Definition: NvInfer.h:11088
@ kPREFER_JIT_PYTHON_PLUGINS
@ kPREFER_AOT_PYTHON_PLUGINS
ElementWiseOperation
Enumerates the binary operations that may be performed by an ElementWise layer.
Definition: NvInfer.h:2455
@ 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:9371
@ kCALIBRATE_BEFORE_FUSION
constexpr int32_t EnumMax< SampleMode >() noexcept
Definition: NvInfer.h:3272
InterpolationMode
Enumerates various modes of interpolation.
Definition: NvInfer.h:4030
@ 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:9449
@ 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.
@ kSPARSE_WEIGHTS
Allow the builder to examine weights and use optimized functions when weights have suitable sparsity.
@ kERROR_ON_TIMING_CACHE_MISS
@ kEDITABLE_TIMING_CACHE
Enable editable timing cache.
@ kPREFER_PRECISION_CONSTRAINTS
@ kDISTRIBUTIVE_INDEPENDENCE
@ 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
@ kREJECT_EMPTY_ALGORITHMS
constexpr int32_t EnumMax< TopKOperation >() noexcept
Definition: NvInfer.h:3548
TENSORRTAPI nvinfer1::IPluginRegistry * getBuilderPluginRegistry(nvinfer1::EngineCapability capability) noexcept
Return the plugin registry for building a Standard engine, or nullptr if no registry exists.
constexpr int32_t EnumMax< MemoryPoolType >() noexcept
Definition: NvInfer.h:9885
TopKOperation
Enumerates the operations that may be performed by a TopK layer.
Definition: NvInfer.h:3537
ReduceOperation
Enumerates the reduce operations that may be performed by a Reduce layer.
Definition: NvInfer.h:2893
constexpr int32_t EnumMax< LoopOutput >() noexcept
Definition: NvInfer.h:4490
constexpr int32_t EnumMax< NetworkDefinitionCreationFlag >() noexcept
Definition: NvInfer.h:11116
TRT_DEPRECATED_API nvinfer1::safe::IPluginRegistry * getBuilderSafePluginRegistry(nvinfer1::EngineCapability capability) noexcept
Return the plugin registry for building a Safety engine, or nullptr if no registry exists.
ScatterMode
Control form of IScatterLayer.
Definition: NvInfer.h:5912
MatrixOperation
Enumerates the operations that may be performed on a tensor by IMatrixMultiplyLayer before multiplica...
Definition: NvInfer.h:3690
@ kTRANSPOSE
Like kNONE, but transpose the matrix dimensions.
ResizeCoordinateTransformation
The resize coordinate transformation function.
Definition: NvInfer.h:4058
constexpr int32_t EnumMax< UnaryOperation >() noexcept
Definition: NvInfer.h:2832
LoopOutput
Enum that describes kinds of loop outputs.
Definition: NvInfer.h:4473
@ 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:6191
constexpr int32_t EnumMax< MatrixOperation >() noexcept
Definition: NvInfer.h:3718
PoolingType
The type of pooling to perform in a pooling layer.
Definition: NvInfer.h:1467
@ 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:10112
constexpr int32_t EnumMax< FillOperation >() noexcept
Definition: NvInfer.h:5124
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:2635
AttentionNormalizationOp
Enumerates the operations that may be performed by the normalization in the attention subgraph.
Definition: NvInfer.h:6759
constexpr int32_t EnumMax< ScatterMode >() noexcept
Definition: NvInfer.h:5923
Represents a permutation of dimensions.
Definition: NvInfer.h:3065
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:9666
Definition: NvInfer.h:9678
uint64_t tacticHash
Hash of the selected tactic.
Definition: NvInfer.h:9680
float timingMSec
Timing of this tactic in milliseconds. Negative numbers and NaN are invalid values.
Definition: NvInfer.h:9682