Skip to content

Latest commit

 

History

History
893 lines (723 loc) · 46.1 KB

CHANGELOG.md

File metadata and controls

893 lines (723 loc) · 46.1 KB
Raw

TensorRT OSS Release Changelog

10.5.0 GA - 2024-09-30

Key Features and Updates:

  • Demo changes
  • Sample changes
    • None
  • Plugin changes
    • Migrated IPluginV2-descendent versions of bertQKVToContextPlugin (1, 2, 3) to newer versions (4, 5, 6 respectively) which implement IPluginV3.
    • Note:
      • The newer versions preserve the attributes and I/O of the corresponding older plugin version
      • The older plugin versions are deprecated and will be removed in a future release
  • Quickstart guide
    • None
  • Parser changes
    • Added support for real-valued STFT operations
    • Improved error handling in IParser

Known issues:

  • Demos:
    • TensorRT engine might not be build successfully when using --fp8 flag on H100 GPUs.

10.4.0 GA - 2024-09-18

Key Features and Updates:

  • Demo changes

    • Added Stable Cascade pipeline.
    • Enabled INT8 and FP8 quantization for Stable Diffusion v1.5, v2.0 and v2.1 pipelines.
    • Enabled FP8 quantization for Stable Diffusion XL pipeline.

  • Sample changes

    • Add a new python sample aliased_io_plugin which demonstrates how in-place updates to plugin inputs can be achieved through I/O aliasing.

  • Plugin changes

    • Migrated IPluginV2-descendent versions (a) of the following plugins to newer versions (b) which implement IPluginV3 (a->b):
      • scatterElementsPlugin (1->2)
      • skipLayerNormPlugin (1->5, 2->6, 3->7, 4->8)
      • embLayerNormPlugin (2->4, 3->5)
      • bertQKVToContextPlugin (1->4, 2->5, 3->6)
    • Note
      • The newer versions preserve the corresponding attributes and I/O of the corresponding older plugin version.
      • The older plugin versions are deprecated and will be removed in a future release.

  • Quickstart guide

  • Parser changes

    • Added support for tensor axes for Pad operations.
    • Added support for BlackmanWindow, HammingWindow, and HannWindow operations.
    • Improved error handling in IParserRefitter.
    • Fixed kernel shape inference in multi-input convolutions.

  • Updated tooling

    • polygraphy-extension-trtexec v0.0.9

10.3.0 GA - 2024-08-02

Key Features and Updates:

  • Demo changes
  • Plugin changes
    • Deprecated Version 1 of ScatterElements plugin. It is superseded by Version 2, which implements the IPluginV3 interface.
  • Quickstart guide
  • Parser changes
    • Added support for tensor axes inputs for Slice node.
    • Updated ScatterElements importer to use Version 2 of ScatterElements plugin, which implements the IPluginV3 interface.
  • Updated tooling
    • Polygraphy v0.49.13

10.2.0 GA - 2024-07-09

Key Features and Updates:

  • Demo changes
  • Plugin changes
    • Version 3 of the InstanceNormalization plugin (InstanceNormalization_TRT) has been added. This version is based on the IPluginV3 interface and is used by the TensorRT ONNX parser when native InstanceNormalization is disabled.
  • Tooling changes
    • Pytorch Quantization development has transitioned to TensorRT Model Optimizer. All developers are encouraged to use TensorRT Model Optimizer to benefit from the latest advancements on quantization and compression.
  • Build containers
    • Updated default cuda versions to 12.5.0.

10.1.0 GA - 2024-06-17

Key Features and Updates:

  • Parser changes
    • Added supportsModelV2 API
    • Added support for DeformConv operation
    • Added support for PluginV3 TensorRT Plugins
    • Marked all IParser and IParserRefitter APIs as noexcept
  • Plugin changes
    • Added version 2 of ROIAlign_TRT plugin, which implements the IPluginV3 plugin interface. When importing an ONNX model with the RoiAlign op, this new version of the plugin will be inserted to the TRT network.
  • Samples changes
  • Updated tooling
    • Polygraphy v0.49.12
    • ONNX-GraphSurgeon v0.5.3

10.0.1 GA - 2024-04-24

Key Features and Updates:

  • Parser changes
    • Added support for building with protobuf-lite.
    • Fixed issue when parsing and refitting models with nested BatchNormalization nodes.
    • Added support for empty inputs in custom plugin nodes.
  • Demo changes
    • The following demos have been removed: Jasper, Tacotron2, HuggingFace Diffusers notebook
  • Updated tooling
    • Polygraphy v0.49.10
    • ONNX-GraphSurgeon v0.5.2
  • Build Containers
    • Updated default cuda versions to 12.4.0.
    • Added Rocky Linux 8 and Rocky Linux 9 build containers

10.0.0 EA - 2024-03-27

Key Features and Updates:

  • Samples changes
    • Added a sample showcasing weight-stripped engines.
    • Added a sample demonstrating the use of custom tactics with IPluginV3.
    • Added a sample to showcase plugins with data-dependent output shapes, using IPluginV3.
  • Parser changes
    • Added a new class IParserRefitter that can be used to refit a TensorRT engine with the weights of an ONNX model.
    • kNATIVE_INSTANCENORM is now set to ON by default.
    • Added support for IPluginV3 interfaces from TensorRT.
    • Added support for INT4 quantization.
    • Added support for the reduction attribute in ScatterElements.
    • Added support for wrap padding mode in Pad
  • Plugin changes
    • A new plugin has been added in compliance with ONNX ScatterElements.
    • The TensorRT plugin library no longer has a load-time link dependency on cuBLAS or cuDNN libraries.
    • All plugins which relied on cuBLAS/cuDNN handles passed through IPluginV2Ext::attachToContext() have moved to use cuBLAS/cuDNN resources initialized by the plugin library itself. This works by dynamically loading the required cuBLAS/cuDNN library. Additionally, plugins which independently initialized their cuBLAS/cuDNN resources have also moved to dynamically loading the required library. If the respective library is not discoverable through the library path(s), these plugins will not work.
    • bertQKVToContextPlugin: Version 2 of this plugin now supports head sizes less than or equal to 32.
    • reorgPlugin: Added a version 2 which implements IPluginV2DynamicExt.
    • disentangledAttentionPlugin: Fixed a kernel bug.
  • Demo changes
    • HuggingFace demos have been removed. For all users using TensorRT to accelerate Large Language Model inference, please use TensorRT-LLM.
  • Updated tooling
    • Polygraphy v0.49.9
    • ONNX-GraphSurgeon v0.5.1
    • TensorRT Engine Explorer v0.1.8
  • Build Containers
    • RedHat/CentOS 7.x are no longer officially supported starting with TensorRT 10.0. The corresponding container has been removed from TensorRT-OSS.

9.3.0 GA - 2024-02-09

Key Features and Updates:

  • Demo changes
    • Faster Text-to-image using SDXL & INT8 quantization using AMMO
  • Updated tooling
    • Polygraphy v0.49.7

9.2.0 GA - 2023-11-27

Key Features and Updates:

  • trtexec enhancement: Added --weightless flag to mark the engine as weightless.
  • Parser changes
    • Added support for Hardmax operator.
    • Changes to a few operator importers to ensure that TensorRT preserves the precision of operations when using strongly typed mode.
  • Plugin changes
    • Explicit INT8 support added to bertQKVToContextPlugin.
    • Various bug fixes.
  • Updated HuggingFace demo to use transformers v4.31.0 and PyTorch v2.1.0.

9.1.0 GA - 2023-10-18

Key Features and Updates:

  • Update the trt_python_plugin sample.
    • Python plugins API reference is part of the offical TRT Python API.
  • Added samples demonstrating the usage of the progress monitor API.
  • Remove dependencies related to python<3.8 in python samples as we no longer support python<3.8 for python samples.
  • Demo changes
    • Added LAMBADA dataset accuracy checks in the HuggingFace demo.
    • Enabled structured sparsity and FP8 quantized batch matrix multiplication(BMM)s in attention in the NeMo demo.
    • Replaced deprecated APIs in the BERT demo.
  • Updated tooling
    • Polygraphy v0.49.1

9.0.1 GA - 2023-09-07

Key Features and Updates:

  • TensorRT plugin autorhing in Python is now supported
  • Updated default CUDA version to 12.2
  • Support for BLIP models, Seq2Seq and Vision2Seq abstractions in HuggingFace demo.
  • demoDiffusion refactoring and SDXL enhancements
  • Additional validation asserts for NV Plugins
  • Updated tooling
    • TensorRT Engine Explorer v0.1.7: graph rendering for TensorRT 9.0 kgen kernels
    • ONNX-GraphSurgeon v0.3.29
    • PyTorch quantization toolkit v2.2.0

9.0.0 EA - 2023-08-06

Key Features and Updates:

  • Added the NeMo demo to demonstrate the performance benefit of using E4M3 FP8 data type with the GPT models trained with the NVIDIA NeMo Toolkit and TransformerEngine.

  • Demo Diffusion updates

    • Added SDXL 1.0 txt2img pipeline
    • Added ControlNet pipeline
    • Huggingface demo updates
      • Added Flan-T5, OPT, BLOOM, BLOOMZ, GPT-Neo, GPT-NeoX, Cerebras-GPT support with accuracy check
      • Refactored code and extracted common utils into Seq2Seq class
      • Optimized shape-changing overhead and achieved a >30% e2e performance gain
      • Added stable KV-cache, beam search and fp16 support for all models
      • Added dynamic batch size TRT inference
      • Added uneven-length multi-batch inference with attention_mask support
      • Added chat command – interactive CLI
      • Upgraded PyTorch and HuggingFace version to support Hopper GPU
      • Updated notebooks with much simplified demo API.

  • Added two new TensorRT samples: sampleProgressMonitor (C++) and simple_progress_reporter (Python) that are examples for using Progress Monitor during engine build.

  • The following plugins were deprecated:

    • BatchedNMS_TRT
    • BatchedNMSDynamic_TRT
    • BatchTilePlugin_TRT
    • Clip_TRT
    • CoordConvAC
    • CropAndResize
    • EfficientNMS_ONNX_TRT
    • CustomGeluPluginDynamic
    • LReLU_TRT
    • NMSDynamic_TRT
    • NMS_TRT
    • Normalize_TRT
    • Proposal
    • SingleStepLSTMPlugin
    • SpecialSlice_TRT
    • Split

  • Ubuntu 18.04 has reached end of life and is no longer supported by TensorRT starting with 9.0, and the corresponding Dockerfile(s) have been removed.

  • Support for aarch64 builds will not be available in this release, and the corresponding Dockerfiles have been removed.

8.6.1 GA - 2023-05-02

TensorRT OSS release corresponding to TensorRT 8.6.1.6 GA release.

Key Features and Updates:

  • Added a new flag --use-cuda-graph to demoDiffusion to improve performance.
  • Optimized GPT2 and T5 HuggingFace demos to use fp16 I/O tensors for fp16 networks.

8.6.0 EA - 2023-03-10

TensorRT OSS release corresponding to TensorRT 8.6.0.12 EA release.

Key Features and Updates:

  • demoDiffusion acceleration is now supported out of the box in TensorRT without requiring plugins.
    • The following plugins have been removed accordingly: GroupNorm, LayerNorm, MultiHeadCrossAttention, MultiHeadFlashAttention, SeqLen2Spatial, and SplitGeLU.
  • Added a new sample called onnx_custom_plugin.

8.5.3 GA - 2023-01-30

TensorRT OSS release corresponding to TensorRT 8.5.3.1 GA release.

Key Features and Updates:

  • Added the following HuggingFace demos: GPT-J-6B, GPT2-XL, and GPT2-Medium
  • Added nvinfer1::plugin namespace
  • Optimized KV Cache performance for T5

8.5.2 GA - 2022-12-12

TensorRT OSS release corresponding to TensorRT 8.5.2.2 GA release.

Key Features and Updates:

22.12 - 2022-12-06

Added

  • Stable Diffusion demo using TensorRT Plugins
  • KV-cache and beam search to GPT2 and T5 demos
  • Perplexity calculation to all HF demos

Changed

  • Updated trex to v0.1.5
  • Increased default workspace size in demoBERT to build BS=128 fp32 engines
  • Use avg_iter=8 and timing cache to make demoBERT perf more stable

Removed

  • None

8.5.1 GA - 2022-11-01

TensorRT OSS release corresponding to TensorRT 8.5.1.7 GA release.

Key Features and Updates:

  • Samples enhancements

    • Added sampleNamedDimensions which works with named dimensions.
    • Updated sampleINT8API and introductory_parser_samples to use ONNX models over Caffe/UFF
    • Removed UFF/Caffe samples including sampleMNIST, end_to_end_tensorflow_mnist, sampleINT8, sampleMNISTAPI, sampleUffMNIST, sampleUffPluginV2Ext, engine_refit_mnist, int8_caffe_mnist, uff_custom_plugin, sampleFasterRCNN, sampleUffFasterRCNN, sampleGoogleNet, sampleSSD, sampleUffSSD, sampleUffMaskRCNN and uff_ssd.

  • Plugin enhancements

    • Added GridAnchorRectPlugin to support rectangular feature maps in gridAnchorPlugin.
    • Added ROIAlignPlugin to support the ONNX operator RoiAlign. The ONNX parser will automatically route ROIAlign ops through the plugin.
    • Added Hopper support for the BERTQKVToContextPlugin plugin.
    • Exposed the use_int8_scale_max attribute in the BERTQKVToContextPlugin plugin to allow users to disable the by-default usage of INT8 scale factors to optimize softmax MAX reduction in versions 2 and 3 of the plugin.

  • ONNX-TensorRT changes

  • Build containers

    • Updated default cuda versions to 11.8.0.

  • Tooling enhancements

22.08 - 2022-08-16

Updated TensorRT version to 8.4.2 - see the TensorRT 8.4.2 release notes for more information

Changed

  • Updated default protobuf version to 3.20.x
  • Updated ONNX-TensorRT submodule version to 22.08 tag
  • Updated sampleIOFormats and sampleAlgorithmSelector to use ONNX models over Caffe

Fixes

  • Fixed missing serialization member in custom ClipPlugin plugin used in uff_custom_plugin sample
  • Fixed various Python import issues

Added

  • Added new DeBERTA demo
  • Added version 2 for disentangledAttentionPlugin to support DeBERTA v2

Removed

  • None

22.07 - 2022-07-21

Added

  • polygraphy-trtexec-plugin tool for Polygraphy
  • Multi-profile support for demoBERT
  • KV cache support for HF BART demo

Changed

  • Updated ONNX-GS to v0.3.20

Removed

  • None

8.4.1 GA - 2022-06-14

TensorRT OSS release corresponding to TensorRT 8.4.1.5 GA release.

Key Features and Updates:

  • Samples enhancements

  • EfficientDet sample

    • Added support for EfficientDet Lite and AdvProp models.
    • Added dynamic batch support.
    • Added mixed precision engine builder.

  • HuggingFace transformer demo

    • Added BART model.
    • Performance speedup of GPT-2 greedy search using GPU implementation.
    • Fixed GPT2 onnx export failure due to 2G file size limitation.
    • Extended Megatron LayerNorm plugins to support larger hidden sizes.
    • Added performance benchmarking mode.
    • Enable tf32 format by default.

  • demoBERT enhancements

    • Add --duration flag to perf benchmarking script.
    • Fixed import of nvinfer_plugins library in demoBERT on Windows.

  • Torch-QAT toolkit

    • quant_bert.py module removed. It is now upstreamed to HuggingFace QDQBERT.
    • Use axis0 as default for deconv.
    • #1939 - Fixed path in classification_flow example.

  • Plugin enhancements

  • Build containers

    • Updated default cuda versions to 11.6.2.
    • CentOS Linux 8 has reached End-of-Life on Dec 31, 2021. The corresponding container has been removed from TensorRT-OSS.
    • Install devtoolset-8 for updated g++ versions in CentOS7 container.

  • Tooling enhancements

  • trtexec enhancements

    • Added --layerPrecisions and --layerOutputTypes flags for specifying layer-wise precision and output type constraints.
    • Added --memPoolSize flag to specify the size of workspace as well as the DLA memory pools via a unified interface. Correspondingly the --workspace flag has been deprecated.
    • "End-To-End Host Latency" metric has been removed. Use the “Host Latency” metric instead. For more information, refer to Benchmarking Network section in the TensorRT Developer Guide.
    • Use enqueueV2() instead of enqueue() when engine has explicit batch dimensions.

22.06 - 2022-06-08

Added

  • None

Changed

  • Disentangled attention (DMHA) plugin refactored
  • ONNX parser updated to 8.2GA

Removed

  • None

22.05 - 2022-05-13

Added

  • Disentangled attention plugin for DeBERTa
  • DMHA (multiscaleDeformableAttnPlugin) plugin for DDETR
  • Performance benchmarking mode to HuggingFace demo

Changed

  • Updated base TensorRT version to 8.2.5.1
  • Updated onnx-graphsurgeon v0.3.19 CHANGELOG
  • fp16 support for pillarScatterPlugin
  • #1939 - Fixed path in quantization classification_flow
  • Fixed GPT2 onnx export failure due to 2G limitation
  • Use axis0 as default for deconv in pytorch-quantization toolkit
  • Updated onnx export script for CoordConvAC sample
  • Install devtoolset-8 for updated g++ version in CentOS7 container

Removed

  • Usage of deprecated TensorRT APIs in samples removed
  • quant_bert.py module removed from pytorch-quantization

22.04 - 2022-04-13

Added

  • TensorRT Engine Explorer v0.1.0 README
  • Detectron 2 Mask R-CNN R50-FPN python sample
  • Model export script for sampleOnnxMnistCoordConvAC

Changed

  • Updated base TensorRT version to 8.2.4.2
  • Updated copyright headers with SPDX identifiers
  • Updated onnx-graphsurgeon v0.3.17 CHANGELOG
  • PyramidROIAlign plugin refactor and bug fixes
  • Fixed MultilevelCropAndResize crashes on Windows
  • #1583 - sublicense ieee/half.h under Apache2
  • Updated demo/BERT performance tables for rel-8.2
  • #1774 Fix python hangs at IndexErrors when TF is imported after TensorRT
  • Various bugfixes in demos - BERT, Tacotron2 and HuggingFace GPT/T5 notebooks
  • Cleaned up sample READMEs

Removed

  • sampleNMT removed from samples

22.03 - 2022-03-23

Added

  • EfficientDet sample enhancements
    • Added support for EfficientDet Lite and AdvProp models.
    • Added dynamic batch support.
    • Added mixed precision engine builder.

Changed

  • Better decoupling of HuggingFace demo tests

22.02 - 2022-02-04

Added

Changed

  • Extend Megatron LayerNorm plugins to support larger hidden sizes
  • Refactored EfficientNMS plugin for TFTRT and added implicit batch mode support
  • Update base TensorRT version to 8.2.3.0
  • GPT-2 greedy search speedup - now runs on GPU
  • Updates to TensorRT developer tools
  • Updated ONNX parser to v8.2.3.0
  • Minor updates and bugfixes
    • Samples: TFOD, GPT-2, demo/BERT
    • Plugins: proposalPlugin, geluPlugin, bertQKVToContextPlugin, batchedNMS

Removed

  • Unused source file(s) in demo/BERT

8.2.1 GA - 2021-11-24

TensorRT OSS release corresponding to TensorRT 8.2.1.8 GA release.

  • Updates since TensorRT 8.2.0 EA release.

  • Please refer to the TensorRT 8.2.1 GA release notes for more information.

  • ONNX parser v8.2.1

    • Removed duplicate constant layer checks that caused some performance regressions
    • Fixed expand dynamic shape calculations
    • Added parser-side checks for Scatter layer support

  • Sample updates

    • Added Tensorflow Object Detection API converter samples, including Single Shot Detector, Faster R-CNN and Mask R-CNN models
    • Multiple enhancements in HuggingFace transformer demos
      • Added multi-batch support
      • Fixed resultant performance regression in batchsize=1
      • Fixed T5 large/T5-3B accuracy issues
      • Added notebooks for T5 and GPT-2
      • Added CPU benchmarking option
    • Deprecated kSTRICT_TYPES (strict type constraints). Equivalent behaviour now achieved by setting PREFER_PRECISION_CONSTRAINTS, DIRECT_IO, and REJECT_EMPTY_ALGORITHMS
    • Removed sampleMovieLens
    • Renamed sampleReformatFreeIO to sampleIOFormats
    • Add idleTime option for samples to control qps
    • Specify default value for precisionConstraints
    • Fixed reporting of TensorRT build version in trtexec
    • Fixed combineDescriptions typo in trtexec/tracer.py
    • Fixed usages of of kDIRECT_IO

  • Plugin updates

    • EfficientNMS plugin support extended to TF-TRT, and for clang builds.
    • Sanitize header definitions for BERT fused MHA plugin
    • Separate C++ and cu files in splitPlugin to avoid PTX generation (required for CUDA enhanced compatibility support)
    • Enable C++14 build for plugins

  • ONNX tooling updates

  • Build and container fixes

    • Add SM86 target to default GPU_ARCHS for platforms with cuda-11.1+
    • Remove deprecated SM_35 and add SM_60 to default GPU_ARCHS
    • Skip CUB builds for cuda 11.0+ #1455
    • Fixed cuda-10.2 container build failures in Ubuntu 20.04
    • Add native ARM server build container
    • Install devtoolset-8 for updated g++ version in CentOS7
    • Added a note on supporting c++14 builds for CentOS7
    • Fixed docker build for large UIDs #1373
    • Updated README instructions for Jetpack builds

  • demo enhancements

    • Updated Tacotron2 instructions and add CPU benchmarking
    • Fixed issues in demoBERT python notebook

  • Documentation updates

    • Updated Python documentation for add_reduce, add_top_k, and ISoftMaxLayer
    • Renamed default GitHub branch to main and updated hyperlinks

8.2.0 EA - 2021-10-05

Added

  • Demo applications showcasing TensorRT inference of HuggingFace Transformers.
    • Support is currently extended to GPT-2 and T5 models.
  • Added support for the following ONNX operators:
    • Einsum
    • IsNan
    • GatherND
    • Scatter
    • ScatterElements
    • ScatterND
    • Sign
    • Round
  • Added support for building TensorRT Python API on Windows.

Updated

  • Notable API updates in TensorRT 8.2.0.6 EA release. See TensorRT Developer Guide for details.
    • Added three new APIs, IExecutionContext: getEnqueueEmitsProfile(), setEnqueueEmitsProfile(), and reportToProfiler() which can be used to collect layer profiling info when the inference is launched as a CUDA graph.
    • Eliminated the global logger; each Runtime, Builder or Refitter now has its own logger.
    • Added new operators: IAssertionLayer, IConditionLayer, IEinsumLayer, IIfConditionalBoundaryLayer, IIfConditionalOutputLayer, IIfConditionalInputLayer, and IScatterLayer.
    • Added new IGatherLayer modes: kELEMENT and kND
    • Added new ISliceLayer modes: kFILL, kCLAMP, and kREFLECT
    • Added new IUnaryLayer operators: kSIGN and kROUND
    • Added new runtime class IEngineInspector that can be used to inspect the detailed information of an engine, including the layer parameters, the chosen tactics, the precision used, etc.
    • ProfilingVerbosity enums have been updated to show their functionality more explicitly.
  • Updated TensorRT OSS container defaults to cuda 11.4
  • CMake to target C++14 builds.
  • Updated following ONNX operators:
    • Gather and GatherElements implementations to natively support negative indices
    • Pad layer to support ND padding, along with edge and reflect padding mode support
    • If layer with general performance improvements.

Removed

  • Removed sampleMLP.
  • Several flags of trtexec have been deprecated:
    • --explicitBatch flag has been deprecated and has no effect. When the input model is in UFF or in Caffe prototxt format, the implicit batch dimension mode is used automatically; when the input model is in ONNX format, the explicit batch mode is used automatically.
    • --explicitPrecision flag has been deprecated and has no effect. When the input ONNX model contains Quantization/Dequantization nodes, TensorRT automatically uses explicit precision mode.
    • --nvtxMode=[verbose|default|none] has been deprecated in favor of --profilingVerbosity=[detailed|layer_names_only|none] to show its functionality more explicitly.

21.10 - 2021-10-05

Added

  • Benchmark script for demoBERT-Megatron
  • Dynamic Input Shape support for EfficientNMS plugin
  • Support empty dimensions in ONNX
  • INT32 and dynamic clips through elementwise in ONNX parser

Changed

  • Bump TensorRT version to 8.0.3.4
  • Use static shape for only single batch single sequence input in demo/BERT
  • Revert to using native FC layer in demo/BERT and FCPlugin only on older GPUs.
  • Update demo/Tacotron2 for TensorRT 8.0
  • Updates to TensorRT developer tools
    • Polygraphy v0.33.0
      • Added various examples, a CLI User Guide and how-to guides.
      • Added experimental support for DLA.
      • Added a data to-input tool that can combine inputs/outputs created by --save-inputs/--save-outputs.
      • Added a PluginRefRunner which provides CPU reference implementations for TensorRT plugins
      • Made several performance improvements in the Polygraphy CUDA wrapper.
      • Removed the to-json tool which was used to convert Pickled data generated by Polygraphy 0.26.1 and older to JSON.
    • Bugfixes and documentation updates in pytorch-quantization toolkit.
  • Bumped up package versions: tensorflow-gpu 2.5.1, pillow 8.3.2
  • ONNX parser enhancements and bugfixes
    • Update ONNX submodule to v1.8.0
    • Update convDeconvMultiInput function to properly handle deconvs
    • Update RNN documentation
    • Update QDQ axis assertion
    • Fix bidirectional activation alpha and beta values
    • Fix opset10 Resize
    • Fix shape tensor unsqueeze
    • Mark BOOL tiles as unsupported
    • Remove unnecessary shape tensor checks

Removed

  • N/A

21.09 - 2021-09-22

Added

  • Add ONNX2TRT_VERSION overwrite in CMake.

Changed

  • Updates to TensorRT developer tools
  • Fix assertion in EfficientNMSPlugin

Removed

  • N/A

21.08 - 2021-08-05

Added

Changed

  • Updated samples and plugins directory structure
  • Updates to TensorRT developer tools
  • README fix to update build command for native aarch64 builds.

Removed

  • N/A

21.07 - 2021-07-21

Identical to the TensorRT-OSS 8.0.1 Release.

8.0.1 - 2021-07-02

Added

  • Added support for the following ONNX operators: Celu, CumSum, EyeLike, GatherElements, GlobalLpPool, GreaterOrEqual, LessOrEqual, LpNormalization, LpPool, ReverseSequence, and SoftmaxCrossEntropyLoss details.
  • Rehauled Resize ONNX operator, now fully supporting the following modes:
    • Coordinate Transformation modes: half_pixel, pytorch_half_pixel, tf_half_pixel_for_nn, asymmetric, and align_corners.
    • Modes: nearest, linear.
    • Nearest Modes: floor, ceil, round_prefer_floor, round_prefer_ceil.
  • Added support for multi-input ONNX ConvTranpose operator.
  • Added support for 3D spatial dimensions in ONNX InstanceNormalization.
  • Added support for generic 2D padding in ONNX.
  • ONNX QuantizeLinear and DequantizeLinear operators leverage IQuantizeLayer and IDequantizeLayer.
    • Added support for tensor scales.
    • Added support for per-axis quantization.
  • Added EfficientNMS_TRT, EfficientNMS_ONNX_TRT plugins and experimental support for ONNX NonMaxSuppression operator.
  • Added ScatterND plugin.
  • Added TensorRT QuickStart Guide.
  • Added new samples: engine_refit_onnx_bidaf builds an engine from ONNX BiDAF model and refits engine with new weights, efficientdet and efficientnet samples for demonstrating Object Detection using TensorRT.
  • Added support for Ubuntu20.04 and RedHat/CentOS 8.3.
  • Added Python 3.9 support.

Changed

  • Update Polygraphy to v0.30.3.
  • Update ONNX-GraphSurgeon to v0.3.10.
  • Update Pytorch Quantization toolkit to v2.1.0.
  • Notable TensorRT API updates
    • TensorRT now declares API’s with the noexcept keyword. All TensorRT classes that an application inherits from (such as IPluginV2) must guarantee that methods called by TensorRT do not throw uncaught exceptions, or the behavior is undefined.
    • Destructors for classes with destroy() methods were previously protected. They are now public, enabling use of smart pointers for these classes. The destroy() methods are deprecated.
  • Moved RefitMap API from ONNX parser to core TensorRT.
  • Various bugfixes for plugins, samples and ONNX parser.
  • Port demoBERT to tensorflow2 and update UFF samples to leverage nvidia-tensorflow1 container.

Removed

  • IPlugin and IPluginFactory interfaces were deprecated in TensorRT 6.0 and have been removed in TensorRT 8.0. We recommend that you write new plugins or refactor existing ones to target the IPluginV2DynamicExt and IPluginV2IOExt interfaces. For more information, refer to Migrating Plugins From TensorRT 6.x Or 7.x To TensorRT 8.x.x.
    • For plugins based on IPluginV2DynamicExt and IPluginV2IOExt, certain methods with legacy function signatures (derived from IPluginV2 and IPluginV2Ext base classes) which were deprecated and marked for removal in TensorRT 8.0 will no longer be available.
  • Removed samplePlugin since it showcased IPluginExt interface, which is no longer supported in TensorRT 8.0.
  • Removed sampleMovieLens and sampleMovieLensMPS.
  • Removed Dockerfile for Ubuntu 16.04. TensorRT 8.0 debians for Ubuntu 16.04 require python 3.5 while minimum required python version for TensorRT OSS is 3.6.
  • Removed support for PowerPC builds, consistent with TensorRT GA releases.

Notes

  • We had deprecated the Caffe Parser and UFF Parser in TensorRT 7.0. They are still tested and functional in TensorRT 8.0, however, we plan to remove the support in a future release. Ensure you migrate your workflow to use tf2onnx, keras2onnx or TensorFlow-TensorRT (TF-TRT).
  • Refer to TensorRT 8.0.1 GA Release Notes for additional details

21.06 - 2021-06-23

Added

  • Add switch for batch-agnostic mode in NMS plugin
  • Add missing model.py in uff_custom_plugin sample

Changed

  • Update to Polygraphy v0.29.2
  • Update to ONNX-GraphSurgeon v0.3.9
  • Fix numerical errors for float type in NMS/batchedNMS plugins
  • Update demoBERT input dimensions to match Triton requirement #1051
  • Optimize TLT MaskRCNN plugins:
    • enable fp16 precision in multilevelCropAndResizePlugin and multilevelProposeROIPlugin
    • Algorithms optimization for NMS kernels and ROIAlign kernel
    • Fix invalid cuda config issue when bs is larger than 32
    • Fix issues found on Jetson NANO

Removed

  • Removed fcplugin from demoBERT to improve latency

21.05 - 2021-05-20

Added

  • Extended support for ONNX operator InstanceNormalization to 5D tensors
  • Support negative indices in ONNX Gather operator
  • Add support for importing ONNX double-typed weights as float
  • ONNX-GraphSurgeon (v0.3.7) support for models with externally stored weights

Changed

  • Update ONNX-TensorRT to 21.05
  • Relicense ONNX-TensorRT under Apache2
  • demoBERT builder fixes for multi-batch
  • Speedup demoBERT build using global timing cache and disable cuDNN tactics
  • Standardize python package versions across OSS samples
  • Bugfixes in multilevelProposeROI and bertQKV plugin
  • Fix memleaks in samples logger

21.04 - 2021-04-12

Added

  • SM86 kernels for BERT MHA plugin
  • Added opset13 support for SoftMax, LogSoftmax, Squeeze, and Unsqueeze.
  • Added support for the EyeLike and GatherElements operators.

Changed

  • Updated TensorRT version to v7.2.3.4.
  • Update to ONNX-TensorRT 21.03
  • ONNX-GraphSurgeon (v0.3.4) - updates fold_constants to correctly exit early.
  • Set default CUDA_INSTALL_DIR #798
  • Plugin bugfixes, qkv kernels for sm86
  • Fixed GroupNorm CMakeFile for cu sources #1083
  • Permit groupadd with non-unique GID in build containers #1091
  • Avoid reinterpret_cast #146
  • Clang-format plugins and samples
  • Avoid arithmetic on void pointer in multilevelProposeROIPlugin.cpp #1028
  • Update BERT plugin documentation.

Removed

  • Removes extra terminate call in InstanceNorm

21.03 - 2021-03-09

Added

  • Optimized FP16 NMS/batchedNMS plugins with n-bit radix sort and based on IPluginV2DynamicExt
  • ProposalDynamic and CropAndResizeDynamic plugins based on IPluginV2DynamicExt

Changed

Removed

  • N/A

21.02 - 2021-02-01

Added

Changed

Removed

  • N/A

20.12 - 2020-12-18

Added

  • Add configurable input size for TLT MaskRCNN Plugin

Changed

  • Update symbol export map for plugins
  • Correctly use channel dimension when creating Prelu node
  • Fix Jetson cross compilation CMakefile

Removed

  • N/A

20.11 - 2020-11-20

Added

Changed

Removed

  • N/A

20.10 - 2020-10-22

Added

  • Polygraphy v0.20.13 - Deep Learning Inference Prototyping and Debugging Toolkit
  • PyTorch-Quantization Toolkit v2.0.0
  • Updated BERT plugins for variable sequence length inputs
  • Optimized kernels for sequence lengths of 64 and 96 added
  • Added Tacotron2 + Waveglow TTS demo #677
  • Re-enable GridAnchorRect_TRT plugin with rectangular feature maps #679
  • Update batchedNMS plugin to IPluginV2DynamicExt interface #738
  • Support 3D inputs in InstanceNormalization plugin #745
  • Added this CHANGELOG.md

Changed

  • ONNX GraphSurgeon - v0.2.7 with bugfixes, new examples.
  • demo/BERT bugfixes for Jetson Xavier
  • Updated build Dockerfile to cuda-11.1
  • Updated ClangFormat style specification according to TensorRT coding guidelines

Removed

  • N/A

7.2.1 - 2020-10-20

Added

  • Polygraphy v0.20.13 - Deep Learning Inference Prototyping and Debugging Toolkit
  • PyTorch-Quantization Toolkit v2.0.0
  • Updated BERT plugins for variable sequence length inputs
    • Optimized kernels for sequence lengths of 64 and 96 added
  • Added Tacotron2 + Waveglow TTS demo #677
  • Re-enable GridAnchorRect_TRT plugin with rectangular feature maps #679
  • Update batchedNMS plugin to IPluginV2DynamicExt interface #738
  • Support 3D inputs in InstanceNormalization plugin #745
  • Added this CHANGELOG.md

Changed

  • ONNX GraphSurgeon - v0.2.7 with bugfixes, new examples.
  • demo/BERT bugfixes for Jetson Xavier
  • Updated build Dockerfile to cuda-11.1
  • Updated ClangFormat style specification according to TensorRT coding guidelines

Removed

  • N/A