TensorRT OSS 21.02 release

Signed-off-by: Rajeev Rao <rajeevrao@nvidia.com>

Author: rajeevsrao

.github/ISSUE_TEMPLATE/bug_report.md

@@ -1,6 +1,6 @@
 ---
-name: Bug report
-about: Create a report to help us improve
+name: TensorRT OSS Bug Report
+about: Report any bugs to help us improve TensorRT.
 title: ''
 labels: ''
 assignees: ''
@@ -15,20 +15,20 @@ assignees: ''
 ## Environment
 
 **TensorRT Version**: 
-**GPU Type**: 
-**Nvidia Driver Version**: 
+**NVIDIA GPU**: 
+**NVIDIA Driver Version**: 
 **CUDA Version**: 
 **CUDNN Version**: 
-**Operating System + Version**: 
+**Operating System**: 
 **Python Version (if applicable)**: 
-**TensorFlow Version (if applicable)**: 
+**Tensorflow Version (if applicable)**: 
 **PyTorch Version (if applicable)**: 
-**Baremetal or Container (if container which image + tag)**: 
+**Baremetal or Container (if so, version)**: 
 
 
 ## Relevant Files
 
-<!-- Please include links to any models, data, files, or scripts necessary to reproduce your issue. (Github repo, Google Drive, Dropbox, etc.) -->
+<!-- Please include links to any models, data, files, or scripts necessary to reproduce your issue. (Github repo, Google Drive/Dropbox, etc.) -->
 
 
 ## Steps To Reproduce

CHANGELOG.md

@@ -1,5 +1,26 @@
 # TensorRT OSS Release Changelog
 
+## [21.02](https://github.com/NVIDIA/TensorRT/releases/tag/21.02) - 2021-02-01
+### Added
+- [TensorRT Python API bindings](python)
+- [TensorRT Python samples](samples/python)
+- FP16 support to batchedNMSPlugin [#1002](https://github.com/NVIDIA/TensorRT/pull/1002)
+- Configurable input size for TLT MaskRCNN Plugin [#986](https://github.com/NVIDIA/TensorRT/pull/986)
+
+### Changed
+- TensorRT version updated to 7.2.2.3
+- [ONNX-TensorRT v21.02 update](https://github.com/onnx/onnx-tensorrt/blob/master/docs/Changelog.md#2102-container-release---2021-01-22)
+- [Polygraphy v0.21.1 update](tools/Polygraphy/CHANGELOG.md#v0211-2021-01-12)
+- [PyTorch-Quantization Toolkit](tools/pytorch-quantization) v2.1.0 update
+  - Documentation update, ONNX opset 13 support, ResNet example
+- [ONNX-GraphSurgeon v0.28 update](tools/onnx-graphsurgeon/CHANGELOG.md#v028-2020-10-08)
+- [demoBERT builder](demo/BERT) updated to work with Tensorflow2 (in compatibility mode)
+- Refactor [Dockerfiles](docker) for OSS container
+
+### Removed
+- N/A
+
+
 ## [20.12](https://github.com/NVIDIA/TensorRT/releases/tag/20.12) - 2020-12-18
 ### Added
 - Add configurable input size for TLT MaskRCNN Plugin

LICENSE

@@ -176,7 +176,7 @@
 
    END OF TERMS AND CONDITIONS
 
-   Copyright 2020 NVIDIA Corporation
+   Copyright 2021 NVIDIA Corporation
 
    Licensed under the Apache License, Version 2.0 (the "License");
    you may not use this file except in compliance with the License.
@@ -193,6 +193,38 @@
 
    PORTIONS LICENSED AS FOLLOWS
 
+   > tools/pytorch-quantization/examples/torchvision/models/classification/resnet.py
+
+     BSD 3-Clause License
+
+     Copyright (c) Soumith Chintala 2016,
+     All rights reserved.
+
+     Redistribution and use in source and binary forms, with or without
+     modification, are permitted provided that the following conditions are met:
+
+     * Redistributions of source code must retain the above copyright notice, this
+       list of conditions and the following disclaimer.
+
+     * Redistributions in binary form must reproduce the above copyright notice,
+       this list of conditions and the following disclaimer in the documentation
+       and/or other materials provided with the distribution.
+
+     * Neither the name of the copyright holder nor the names of its
+       contributors may be used to endorse or promote products derived from
+       this software without specific prior written permission.
+
+     THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+     AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+     IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+     DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
+     FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+     DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+     SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
+     CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
+     OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+     OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+
    > samples/common/windows/getopt.c
 
      Copyright (c) 2002 Todd C. Miller <Todd.Miller@courtesan.com>

NOTICE

@@ -0,0 +1,8 @@
+TensorRT Open Source Software
+Copyright (c) 2021, NVIDIA CORPORATION.
+
+This product includes software developed at
+NVIDIA CORPORATION (https://www.nvidia.com/).
+
+This software contains code derived by Soumith Chintala.
+BSD 3-Clause License (https://github.com/pytorch/vision/blob/master/LICENSE)

README.md

@@ -12,7 +12,7 @@ This repository contains the Open Source Software (OSS) components of NVIDIA Ten
 To build the TensorRT-OSS components, you will first need the following software packages.
 
 **TensorRT GA build**
-* [TensorRT](https://developer.nvidia.com/nvidia-tensorrt-download) v7.2.1
+* [TensorRT](https://developer.nvidia.com/nvidia-tensorrt-download) v7.2.2
     - See [Downloading TensorRT Builds](#downloading-tensorrt-builds) for details
 
 **System Packages**
@@ -56,69 +56,55 @@ To build the TensorRT-OSS components, you will first need the following software
 	git clone -b master https://github.com/nvidia/TensorRT TensorRT
 	cd TensorRT
 	git submodule update --init --recursive
-	export TRT_SOURCE=`pwd`
 	```
 	**On Windows: Powershell**
 	```powershell
 	git clone -b master https://github.com/nvidia/TensorRT TensorRT
 	cd TensorRT
 	git submodule update --init --recursive
-	$Env:TRT_SOURCE = $(Get-Location)
 	```
 
-2. #### Download TensorRT GA
-	To build TensorRT OSS, obtain the corresponding TensorRT GA build from [NVIDIA Developer Zone](https://developer.nvidia.com/nvidia-tensorrt-download).
+2. #### Specify the TensorRT Release build
 
-	**Example: Ubuntu 18.04 on x86-64 with cuda-11.1**
+    If using NVIDIA build containers, TensorRT is preinstalled under `/usr/lib/x86_64-linux-gnu`.
 
-	Download and extract the latest *TensorRT 7.2.1 GA package for Ubuntu 18.04 and CUDA 11.1*
-	```bash
-	cd ~/Downloads
-	tar -xvzf TensorRT-7.2.1.6.Ubuntu-18.04.x86_64-gnu.cuda-11.1.cudnn8.0.tar.gz
-	export TRT_RELEASE=`pwd`/TensorRT-7.2.1.6
-	```
-	**Example: Ubuntu 18.04 on PowerPC with cuda-11.0**
+    Else download and extract the TensorRT build from [NVIDIA Developer Zone](https://developer.nvidia.com/nvidia-tensorrt-download).
 
-	Download and extract the latest *TensorRT 7.2.1 GA package for Ubuntu 18.04 and CUDA 11.0*
-	```bash
-	cd ~/Downloads
-	tar -xvzf TensorRT-7.2.1.6.Ubuntu-18.04.powerpc64le-gnu.cuda-11.0.cudnn8.0.tar.gz
-	export TRT_RELEASE=`pwd`/TensorRT-7.2.1.6
-	```
-	**Example: CentOS/RedHat 7 on x86-64 with cuda-11.0**
+    **Example: Ubuntu 18.04 on x86-64 with cuda-11.1**
 
-	Download and extract the *TensorRT 7.2.1 GA for CentOS/RedHat 7 and CUDA 11.0 tar package*
-	```bash
-	cd ~/Downloads
-	tar -xvzf TensorRT-7.2.1.6.CentOS-7.6.x86_64-gnu.cuda-11.0.cudnn8.0.tar.gz
-	export TRT_RELEASE=`pwd`/TensorRT-7.2.1.6
-	```
-	**Example: Ubuntu18.04 Cross-Compile for QNX with cuda-10.2**
+    ```bash
+    cd ~/Downloads
+    tar -xvzf TensorRT-7.2.2.3.Ubuntu-18.04.x86_64-gnu.cuda-11.1.cudnn8.0.tar.gz
+    export TRT_LIBPATH=`pwd`/TensorRT-7.2.2.3
+    ```
 
-	Download and extract the *TensorRT 7.2.1 GA for QNX and CUDA 10.2 tar package*
-	```bash
-	cd ~/Downloads
-	tar -xvzf TensorRT-7.2.1.6.Ubuntu-18.04.aarch64-qnx.cuda-10.2.cudnn7.6.tar.gz
-	export TRT_RELEASE=`pwd`/TensorRT-7.2.1.6
-	export QNX_HOST=/<path-to-qnx-toolchain>/host/linux/x86_64
-	export QNX_TARGET=/<path-to-qnx-toolchain>/target/qnx7
-	```
-	**Example: Windows on x86-64 with cuda-11.0**
+    **Example: Ubuntu18.04 Cross-Compile for QNX with cuda-10.2**
+
+    ```bash
+    cd ~/Downloads
+    tar -xvzf TensorRT-7.2.2.3.Ubuntu-18.04.aarch64-qnx.cuda-10.2.cudnn7.6.tar.gz
+    export TRT_LIBPATH=`pwd`/TensorRT-7.2.2.3
+    export QNX_HOST=/<path-to-qnx-toolchain>/host/linux/x86_64
+    export QNX_TARGET=/<path-to-qnx-toolchain>/target/qnx7
+    ```
+
+    **Example: Windows on x86-64 with cuda-11.0**
+
+    ```powershell
+    cd ~\Downloads
+    Expand-Archive .\TensorRT-7.2.2.3.Windows10.x86_64.cuda-11.0.cudnn8.0.zip
+    $Env:TRT_LIBPATH = '$(Get-Location)\TensorRT-7.2.2.3'
+    $Env:PATH += 'C:\Program Files (x86)\Microsoft Visual Studio\2017\Professional\MSBuild\15.0\Bin\'
+    ```
 
-	Download and extract the *TensorRT 7.2.1 GA for Windows and CUDA 11.0 zip package* and add *msbuild* to *PATH*
-	```powershell
-	cd ~\Downloads
-	Expand-Archive .\TensorRT-7.2.1.6.Windows10.x86_64.cuda-11.0.cudnn8.0.zip
-	$Env:TRT_RELEASE = '$(Get-Location)\TensorRT-7.2.1.6'
-	$Env:PATH += 'C:\Program Files (x86)\Microsoft Visual Studio\2017\Professional\MSBuild\15.0\Bin\'
-	```
 
 3. #### (Optional) JetPack SDK for Jetson builds
     Using the JetPack SDK manager, download the host components. Steps:
     1. Download and launch the SDK manager. Login with your developer account.
     2. Select the  platform and target OS  (example: Jetson AGX Xavier, `Linux Jetpack 4.4`), and click Continue.
     3. Under `Download & Install Options` change the download folder and select `Download now, Install later`. Agree to the license terms and click Continue.
-    4. Move the extracted files into the `$TRT_SOURCE/docker/jetpack_files` folder.
+    4. Move the extracted files into the `<TensorRT-OSS>/docker/jetpack_files` folder.
+
 
 ## Setting Up The Build Environment
 
@@ -129,25 +115,25 @@ For native builds, install the [prerequisite](#prerequisites) *System Packages*.
 
     **Example: Ubuntu 18.04 on x86-64 with cuda-11.1**
     ```bash
-    ./docker/build.sh --file docker/ubuntu.Dockerfile --tag tensorrt-ubuntu --os 18.04 --cuda 11.1
+    ./docker/build.sh --file docker/ubuntu-18.04.Dockerfile --tag tensorrt-ubuntu-1804 --cuda 11.1
     ```
-    **Example: Ubuntu 18.04 on PowerPC with cuda-11.0**
+    **Example: Ubuntu 18.04 cross-compile for PowerPC with cuda-11.0**
     ```bash
-    ./docker/build.sh --file docker/ubuntu-cross-ppc64le.Dockerfile --tag tensorrt-ubuntu-ppc --os 18.04 --cuda 11.0
+    ./docker/build.sh --file docker/ubuntu-cross-ppc64le.Dockerfile --tag tensorrt-ubuntu-ppc --cuda 11.0
     ```
     **Example: CentOS/RedHat 7 on x86-64 with cuda-11.0**
     ```bash
-    ./docker/build.sh --file docker/centos.Dockerfile --tag tensorrt-centos --os 7 --cuda 11.0
+    ./docker/build.sh --file docker/centos-7.Dockerfile --tag tensorrt-centos --cuda 11.0
     ```
-    **Example: Ubuntu 18.04 Cross-Compile for Jetson (arm64) with cuda-10.2 (JetPack)**
+    **Example: Ubuntu 18.04 cross-compile for Jetson (arm64) with cuda-10.2 (JetPack SDK)**
     ```bash
-    ./docker/build.sh --file docker/ubuntu-cross-aarch64.Dockerfile --tag tensorrt-cross-jetpack --os 18.04 --cuda 10.2
+    ./docker/build.sh --file docker/ubuntu-cross-aarch64.Dockerfile --tag tensorrt-cross-jetpack --cuda 10.2
     ```
 
 2. #### Launch the TensorRT-OSS build container.
     **Example: Ubuntu 18.04 build container**
 	```bash
-	./docker/launch.sh --tag tensorrt-ubuntu --gpus all --release $TRT_RELEASE --source $TRT_SOURCE
+	./docker/launch.sh --tag tensorrt-ubuntu-1804 --gpus all
 	```
 	> NOTE:
 	1. Use the tag corresponding to the build container you generated in 
@@ -158,37 +144,37 @@ For native builds, install the [prerequisite](#prerequisites) *System Packages*.
 
    **Example: Linux (x86-64) build with default cuda-11.1**
 	```bash
-	cd $TRT_SOURCE
+	cd $TRT_OSSPATH
 	mkdir -p build && cd build
-	cmake .. -DTRT_LIB_DIR=$TRT_RELEASE/lib -DTRT_OUT_DIR=`pwd`/out
+	cmake .. -DTRT_LIB_DIR=$TRT_LIBPATH -DTRT_OUT_DIR=`pwd`/out
 	make -j$(nproc)
 	```
     **Example: Native build on Jetson (arm64) with cuda-10.2**
     ```bash
-    cd $TRT_SOURCE
+    cd $TRT_OSSPATH
     mkdir -p build && cd build
-    cmake .. -DTRT_LIB_DIR=$TRT_RELEASE/lib -DTRT_OUT_DIR=`pwd`/out -DTRT_PLATFORM_ID=aarch64 -DCUDA_VERSION=10.2
+    cmake .. -DTRT_LIB_DIR=$TRT_LIBPATH -DTRT_OUT_DIR=`pwd`/out -DTRT_PLATFORM_ID=aarch64 -DCUDA_VERSION=10.2
     make -j$(nproc)
     ```
     **Example: Ubuntu 18.04 Cross-Compile for Jetson (arm64) with cuda-10.2 (JetPack)**
 	```bash
-	cd $TRT_SOURCE
+	cd $TRT_OSSPATH
 	mkdir -p build && cd build
-	cmake .. -DTRT_LIB_DIR=$TRT_RELEASE/lib -DTRT_OUT_DIR=`pwd`/out -DCMAKE_TOOLCHAIN_FILE=$TRT_SOURCE/cmake/toolchains/cmake_aarch64.toolchain -DCUDA_VERSION=10.2
+	cmake .. -DTRT_LIB_DIR=$TRT_LIBPATH -DTRT_OUT_DIR=`pwd`/out -DCMAKE_TOOLCHAIN_FILE=$TRT_OSSPATH/cmake/toolchains/cmake_aarch64.toolchain -DCUDA_VERSION=10.2
 	make -j$(nproc)
 	```
     **Example: Cross-Compile for QNX with cuda-10.2**
 	```bash
-	cd $TRT_SOURCE
+	cd $TRT_OSSPATH
 	mkdir -p build && cd build
-	cmake .. -DTRT_LIB_DIR=$TRT_RELEASE/lib -DTRT_OUT_DIR=`pwd`/out -DCMAKE_TOOLCHAIN_FILE=$TRT_SOURCE/cmake/toolchains/cmake_qnx.toolchain -DCUDA_VERSION=10.2
+	cmake .. -DTRT_LIB_DIR=$TRT_LIBPATH -DTRT_OUT_DIR=`pwd`/out -DCMAKE_TOOLCHAIN_FILE=$TRT_OSSPATH/cmake/toolchains/cmake_qnx.toolchain -DCUDA_VERSION=10.2
 	make -j$(nproc)
 	```
     **Example: Windows (x86-64) build in Powershell**
 	```powershell
-	cd $Env:TRT_SOURCE
+	cd $Env:TRT_OSSPATH
 	mkdir -p build ; cd build
-	cmake .. -DTRT_LIB_DIR=$Env:TRT_RELEASE\lib -DTRT_OUT_DIR='$(Get-Location)\out' -DCMAKE_TOOLCHAIN_FILE=..\cmake\toolchains\cmake_x64_win.toolchain
+	cmake .. -DTRT_LIB_DIR=$Env:TRT_LIBPATH -DTRT_OUT_DIR='$(Get-Location)\out' -DCMAKE_TOOLCHAIN_FILE=..\cmake\toolchains\cmake_x64_win.toolchain
 	msbuild ALL_BUILD.vcxproj
 	```
 	> NOTE:
@@ -215,15 +201,6 @@ For native builds, install the [prerequisite](#prerequisites) *System Packages*.
         - Multiple SMs: `-DGPU_ARCHS="80 75"`
 	- `TRT_PLATFORM_ID`: Bare-metal build (unlike containerized cross-compilation) on non Linux/x86 platforms must explicitly specify the target platform. Currently supported options: `x86_64` (default), `aarch64`
 
-#### (Optional) Install TensorRT python bindings
-
-* The TensorRT python API bindings must be installed for running TensorRT python applications
-
-    **Example: install TensorRT wheel for python 3.6**
-    ```bash
-    pip3 install $TRT_RELEASE/python/tensorrt-7.2.1.6-cp36-none-linux_x86_64.whl
-    ```
-
 # References
 
 ## TensorRT Resources
@@ -236,5 +213,5 @@ For native builds, install the [prerequisite](#prerequisites) *System Packages*.
 
 ## Known Issues
 
-#### TensorRT 7.2.1
+#### TensorRT 7.2.2
 * None

VERSION

@@ -1 +1 @@
-7.2.1.6
+7.2.2.3

demo/BERT/CMakeLists.txt

@@ -33,10 +33,10 @@ endif()
 
 set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wno-deprecated-declarations")
 
-include($ENV{TRT_SOURCE}/cmake/modules/set_ifndef.cmake)
-set_ifndef(TRT_INC_DIR $ENV{TRT_SOURCE}/include)
-set_ifndef(TRT_LIB_DIR $ENV{TRT_RELEASE}/lib)
-set_ifndef(TRT_OUT_DIR $ENV{TRT_SOURCE}/build/out)
+include($ENV{TRT_OSSPATH}/cmake/modules/set_ifndef.cmake)
+set_ifndef(TRT_INC_DIR $ENV{TRT_OSSPATH}/include)
+set_ifndef(TRT_LIB_DIR $ENV{TRT_LIBPATH})
+set_ifndef(TRT_OUT_DIR $ENV{TRT_OSSPATH}/build/out)
 
 include_directories(
     infer_c