chore: Remove CUDNN dependencies (#2804)

Author: zewenli98

.github/workflows/docker_builder.yml

@@ -44,18 +44,16 @@ jobs:
         username: ${{ github.actor }}
         password: ${{ secrets.GITHUB_TOKEN }}
 
-    # Automatically detect TensorRT and cuDNN default versions for Torch-TRT build
+    # Automatically detect TensorRT default versions for Torch-TRT build
     - name: Build Docker image
       env:
         DOCKER_TAG: ${{ env.DOCKER_REGISTRY }}/${{ steps.fix_slashes.outputs.container_name }}
       run: |
         python3 -m pip install pyyaml
         TRT_VERSION=$(python3 -c "import versions; versions.tensorrt_version()")
         echo "TRT VERSION = ${TRT_VERSION}"
-        CUDNN_VERSION=$(python3 -c "import versions; versions.cudnn_version()")
-        echo "CUDNN VERSION = ${CUDNN_VERSION}"
 
-        DOCKER_BUILDKIT=1 docker build --build-arg TENSORRT_VERSION=$TRT_VERSION --build-arg CUDNN_VERSION=$CUDNN_VERSION -f docker/Dockerfile --tag $DOCKER_TAG .
+        DOCKER_BUILDKIT=1 docker build --build-arg TENSORRT_VERSION=$TRT_VERSION -f docker/Dockerfile --tag $DOCKER_TAG .
 
     - name: Push Docker image
       env:

README.md

@@ -19,7 +19,7 @@ Torch-TensorRT is distributed in the ready-to-run NVIDIA [NGC PyTorch Container]
 
 ## Building a docker container for Torch-TensorRT
 
-We provide a `Dockerfile` in `docker/` directory. It expects a PyTorch NGC container as a base but can easily be modified to build on top of any container that provides, PyTorch, CUDA, cuDNN and TensorRT. The dependency libraries in the container can be found in the <a href="https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/index.html">release notes</a>.
+We provide a `Dockerfile` in `docker/` directory. It expects a PyTorch NGC container as a base but can easily be modified to build on top of any container that provides, PyTorch, CUDA, and TensorRT. The dependency libraries in the container can be found in the <a href="https://docs.nvidia.com/deeplearning/frameworks/pytorch-release-notes/index.html">release notes</a>.
 
 Please follow this instruction to build a Docker container.
 
@@ -152,14 +152,13 @@ bash ./compile.sh
 You need to start by having CUDA installed on the system, LibTorch will automatically be pulled for you by bazel,
 then you have two options.
 
-#### 1. Building using cuDNN & TensorRT tarball distributions
+#### 1. Building using TensorRT tarball distributions
 
 > This is recommended so as to build Torch-TensorRT hermetically and insures any bugs are not caused by version issues
 
 > Make sure when running Torch-TensorRT that these versions of the libraries are prioritized in your `$LD_LIBRARY_PATH`
 
-1. You need to download the tarball distributions of TensorRT and cuDNN from the NVIDIA website.
-   - https://developer.nvidia.com/cudnn
+1. You need to download the tarball distributions of TensorRT from the NVIDIA website.
    - https://developer.nvidia.com/tensorrt
 2. Place these files in a directory (the directories `third_party/dist_dir/[x86_64-linux-gnu | aarch64-linux-gnu]` exist for this purpose)
 3. Compile using:
@@ -168,25 +167,16 @@ then you have two options.
 bazel build //:libtorchtrt --compilation_mode opt --distdir third_party/dist_dir/[x86_64-linux-gnu | aarch64-linux-gnu]
 ```
 
-#### 2. Building using locally installed cuDNN & TensorRT
+#### 2. Building using locally installed TensorRT
 
 > If you find bugs and you compiled using this method please disclose you used this method in the issue
 > (an `ldd` dump would be nice too)
 
-1. Install TensorRT, CUDA and cuDNN on the system before starting to compile.
+1. Install TensorRT and CUDA on the system before starting to compile.
 2. In `WORKSPACE` comment out
 
 ```py
 # Downloaded distributions to use with --distdir
-http_archive(
-    name = "cudnn",
-    urls = ["<URL>",],
-
-    build_file = "@//third_party/cudnn/archive:BUILD",
-    sha256 = "<TAR SHA256>",
-    strip_prefix = "cuda"
-)
-
 http_archive(
     name = "tensorrt",
     urls = ["<URL>",],
@@ -201,12 +191,6 @@ and uncomment
 
 ```py
 # Locally installed dependencies
-new_local_repository(
-    name = "cudnn",
-    path = "/usr/",
-    build_file = "@//third_party/cudnn/local:BUILD"
-)
-
 new_local_repository(
    name = "tensorrt",
    path = "/usr/",

cmake/Modules/FindcuDNN.cmake

@@ -7,11 +7,9 @@ endif()
 
 # If the custom finders are needed at this point, there are good chances that they will be needed when consuming the library as well
 install(FILES "${CMAKE_SOURCE_DIR}/cmake/Modules/FindTensorRT.cmake" DESTINATION "${CMAKE_INSTALL_LIBDIR}/cmake/torchtrt/Modules")
-install(FILES "${CMAKE_SOURCE_DIR}/cmake/Modules/FindcuDNN.cmake" DESTINATION "${CMAKE_INSTALL_LIBDIR}/cmake/torchtrt/Modules")
 
 # CUDA
 find_package(CUDAToolkit REQUIRED)
-find_package(cuDNN REQUIRED) # Headers are needed somewhere
 
 # libtorch
 find_package(Torch REQUIRED)

cmake/dependencies.cmake

@@ -23,7 +23,6 @@ target_link_libraries(${lib_name}
         TensorRT::nvinfer_plugin
         torch
         core_util
-        cuDNN::cuDNN
     PRIVATE
         Threads::Threads
 )

core/plugins/CMakeLists.txt

@@ -1,4 +1,3 @@
 __version__: "2.4.0.dev0"
 __cuda_version__: "12.1"
-__cudnn_version__: "8.9"
 __tensorrt_version__: "10.0.1"

dev_dep_versions.yml

@@ -8,9 +8,6 @@ ENV BASE_IMG=nvidia/cuda:12.1.1-devel-ubuntu22.04
 ARG TENSORRT_VERSION
 ENV TENSORRT_VERSION=${TENSORRT_VERSION}
 RUN test -n "$TENSORRT_VERSION" || (echo "No tensorrt version specified, please use --build-arg TENSORRT_VERSION=x.y to specify a version." && exit 1)
-ARG CUDNN_VERSION
-ENV CUDNN_VERSION=${CUDNN_VERSION}
-RUN test -n "$CUDNN_VERSION" || (echo "No cudnn version specified, please use --build-arg CUDNN_VERSION=x.y to specify a version." && exit 1)
 
 ARG PYTHON_VERSION=3.10
 ENV PYTHON_VERSION=${PYTHON_VERSION}
@@ -35,13 +32,12 @@ RUN wget -L https://github.com/pyenv/pyenv-installer/raw/master/bin/pyenv-instal
 RUN pyenv install -v ${PYTHON_VERSION}
 RUN pyenv global ${PYTHON_VERSION}
 
-# Install CUDNN + TensorRT + dependencies
+# Install TensorRT + dependencies
 RUN wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/cuda-ubuntu2204.pin
 RUN mv cuda-ubuntu2204.pin /etc/apt/preferences.d/cuda-repository-pin-600
 RUN apt-key adv --fetch-keys https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/7fa2af80.pub
 RUN add-apt-repository "deb https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/ /"
 RUN apt-get update
-RUN apt-get install -y libcudnn8=${CUDNN_VERSION}* libcudnn8-dev=${CUDNN_VERSION}*
 
 RUN apt-key adv --fetch-keys https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/3bf863cc.pub
 RUN add-apt-repository "deb https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/ /"