Install TensorFlow with pip

Linux

conda install -c conda-forge cudatoolkit=11.8.0
python3 -m pip install nvidia-cudnn-cu11==8.6.0.163 tensorflow==2.12.*
CUDNN_PATH=$(dirname $(python -c "import nvidia.cudnn;print(nvidia.cudnn.__file__)"))
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$CONDA_PREFIX/lib/:$CUDNN_PATH/lib
# Verify install:
python3 -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))"

MacOS

# There is currently no official GPU support for MacOS.
python3 -m pip install tensorflow
# Verify install:
python3 -c "import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))"

Windows Native

conda install -c conda-forge cudatoolkit=11.2 cudnn=8.1.0
# Anything above 2.10 is not supported on the GPU on Windows Native
python -m pip install "tensorflow<2.11"
# Verify install:
python -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))"

Windows WSL2

conda install -c conda-forge cudatoolkit=11.8.0
python3 -m pip install nvidia-cudnn-cu11==8.6.0.163 tensorflow==2.12.*
CUDNN_PATH=$(dirname $(python -c "import nvidia.cudnn;print(nvidia.cudnn.__file__)"))
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$CONDA_PREFIX/lib/:$CUDNN_PATH/lib
# Verify install:
python3 -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))"

CPU

python3 -m pip install tensorflow
# Verify install:
python3 -c "import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))"

Nightly

python3 -m pip install tf-nightly
# Verify install:
python3 -c "import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))"

Linux

1. System requirements

• Ubuntu 16.04 or higher (64-bit)

TensorFlow only officially support Ubuntu. However, the following instructions may also work for other Linux distros.

2. Install Miniconda

Miniconda is the recommended approach for installing TensorFlow with GPU support. It creates a separate environment to avoid changing any installed software in your system. This is also the easiest way to install the required software especially for the GPU setup.

You can use the following command to install Miniconda. During installation, you may need to press enter and type "yes".

curl https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh -o Miniconda3-latest-Linux-x86_64.sh
bash Miniconda3-latest-Linux-x86_64.sh

You may need to restart your terminal or source ~/.bashrc to enable the conda command. Use conda -V to test if it is installed successfully.

3. Create a conda environment

Create a new conda environment named tf with the following command.

conda create --name tf python=3.9

You can deactivate and activate it with the following commands.

conda deactivate
conda activate tf

Make sure it is activated for the rest of the installation.

4. GPU setup

You can skip this section if you only run TensorFlow on the CPU.

First install the NVIDIA GPU driver if you have not. You can use the following command to verify it is installed.

nvidia-smi

Then install CUDA and cuDNN with conda and pip.

conda install -c conda-forge cudatoolkit=11.8.0
pip install nvidia-cudnn-cu11==8.6.0.163

Configure the system paths. You can do it with the following command every time you start a new terminal after activating your conda environment.

CUDNN_PATH=$(dirname $(python -c "import nvidia.cudnn;print(nvidia.cudnn.__file__)"))
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$CONDA_PREFIX/lib/:$CUDNN_PATH/lib

For your convenience it is recommended that you automate it with the following commands. The system paths will be automatically configured when you activate this conda environment.

mkdir -p $CONDA_PREFIX/etc/conda/activate.d
CUDNN_PATH=$(dirname $(python -c "import nvidia.cudnn;print(nvidia.cudnn.__file__)"))
echo 'export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$CONDA_PREFIX/lib/:$CUDNN_PATH/lib' > $CONDA_PREFIX/etc/conda/activate.d/env_vars.sh

5. Install TensorFlow

TensorFlow requires a recent version of pip, so upgrade your pip installation to be sure you're running the latest version.

pip install --upgrade pip

Then, install TensorFlow with pip.

pip install tensorflow==2.12.*

6. Verify install

Verify the CPU setup:

python3 -c "import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))"

If a tensor is returned, you've installed TensorFlow successfully.

Verify the GPU setup:

python3 -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))"

If a list of GPU devices is returned, you've installed TensorFlow successfully.

Ubuntu 22.04

In Ubuntu 22.04, you may encounter the following error:

Can't find libdevice directory ${CUDA_DIR}/nvvm/libdevice.
...
Couldn't invoke ptxas --version
...
InternalError: libdevice not found at ./libdevice.10.bc [Op:__some_op]

To fix this error, you will need to run the following commands.

# Install NVCC
conda install -c nvidia cuda-nvcc=11.3.58
# Configure the XLA cuda directory
mkdir -p $CONDA_PREFIX/etc/conda/activate.d
printf 'export XLA_FLAGS=--xla_gpu_cuda_data_dir=$CONDA_PREFIX/lib/\n' >> $CONDA_PREFIX/etc/conda/activate.d/env_vars.sh
source $CONDA_PREFIX/etc/conda/activate.d/env_vars.sh
# Copy libdevice file to the required path
mkdir -p $CONDA_PREFIX/lib/nvvm/libdevice
cp $CONDA_PREFIX/lib/libdevice.10.bc $CONDA_PREFIX/lib/nvvm/libdevice/

MacOS

1. System requirements

• macOS 10.12.6 (Sierra) or higher (64-bit)

Currently there is no official GPU support for running TensorFlow on MacOS. The following is instructions are for running on CPU.

2. Check Python version

Check if your Python environment is already configured:

python3 --version
python3 -m pip --version

2. Install Miniconda

Miniconda is the recommended approach for installing TensorFlow with GPU support. It creates a separate environment to avoid changing any installed software in your system. This is also the easiest way to install the required software especially for the GPU setup.

curl https://repo.anaconda.com/miniconda/Miniconda3-latest-MacOSX-x86_64.sh -o Miniconda3-latest-MacOSX-x86_64.sh
bash Miniconda3-latest-MacOSX-x86_64.sh

You may need to restart your terminal or source ~/.bashrc to enable the conda command. Use conda -V to test if it is installed successfully.

4. Create a conda environment

Create a new conda environment named tf with the following command.

conda create --name tf python=3.9

You can deactivate and activate it with the following commands.

conda deactivate
conda activate tf

Make sure it is activated for the rest of the installation.

5. Install TensorFlow

TensorFlow requires a recent version of pip, so upgrade your pip installation to be sure you're running the latest version.

pip install --upgrade pip

Then, install TensorFlow with pip.

pip install tensorflow

6. Verify install

python3 -c "import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))"

If a tensor is returned, you've installed TensorFlow successfully.

Windows Native

1. System requirements

• Windows 7 or higher (64-bit)

2. Install Microsoft Visual C++ Redistributable

Install the Microsoft Visual C++ Redistributable for Visual Studio 2015, 2017, and 2019. Starting with the TensorFlow 2.1.0 version, the msvcp140_1.dll file is required from this package (which may not be provided from older redistributable packages). The redistributable comes with Visual Studio 2019 but can be installed separately:

1. Go to the Microsoft Visual C++ downloads.
2. Scroll down the page to the Visual Studio 2015, 2017 and 2019 section.
3. Download and install the Microsoft Visual C++ Redistributable for Visual Studio 2015, 2017 and 2019 for your platform.

Make sure long paths are enabled on Windows.

3. Install Miniconda

Miniconda is the recommended approach for installing TensorFlow with GPU support. It creates a separate environment to avoid changing any installed software in your system. This is also the easiest way to install the required software especially for the GPU setup.

Download the Miniconda Windows Installer. Double-click the downloaded file and follow the instructions on the screen.

4. Create a conda environment

Create a new conda environment named tf with the following command.

conda create --name tf python=3.9

You can deactivate and activate it with the following commands.

conda deactivate
conda activate tf

Make sure it is activated for the rest of the installation.

5. GPU setup

You can skip this section if you only run TensorFlow on CPU.

First install NVIDIA GPU driver if you have not.

Then install the CUDA, cuDNN with conda.

conda install -c conda-forge cudatoolkit=11.2 cudnn=8.1.0

6. Install TensorFlow

TensorFlow requires a recent version of pip, so upgrade your pip installation to be sure you're running the latest version.

pip install --upgrade pip

Then, install TensorFlow with pip.

# Anything above 2.10 is not supported on the GPU on Windows Native
pip install "tensorflow<2.11" 

7. Verify install

Verify the CPU setup:

python -c "import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))"

If a tensor is returned, you've installed TensorFlow successfully.

Verify the GPU setup:

python -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))"

If a list of GPU devices is returned, you've installed TensorFlow successfully.

Windows WSL2

1. System requirements

• Windows 10 19044 or higher (64-bit). This corresponds to Windows 10 version 21H2, the November 2021 update.

See the following documents to:

• Download the latest Windows 10 update.
• Install WSL2
• Setup NVIDIA® GPU support in WSL2

2. Install Miniconda

Miniconda is the recommended approach for installing TensorFlow with GPU support. It creates a separate environment to avoid changing any installed software in your system. This is also the easiest way to install the required software especially for the GPU setup.

You can use the following command to install Miniconda. During installation, you may need to press enter and type "yes".

curl https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh -o Miniconda3-latest-Linux-x86_64.sh
bash Miniconda3-latest-Linux-x86_64.sh

You may need to restart your terminal or source ~/.bashrc to enable the conda command. Use conda -V to test if it is installed successfully.

3. Create a conda environment

Create a new conda environment named tf with the following command.

conda create --name tf python=3.9

You can deactivate and activate it with the following commands.

conda deactivate
conda activate tf

Make sure it is activated for the rest of the installation.

4. GPU setup

You can skip this section if you only run TensorFlow on the CPU.

First install the NVIDIA GPU driver if you have not. You can use the following command to verify it is installed.

nvidia-smi

Then install CUDA and cuDNN with conda and pip.

conda install -c conda-forge cudatoolkit=11.8.0
pip install nvidia-cudnn-cu11==8.6.0.163

Configure the system paths. You can do it with following command everytime your start a new terminal after activating your conda environment.

CUDNN_PATH=$(dirname $(python -c "import nvidia.cudnn;print(nvidia.cudnn.__file__)"))
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$CONDA_PREFIX/lib/:$CUDNN_PATH/lib

For your convenience it is recommended that you automate it with the following commands. The system paths will be automatically configured when you activate this conda environment.

mkdir -p $CONDA_PREFIX/etc/conda/activate.d
CUDNN_PATH=$(dirname $(python -c "import nvidia.cudnn;print(nvidia.cudnn.__file__)"))
echo 'export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$CONDA_PREFIX/lib/:$CUDNN_PATH/lib' > $CONDA_PREFIX/etc/conda/activate.d/env_vars.sh

5. Install TensorFlow

TensorFlow requires a recent version of pip, so upgrade your pip installation to be sure you're running the latest version.

pip install --upgrade pip

Then, install TensorFlow with pip.

pip install tensorflow==2.12.*

6. Verify install

Verify the CPU setup:

python3 -c "import tensorflow as tf; print(tf.reduce_sum(tf.random.normal([1000, 1000])))"

If a tensor is returned, you've installed TensorFlow successfully.

Verify the GPU setup:

python3 -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))"

If a list of GPU devices is returned, you've installed TensorFlow successfully.