- Environment :: GPU
- Environment :: GPU :: NVIDIA CUDA :: 11.2
When you know about this project and you want to new install tf-nightly-gpu to support your project or you get trouble as ModuleNotFoundError: No module named "tf-nightly-gpu" or ImportError: cannot import name "tf-nightly-gpu" in your project, let follow this tutorial to install tf-nightly-gpu
In Windows (CMD):
py -m pip --version
In Unix/macOS:
python3 -m pip --version
Ensure pip, setuptools, and wheel are up to date:
In Windows (CMD):
py -m pip install --upgrade pip setuptools wheel
In Unix/macOS:
python3 -m pip install --upgrade pip setuptools wheel
Optional - If you want to install
In Windows (CMD):
- Install virtualenv - if you installed it, please ignore
py -m pip install --user virtualenv
- Create a virtual environment
py -m venv test_tf-nightly-gpu_env
- Active the virtual environment
test_tf-nightly-gpu_env\Scripts\active
In Unix/macOS:
- Install virtualenv - if you installed it, please ignore
pip3 install virtualenv
- Create a virtual environment
python3 -m venv test_tf-nightly-gpu_env
- Active the virtual environment
source test_tf-nightly-gpu_env/bin/active
To install tf-nightly-gpu on Windows(CMD):
py -m pip install tf-nightly-gpu
To install tf-nightly-gpu on Unix/macOs:
pip install tf-nightly-gpu
Example:
pip install tf-nightly-gpu==2.9.0.dev20220201 pre-release
Please see the version list below table:
Version | Released date | Command |
---|
Download the distribution file from tf_nightly_gpu-2.11.0.dev20220821-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl or the specific tf-nightly-gpu version in the below list of distribution
After that, install by command:
On Windows(CMD):
py -m pip install <path_to_tf-nightly-gpu_downloaded_file>
On Unix/macOs:
pip install <path_to_tf-nightly-gpu_downloaded_file>
List distribution:
Project link: