无法使用 PackagesNotFoundError 在 anaconda 上安装 tensorflow-gpu

Question

I have created new environment in anaconda, webapp-env, to use streamlit and flask.

I have installed cuda (8.1) and cudatoolkit (11.2) so it is compatible with my environment and am trying to install tensorflow-gpu=2.8.0 (or 2.9.0) as per the table here: https://www.tensorflow.org/install/source_windows

Unfortunately, I am getting this error (below) and I dont know how to solve this.

(webapp-env) PS C:\Users\me\Coursera_NN_visualizer_web_app_MNIST_project> conda config --append channels conda-forge tensorflow-gpu = 2.9.0
usage: conda-script.py [-h] [-V] command ...
conda-script.py: error: unrecognized arguments: tensorflow-gpu = 2.9.0

(webapp-env) PS C:\Users\me\Coursera_NN_visualizer_web_app_MNIST_project> conda install -c conda-forge tensorflow-gpu=2.8.0
Collecting package metadata (current_repodata.json): done
Solving environment: failed with initial frozen solve. Retrying with flexible solve.
Collecting package metadata (repodata.json): done
Solving environment: failed with initial frozen solve. Retrying with flexible solve.

PackagesNotFoundError: The following packages are not available from current channels:

  - tensorflow-gpu=2.8.0

Current channels:

  - https://conda.anaconda.org/conda-forge/win-64
  - https://conda.anaconda.org/conda-forge/noarch
  - https://repo.anaconda.com/pkgs/main/win-64
  - https://repo.anaconda.com/pkgs/main/noarch
  - https://repo.anaconda.com/pkgs/r/win-64
  - https://repo.anaconda.com/pkgs/r/noarch
  - https://repo.anaconda.com/pkgs/msys2/win-64
  - https://repo.anaconda.com/pkgs/msys2/noarch

To search for alternate channels that may provide the conda package you're
looking for, navigate to

    https://anaconda.org

and use the search bar at the top of the page.

I have an Nvidia 3060 graphics card and can use tensorflow-gpu version 2.8.0 successfully in my anaconda base virtual environment. any advice?

Answer 1

Tnesorflow-GPU in the main channel is at version 2.6 and from conda-forge , only a linux version is available, see here and here

You probably used pip to install it in your base env. You can do so in your new env just as well