Hello - I had previously set up TensorFlow/keras in a win 10 laptop to run the DeepLearn script included in Zorro. I built a new workstation with a RTX A4000 GPU specifically for DeepLearning algos. While preparing my environment I stumbled across the latest TensorFlow version (2.11) and noticed that it can only be installed on WSL2 environment.

Has anyone in this forum successfully installed TF2.11/keras on WSL2? If so, could you share some pointers? I just want to make sure I take the right approach before I jump into the python/conda/cudatoolkit/cudnn version compatibility nightmare rabbit hole.

Or should I completely avoid TensorFlow 2.11 and just stick to 2.10 and install the GPU version and live with the reduced GPU efficiency?

Please share your thoughts/experiences. Believe me, I went through tons of articles/YouTube videos before posting here. :-)

Thanks

Here are some of the articles/videos I have gone through:
https://www.tensorflow.org/install/source#gpu
https://towardsdatascience.com/how-to-finally-install-tensorflow-gpu-on-windows-10-63527910f255
https://towardsdatascience.com/how-to-finally-install-tensorflow-gpu-on-wsl2-7be59e278f92
https://towardsdatascience.com/inst...d-gpu-support-on-windows-10-60693e46e781
https://zorro-project.com/manual/en/deeplearning.htm
https://opserver.de/ubb7/ubbthreads.php?ubb=showflat&Number=486386&page=1
https://financial-hacker.com/build-better-strategies-part-4-machine-learning/
https://robotwealth.com/deep-learning-trading-tensorflow-keras-gpu/
https://docs.nvidia.com/cuda/wsl-user-guide/index.html#getting-started-with-cuda-on-wsl-2
https://www.youtube.com/watch?v=YozfiLI1ogY
https://www.youtube.com/watch?v=hHWkvEcDBO0