Configuration for Trax-jax[cuda11] on linux(WSL2)

I was able to install trax-jax[cpu]-jaxlib-tensorflow-tensorflow-text-jupiter notebook , but I have problem to install it with GPU (on my notebook with Quadro).
Do you share some recent configuration for linux?
Cuda Toolkit =11, cudnn=8, jax[cuda11]-jaxlib-tensorflow[gpu]-tensorflow-text jupiter notebook for linux (probably python38-310)
p.s.not a big fun of building it from source
Thank you!

P.s.I asked our SP OPS for help and nobody heard of this kind of configuration on the production. Therefore I need your help.

Ok, got it!
my working configuration:

python 39,
jaxlib 0.4.14+cuda11.cudnn86,
tensorflow 2.13.0