cell2location installation on a Windows machine with GPU usage

[hkeremcil]

Is there currently a working way of installing cell2location on a Windows machine with a GPU? I have tried many different methods, and I always end up getting errors or conflicts with scvi-tools and jax. The installations I tried mainly look like this:

setx PYTHONNOUSERSITE "literallyanyletters"
setx PYTHONPATH ""
conda create -y -n cell2loc_env_gpu python=3.9
conda activate cell2loc_env_gpu
pip install cell2location[tutorials]==0.1.3
pip install scvi-tools==1.0.4
pip install jaxlib==0.4.23
pip install jax==0.4.23
pip install jupyter notebook
pip install -U "jax[cuda12_pip]==0.4.23" -f https://storage.googleapis.com/jax-releases/jax_cuda_releases.html

I have also tried omitting the two lines between scvi-tools and jupyter notebook, not specifying jax versions, not specifying scvi-tools version, and installing these in every possible combination and order. The problem is something different for most methods. I would really appreciate it if I could get some guidance on installation. CPU version works but I need GPU usage enabled.

Many thanks in advance!