GPU doesnt seem to be detected

[myselfsama-git]

Not 100% sure if the GPU is being detected I have 3080Ti.

When initially installing and doing the Setup.bat I get a message saying:

=============================================================
Modules installed outside the virtual environment were found.
This can cause issues. Please review the installed modules.

You can uninstall all local modules with:

deactivate
pip freeze > uninstall.txt
pip uninstall -y -r uninstall.txt

=============================================================

Do you want to:
[1] - Quit
[2] - Continue with the setup
Enter your choice (1 or 2):

I pressed 2 and proceeded to install the rest, this was the only part of the installation that differed from what I've seen that you will be asked prior to installing.

Then when launching the gui.bat I got another set of messages which begin with the same message:

=============================================================
Modules installed outside the virtual environment were found.
This can cause issues. Please review the installed modules.

You can uninstall all local modules with:

deactivate
pip freeze > uninstall.txt
pip uninstall -y -r uninstall.txt
=============================================================
23:45:43-576197 INFO     nVidia toolkit detected
23:45:47-756761 INFO     Torch 2.0.1+cpu
23:45:47-759752 WARNING  Torch reports CUDA not available
23:45:47-760777 INFO     Validating that requirements are satisfied.
23:45:48-998347 INFO     All requirements satisfied.
23:45:54-043787 INFO     headless: False
23:45:54-050776 INFO     Load CSS...
Running on local URL:  http://127.0.0.1:7860

To create a public link, set `share=True` in `launch()`.

To create a public link, set 'share=True

Not sure what what I'm doing wrong. I have Auto1111 installed. I would like my GPU to be used when training LORAs

[myselfsama-git]

So far I have...
Uninstalled the whole thing, Installed CUDA toolkit 11.8, reinstalled Kohya, allowed it to install its own Torch thing. Now I get a new error after going through the whole install process.

11:02:32-814149 INFO     nVidia toolkit detected
11:02:34-208450 INFO     Torch 2.0.1+cu118
11:02:34-227359 ERROR    Could not load torch: cuDNN version incompatibility: PyTorch was compiled  against (8, 7, 0)
                         but found runtime version (8, 6, 0). PyTorch already comes bundled with cuDNN. One option to
                         resolving this error is to ensure PyTorch can find the bundled cuDNN.

[myselfsama-git]

I installed CUDNN 8.7, grabbed the files from Nvidia's site, and replaced the CUDNN all the DLL files within the CUDNN_Windows folder to updated ones and now it seems to work.