AssertionError: Torch not compiled with CUDA enabled

[nakumiayumu144]

[Rubiksman78]

Run setup_gpu.bat after you've setup CUDA, cudNN and stuff for using your GPU. Local chatbots need GPU to run for now.

[nakumiayumu144]

where is that?

[Rubiksman78]

Sorry I mixed up, did you install it with your own python or did you use the built in from the drive?

[nakumiayumu144]

What do you mean? Is it the CUDA?

[Rubiksman78]

Yes it is

[nakumiayumu144]

I already installed it

[Rubiksman78]

Sorry I mixed up, did you install it with your own python or did you use the built in from the drive?

Can you answer this question too?

[nakumiayumu144]

I installed it on my own

[Rubiksman78]

Did you install torch version for CUDA ? I suggest you to look at the Python installation section on the README if you did the setup with your own python build.

[nakumiayumu144]

every time I change the version it is always incompatible with other modules pip install --force torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/cu117

[Rubiksman78]

Just download torch as explained and run the game usually, don't pay attention to the warnings it should work.

[nakumiayumu144]

can u give me example for character json? and where is the location save file i want to reset?

[Rubiksman78]

What do you mean by location save file?

There is a template for character json on the README.

[nakumiayumu144]

I mean I want to reset it, because every time I create a new one it still has the same history.

[Rubiksman78]

In char_history.txt

[nakumiayumu144]

thanks for helping me 👍