Torch not compiled with CUDA enabled about audioldm HOT 2 OPEN

Wrong torch install ? try un-installing torch then installing the cpu version ... :D . check on [ https://pytorch.org/get-started/locally/ ] and just go for the newest version choose pip or conda . then tun the code .

But! preferably use conda (anaconda) for the entire thing because you can do [ conda create --name audioldm python=3.8 ] to basically create a new environment that doesn't affect your other local installs of whatever you got going on there , just to be safe . Hopefully you followed the install guide haha xD

Like this you can just add the mentioned code from the website afterwards and just run the gradio by typing cd [enter your path to the audioldm you git pulled] . then type python app.py to run gradio

from audioldm.

Thank you - I am getting a crash though...
I know this is asking for a lot but:

What would be full order of console commands to create a conda environment with the CPU version of conda and run the CLI version of AudioDLM with the example command:
audioldm -t "A hammer is hitting a wooden surface"

I've tried this within the conda environment:
conda install pytorch torchvision -c pytorch
But not sure if that's correct? I am no coder unfortunately... :)

from audioldm.