search

hollowstrawberry / kohya-colab

Accessible Google Colab notebooks for Stable Diffusion Lora training, based on the work of kohya-ss and Linaqruf
GNU General Public License v3.0
564 stars 79 forks source link

Falling back to cpu #133

Open remote3513 opened 2 months ago

remote3513 commented 2 months ago

Hi,

I've noticed since yesterday i get a line in the script of the Colab page: "An NVIDIA GPU may be present on this machine, but a CUDA-enabled jaxlib is not installed. Falling back to cpu."

I have the GPU enabled in the Colab page and i've also tried all of the other GPU's but i get the error message with every GPU.

However, I am able to train despite "falling back to the CPU".

I just want to know if this an error on my side or perhaps in the script?

Cheers!

ArmyOfPun1776 commented 2 months ago

I get the same message when training. It's definitely a collab thing as nothing run in collab is using anything on your local system. I'm not even sure if this an actual issue or intended behavior. Seeing that the trainer doesn't break because of it, I'm leaning toward intended behavior.

hollowstrawberry commented 2 months ago

The training works fine despite this warning, and it's clearly using the gpu, so you can ignore it.

remote3513 commented 2 months ago

Ok! Cheers guys!

firofame commented 1 month ago

Screenshot 2024-06-01 at 10 15 19 PM

"An NVIDIA GPU may be present on this machine, but a CUDA-enabled jaxlib is not installed. Falling back to cpu."