UBBK

Results 8 comments of UBBK

Couldnt manage to make this work. idk if its not working as intended or its on my end

Ok managed to make it work but now it only adds 1 lora per generation. If your prompt has 2 lora it ignores the second one.

Didnt help, still getting paging file error even after doubling pagefile size ``` CUDA SETUP: Loading binary C:\Users\...\sd-scripts\venv\lib\site-packages\bitsandbytes\libbitsandbytes_cuda116.dll... use 8-bit Adam optimizer running training / 学習開始 num train images *...

Increased page file size to the max amount i can and now i am getting a different error. ``` ================================================================================ CUDA SETUP: Loading binary C:\Users\...\sd-scripts\venv\lib\site-packages\bitsandbytes\libbitsandbytes_cuda116.dll... use 8-bit Adam optimizer running...

Wasnt 8bit optimizer for bad GPUs? Mine is a 1070ti with 8GB VRAM so don't think i can run this without 8bit optimizer?

Hmm...New DLL didn't seem to change anything. Still getting same error. ![image](https://user-images.githubusercontent.com/116608006/212418445-f3e40802-951c-4711-832e-4148170a9f9a.png)

Oh nevermind. Doing this AND increasing paging size at the same time solved the issue.

it is incomprehensible to me that this is something impossible to fix it seems. This is literally the one and only thing holding me back