This is the extracted Lora from my full finetune using 600+ images with a variety of themes and subjects.
The data set has been designed with an emphasis on Oatmeal, gunge and custard. The low learning rate and high number of steps gives the model a high degree of flexibility and realism.
I have uploaded the full FP8 checkpoint version to my profile also. This will always give better results than the lora
My sample images have the prompts embedded, Please share any interesting images, it'd be great to see what can be done with this.
Total training time was just under 8 days end to end on an A6000. Training does not come cheap so any contribution of Buzz will help to support continuing development
Description
initial extracted Lora
FAQ
Looks like we don't have an active mirror for this file right now.
CivArchive is a community-maintained index — we catalog mirrors that volunteers upload to HuggingFace, torrents, and other public hosts. Looks like no one has uploaded a copy of this file yet.
Some files do get recovered over time through contributions. If you're looking for this one, feel free to ask in Discord, or help preserve it if you have a copy.