Style based on the artist 03_baraTrained using 136 images scraped from gelbooruI used <a target="_blank" rel="ugc" href="https://civarchive.com/articles/4/make-your-own-loras-easy-and-free">this article</a> to train loras in the past. But with PonyXL models I only use the linked Collab Notebooks to produce my datasets then use Civitai's on-site training to train the models themselves.Dataset was tagged both by the Collab Notebook's tagging step and then since it skipped over some images I retagged everything using Civitai's on-site auto trainer prepending the artist's name to the tags.First sample images were done during on-site training

Files with SHA256: b9abd328d43aa4c26df18f8a06b9077099c447b29bf6effe1ed3a5be9844e99d