Vastly improved ????? model, with strong prompt following (using selected 6k images, hand captionned over a year!).
The kinks can be combined and the images should look good from your first tries thanks to an enhanced base model style, strong at both photography, colorful digital painting, and anime style with advanced shading (not focused on lineart style). As before, you should refer to the captions page here to get an idea of what to ask the AI to generate : https://pickled-fight-817.notion.site/Eldgard-Prompting-guide-bb11234a0d664299b539db88bae942d0?pvs=74

