4 files that help to input tags on stable-diffusion-webui for generating image via animagine-xl-3.0
- animagine-xl-3.0.csv : tag data file for a1111-sd-webui-tagcomplete
- character_plus.txt : character wildcards
- animagine_xl_3.0_artists.txt : artist wildcards
- danbooru2023.csv : tag data file using danbooru2023 dataset
You can modify or create new using ipynb files.
A tag data file from Linaqruf/1.2m-ordered-tags dataset, which I believe the data for training animagine-xl-3.0.
- Put file on
extensions/a1111-sd-webui-tagcomplete/tags
. - Select animagine-xl-3.0.csv' on Settings > Tag Autocomplete > Tag filename, then press apply button.
- Type on prompt.
Wildcard files for character tags and artist tags, that
- based on chracter wildcards from Linaqruf/animagine-xl space, and I add 10 tags mostly used with each characters.
- about 1000 mostly used artist tags and 10 additional tags that mostly used with.
- Put files on valid wildcards location.
(In case of sd-dynamic-prompts, it'sextensions/sd-dynamic-prompts/wildcards
.) - Input
__character_plus__
or__animagine_xl_3.0_artists__
on prompt and generate. - a1111-sd-webui-tagcomplete may helps to input wildcard on prompt.