Instructions to use circlestone-labs/Anima with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Diffusion Single File
How to use circlestone-labs/Anima with Diffusion Single File:
# No code snippets available yet for this library. # To use this model, check the repository files and the library's documentation. # Want to help? PRs adding snippets are welcome at: # https://github.com/huggingface/huggingface.js
- Notebooks
- Google Colab
- Kaggle
Getting censor bars with certain artists, even with censored in negative prompt.
Title kinda says it all. Tried various combinations of negative prompt, including (censored:2) etc. nothing seemed to make much of a difference.
I also still get watermarks quite frequently, despite them being in my negative prompt.
See: https://huggingface.co/circlestone-labs/Anima/discussions/94
If all (or most) of the images of the artist you selected in the dataset are watermarked or censored, the model will recognize them as part of that artist's work, so even if you add neg, the watermarks or censorship will still appear.
data set is not sanitized (not necessarily a bad thing, mind you), you will get this outputs. Nothing anyone can do about it, maybe full version will be better, maybe not.
just put uncensored in positive. it's much stronger than any negative prompt