Open
Description
Hello Kj, love your work, I have added an extra model and bent your code to work with it, it does work fine but I have added:
florence-2-large-nsfw-pretrain-gt
This model does not hallucinate or add clothing to images as it is not prude like the other VLMs, for real life instances this model is actually very good especially if you want accuracy.
Are you happy for me to fork, if not I can understand that. Or would you want the new code and the tweak you need to do to the model so you can push it as an update in this repo?
Metadata
Metadata
Assignees
Labels
No labels