mudler/LocalAI

feat: moderation endpoint

mudler opened this issue · 1 comments

mudler commented

Now that grammars are in, should be trivial to at least offer a minimum viable implementation of the moderation endpoint relying on constrained grammars

mudler commented

We could actually wire up text-classificators in the moderation endpoint. For example: https://huggingface.co/Titeiiko/OTIS-Official-Spam-Model/