You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I was using guardrails-AI repo using NSFWText guardrail. I noticed it uses, as validator, the pre-trained multi-class model from HuggingFace - michellejieli/NSFW_text_classifier.
In your repo, it's only possible to define a threshold to check if the text passes or not. So, the output is only True or False. I was interested in getting, as output, also the score value of the text when it's evaluated by the validator.
In the documentation of the hugging face related to this model, it appears to be possible to get the score value.
Is it possible to get that in this repo or not? If yes, how?
Thanks in advance for the availability to read my question.