ProfanityDetector is now available!

ProfanityDetector

• Based on xlm-roberta-base
• Predictive text classification (normal, profanity)
• The model was only trained for 8 epochs for experimental purposes only (about 84% accuracy)

How to use?

• Create model_output folder in project
• Unzip ProfanityDetector.7z into the model_output folder
• Start flask service