WARNING: THE EXAMPLES PROVIDED IN THE TEST AND TRAIN FILES CONTAIN EXTREME PROFANITY AND OFFENSIVE LANGUAGE.
A simple MLP regressor trained with sklearn.
Uses UCBerkley's hate speech dataset
If you want to use the model, run pip install -r requirements.txt. Then, run train.py. It will take 4-8 hours. Remember to credit me and UCBerkeley as per the license terms.
Visit the HuggingFace to download the pretrained model and vectorizer.