To build a multi-headed model that’s capable of detecting different types of of toxicity like threats, obscenity, insults, and identity-based hate.
Word embedding done by glove.6b 100d vectors
1.kaggle_bi.py is a simple Bi-LSTM model,
2.kaggle_cnn.py is a model built using cnn,
3.kaggle_bi3.py is model built using bi-lstm and cnn,