The aim of the project is to develop a web application that detects abusive comments. The project used Flask-API framework and Heroku cloud platform to deploy. Users are able to send a text to the model, via an API, and get back predictions on the web-server. I have used the Transformers library by Hugging Face and DistilBERT model to build a classifier on Pytorch deep learning library.
You can find descriptions of the models and the project itself on the site. LINK
- google-research/bert,
- huggingface,
- BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding,
- Training Sentiment Model Using BERT and Serving it with Flask API,
- L11 Language Models - Alec Radford (OpenAI),
- The Illustrated BERT, ELMo, and co.,
- BERT Fine-Tuning Tutorial with PyTorch,
- How to Fine-Tune BERT for Text Classification?,
- Huggingface Transformers,
- BERT Explained: State of the art language model for NLP