This repo present a method of finetuning BERT model for classifying text data.
The data used for testing this build was obtained from Twitter. SO, the repo contains scripts for cleaning Twitter data as well. The scripts can be adapted to any text classification use case.
- The code for prprocessing the data can be found in src/preprocess
- The code for building the model can be found in src/model folder
- src/train contains code for training the model.
- Remember to change the parameters in the code to suit your needs
- Contribuitions are welcome