This project demonstrates how to use a fine-tuned BERT model for Named Entity Recognition (NER) using Streamlit for a simple web-based interface.
This application allows users to input text and receive NER predictions from a BERT model. The BERT model is pre-trained and fine-tuned on the CoNLL-2003 dataset for NER tasks.
- It contains a
.ipynbfile which contains the full code of fine-tuning BERT (base_uncased, 110M parameters) on the CoNLL dataset and saving this model. - It contains
app.py, which is an app built on the trained model.
-
To use the app, type the following command:
streamlit run app.py
-
Here is an image of running app
