BIO Tagger using BiLSTM-CRF Model for Named Entity Recognition application in Natural Language Processing. A neat little blog for background information can be found here.

The requirements.txt file has all the dependencies needed for the successful execution of the project till the end. It assumes you are having a CUDA compatible GPU for GPU accelerated PyTorch training. While the code automatically handles CPU/GPU device usage based on availability, you can also modify the requirements file yourself according to your needs.

To install the dependencies, ideally in a separate venv or a conda environment run:

pip install -r requirements.txt

Which should pull all the needed libraries.

The main execution flow is inside src/main.ipynb which offers an interactive jupyter notebook, which imports the other required local libraries from python files. Please check the documentation for each class and function for more information.