elder453 / cs477-final

The code for Yale University's Fall 2024 CPSC477: Natural Language Processing final project.

CPSC 477 Final Project

Predicting Primary Sub-Categories of Statistics ArXiv Papers

Team Members: Eugene Han, Ali Aldous, Elder Veliz

Setup

1. Clone the repository.
2. Create a new conda environment with the following command:

conda env create -f env.yml

Please refer to the YAML file to see all dependencies/libraries used and their versions.

3. Download the baseline and fine-tuned RoBERTa models from this link.
4. Place the downloaded models in the models directory.
5. Run the cells of full_pipeline.ipynb inside the notebooks folder. We recommend skipping the training cells as the models are already trained and saved in the models directory. To replicate the results in the report, go to the evaluation section.