scrape_dataset.py - This code calls the arxiv api to get summaries of different papers.
create_dataset.py - This code takes the summaries obtained in scrape_dataset.py and extracts problem-approach pairs from it.
train.py - This file contains the code for finetuning the models via LoRA, QLoRA, SFT.
eval.py - This file contains code for computing the metrics (bleu, rougel, meteor).
CSE-587_Final_project.pdf - Report for the final project.
Team Members - Aashrith Madasu (asm6590), Darshan Chudiwal (dsc5636), Yash Priydarshi (yvp5218)