Tabular Transfer Learning via Prompting LLMs

Official implementation of "Tabular Transfer Learning via Prompting LLMs" by Jaehyun Nam, Woomin Song, Seong Hyeon Park, Jihoon Tack, Sukmin Yun, Jaehyung Kim, Kyu Hwan Oh, Jinwoo Shin.

TL;DR: We propose a novel tabular transfer learning framework that leverages LLM's in-context learning capabilities.

1. Dependencies

conda create -n p2t python=3.8 -y
conda activate p2t

pip install openai==0.28.1

2. P2T

See main.ipynb to run P2T stey by step. We have provided (i) instructions on downloading the dataset from the OpenML repository and (ii) some example indexes in the /data folder.

Citation

@inproceedings{nam2024tabular,
  title={Tabular Transfer Learning via Prompting LLMs},
  author={Jaehyun Nam and and Woomin Song and Seong Hyeon Park and Jihoon Tack and Sukmin Yun and Jaehyung Kim and Kyu Hwan Oh and Jinwoo Shin},
  booktitle={Conference on Language Modeling},
  year={2024}
}