This is a personal reproduction of the Repoformer-3B model in this paper. This model is trained on ~300k Python chunk and function completion instances created from the Stack.
You can follow the instructions here to use the model.
If you find this useful, please cite:
@inproceedings{DBLP:conf/icml/0054AZR024,
author = {Di Wu and
Wasi Uddin Ahmad and
Dejiao Zhang and
Murali Krishna Ramanathan and
Xiaofei Ma},
title = {Repoformer: Selective Retrieval for Repository-Level Code Completion},
booktitle = {Forty-first International Conference on Machine Learning, {ICML} 2024,
Vienna, Austria, July 21-27, 2024},
publisher = {OpenReview.net},
year = {2024},
url = {https://openreview.net/forum?id=moyG54Okrj},
timestamp = {Mon, 02 Sep 2024 16:55:25 +0200},
biburl = {https://dblp.org/rec/conf/icml/0054AZR024.bib},
bibsource = {dblp computer science bibliography, https://dblp.org}
}