This is the official repository for the paper 'LLMs Could Autonomously Learn Without External Supervision'.
- Python == 3.10.13
- torch == 2.0.1
- transformers == 4.35.0
- alignment-handbook == 0.3.0.dev0
- numpy == 1.26.3
- vllm == 0.2.1
- trl == 0.7.4
- accelerate == 0.23.2
bash run_AL.sh
We experiment on 8 Tesla A100-80GB GPU.
If you have any questions related to the code or the paper, feel free to email Ke Ji (keji@link.cuhk.edu.cn).
Please cite our paper if you find the repo helpful in your work:
@article{ji2024llms,
title={LLMs Could Autonomously Learn Without External Supervision},
author={Ji, Ke and Chen, Junying and Gao, Anningzhe and Xie, Wenya and Wan, Xiang and Wang, Benyou},
journal={arXiv preprint arXiv:2406.00606},
year={2024}
}