Aligning pretrained language models with instruction data generated by themselves.
No reviews for this project.