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