Fine Tuning GPT-3 with no prompt?

This sounds very interesting. How can I learn more on this? Do you have any reading recommends?