Definition. Supervised fine-tuning in InstructGPT [ouyang2022training, Section 3.5 and Appendix C.1] [ftip-006H]
Definition. Supervised fine-tuning in InstructGPT [ouyang2022training, Section 3.5 and Appendix C.1] [ftip-006H]
In the cited pipeline, supervised fine-tuning (SFT) starts from a pretrained GPT-3 checkpoint and updates it by supervised learning on labeler demonstrations. The resulting policy supplies the starting and reference policy for later stages.
Appendix C.1 reports optimization choices for that construction. It does not define a universal mask, weighting rule, or fine-tuning protocol.