Instruction tuning
The fine-tuning stage that turns a base text-predictor into an assistant that follows instructions — supervised examples plus preference optimization.
The fine-tuning stage that turns a base text-predictor into an assistant that follows instructions — supervised examples plus preference optimization.