DPO fine-tuning outperforms SFT

1 pointsposted a year ago
by kcorbitt
(openpipe.ai)

No comments yet