SRFT: A Single-Stage Method with Supervised and Reinforcement Fine-Tuning for Reasoning

Devs

SRFT: A Single-Stage Method with Supervised and Reinforcement Fine-Tuning for Reasoning | Read Paper on Bytez