PaulRoeseler

SFT or RL? An Early Investigation into Training R1-Like Reasoning Large Vision-Language Models