I have a question regarding the NFEs used in SiT. The original SiT model employed 125 steps of Heun's method, resulting in 250 NFEs. However, you mentioned that 25 steps (50 NFEs) can achieve a similar score in the ODE setting. Could you provide some insight into why this might be the case?
Hi, thank you for your awesome work.
I have a question regarding the NFEs used in SiT. The original SiT model employed 125 steps of Heun's method, resulting in 250 NFEs. However, you mentioned that 25 steps (50 NFEs) can achieve a similar score in the ODE setting. Could you provide some insight into why this might be the case?
Thank you in advance.