Closed kimihailv closed 1 year ago
Hi @kimihailv! I improvised that part since there aren't many details about determining sampling schedules.
That's basically the same schedule used when training is almost finished, so picking evenly spaced points (in _timestep
space) from there seems to be a natural choice. Maybe it could be overfitted to those points, who knows? 😂
I think that's the reason why continuous-time training is better, without discretization.
But practically, you could freely choose whatever sampling schemes you want. There shouldn't be a noticeable quality difference.
@junhsss Thank you for such detailed answer! By the way, it is interesting that authors found optimal timesteps by greedy algorithm
@kimihailv Oh, I must have missed that part! Thanks for pointing that out.
It is indeed interesting. So we need to sample more than N steps to perform N-step sampling... 😅 I'll implement it as soon as possible. Thanks again!
Hello. Could you please explain why do you use this tilmestep schedule:
May be it is from some paper?