lqcfcjx·2 года назад·discussThis is very impressive. I know in general people are iffy about research benchmark. How does it work to evaluate text-to-video types of use cases? I want to have some intuition on how much this is better than other systems like pika quantatively.