Unfortunately, this is the only version of this video and we could not find a version with a better quality. Since Meta Learning is a new and a very important topic, we uploaded the video anyway. We hope that you enjoy it!
@21:30 It seems like the "support" and "query" sets should point to the training/test sets in the meta-training dataset, rather than in the meta-testing dataset as marked in the slide. Could someone confirm? or am I misunderstanding what he is explaining?
Unfortunately, this is the only version of this video and we could not find a version with a better quality. Since Meta Learning is a new and a very important topic, we uploaded the video anyway. We hope that you enjoy it!
looks fine to me :D
thank you for sharing it.
Very nice talk! Video looks great (we see both slides and presenter), sounds like audio drops at 25:38
if you are looking for the second part of the video: facebook.com/icml.imls/videos/2970931166257998
Can't wait for Chelsea's CS330 videos to be released. Both Sergey and herslf explain things in such a clear way
Yep!
@21:30 It seems like the "support" and "query" sets should point to the training/test sets in the meta-training dataset, rather than in the meta-testing dataset as marked in the slide. Could someone confirm? or am I misunderstanding what he is explaining?
Had same confusion... but I also think it should point to the former (meta-training)
Anyone. at 31:25, shouldn't the final equation at bottom-right be about minimizing the loss. think that's a typo.