at 5:58 " ... the variance comes from the network" This is not right. In DDPM, the authors made it constant and then in later studies people started to make those learnable as well.
@@moeinshariatnia59 In that formulation it is. Later in video I mentioned that it’s not necessary and model has to only predict the noise sampled from zero mean unit variance.
You helped me a lot! Thanks! Please keep going~
@@jiananwang2681 Thanks for watching🙂
clear!
@@yipengsun8624 Thanks!
Awesome, I'd love to see a video about how high-fidelity VAEs work and how they're trained.
awesome soroush. nice and clear explain
@@armanhatami5706 Thanks 😃
you are great , please keep going
Thanks Ahmed! Appreciate it.
Best❤
Nice short explanation!
Thanks!
Brilliant! thank you!!
@@oblivitus. Thanks for watching
Wow! great video. thanks a lot.
@@HassanHamidi-v8s Thanks for watching it 🙂
Great video, keep it up!
@@vidaadelimosabeb6689 Thanks😃
Ariane Gateway
at 5:58 " ... the variance comes from the network"
This is not right. In DDPM, the authors made it constant and then in later studies people started to make those learnable as well.
@@moeinshariatnia59 In that formulation it is. Later in video I mentioned that it’s not necessary and model has to only predict the noise sampled from zero mean unit variance.
@@soroushmehraban It's so funny you delete the comment instead of correcting the mistake :))
@@moeinshariatnia59 I didn't delete anything. It wasn't a mistake. Sorry if my explanation was ambiguous.