Diffusion with Transformers AND Diffusion In-Painting from Scratch! PyTorch Deep Tutorial

preview_player
Показать описание
In this Tutorial we revisit Latent Diffusion in Pytorch and have at look at how we can use an Image Transformer instead of a Unet! Latent Diffusion is a very popular method used in image diffusion models such as Stable Diffusion and other modern diffusion models! Diffusion transformers are also popular in modern Video Generation models like SORA!

Donations, Help Support this work!

The corresponding code is available here! (Section 9)

Discord Server:
Рекомендации по теме
Комментарии
Автор

Wohooo, new content! Thanks a lot already.
Usually I need a few attempts to absorb it all 😅

Omsip
Автор

This video is absolutely amazing!
Great job!

jensbauman
Автор

another amazing video Luke, just please try to raise the sound to be clearer to hear you. Thanks...

MoMaz-tq
Автор

Could you please do a VQ VAE implementation from scratch? would be a great addition to the autoencoder series

ramzibouderrah
Автор

Quite a great video! By the way, where can i find the dataset for this lecture?

xiaolongli-lxhz
Автор

Is there a way to return a piece of an image generated for each text token to create automaticaly masks for inpainting as it could be done in stable diffusion?

kasias
visit shbcf.ru