latent-diffusion icon indicating copy to clipboard operation
latent-diffusion copied to clipboard

Published 20 hours ago • verified publisher icon CompVis

How to train ldm on a low-resolution image dataset (128*128)

Open crowningwang opened this issue 1 year ago • 4 comments

crowningwang avatar Jul 24 '23 03:07 crowningwang

You should train a new autoencoder to encode the source image 128x128x3 into a latent code 32x32x4.

GrandpaXun242 avatar Jul 28 '23 07:07 GrandpaXun242

You should train a new autoencoder to encode the source image 128x128x3 into a latent code 32x32x4.

Could you tell me how to train? Thanks

tangwy98 avatar Dec 18 '23 09:12 tangwy98

You should train a new autoencoder to encode the source image 128x128x3 into a latent code 32x32x4.

Could you tell me how to train? Thanks

That is a simple processing. You can search "how to train a autoencoder?" in google. You should prepare your datasets and design a network containing latent code 32x32x4. And ensure the image size with 128x128x3.

GrandpaXun242 avatar Dec 18 '23 10:12 GrandpaXun242