Questions about LCM model training

Open luochunfen opened this issue 1 year ago • 1 comments

Hi! I'm trying to train the LCM model. However, the model I trained is worse than the LCM model you released. Could you give more training details about LCM distillation ? Such as EMA decay, learning rate, training steps, optimizer. Is there any difference with the implementation of the official LCM (https://github.com/luosiallen/latent-consistency-model.git). Thank you for your help!

Aug 01 '24 03:08 luochunfen

Hey thanks for your interest, we are preparing an LCM drop (paper, code), follow the repo/us on social media to learn when it is out

Aug 10 '24 13:08 toshas

Hey thanks for your interest, we are preparing an LCM drop (paper, code), follow the repo/us on social media to learn when it is out

OK. Thank you for your reply.

Aug 15 '24 08:08 luochunfen

Hi, any ideas when you'll be able to release the LCM drop?

Aug 17 '24 08:08 omrastogi