search

Stability-AI / stablediffusion

High-Resolution Image Synthesis with Latent Diffusion Models
MIT License
38.65k stars 4.99k forks source link

Details on the training procedure #81

Open nicolas-dufour opened 1 year ago

nicolas-dufour commented 1 year ago

Hi, Could you provide details on the training procedure of this model? How much steps were used on the stage 1 of the model (autoencoder)? On which dataset? Is it like with V1 where multiple subsets of LAION were used at different stages of the training? Similar question for stage 2?

This info would be very valuable for the reproducibility of this work and for future research in this domain! Would be super cool to have a log book like with OPT-175B, with the challenges faced during training!

Thank you very much

zzw-zwzhang commented 1 year ago

+1

wtliao commented 1 year ago

@nicolas-dufour do you have any update?

rzzdr commented 3 months ago

@dmarx pls remove good first issue from this

dmarx commented 3 months ago

@rzzdr you're gonna have to ask someone with perms on the repo, I haven't been in stability's employ for almost a year now.

rzzdr commented 3 months ago

OK