CypherQube
CypherQube
> I'll need to devise a way to detect whether or not the model is 512 or 768 and copy the proper config file. Ideally, the webui would just do...
@d8ahazard Has this issue been solved?
> Does 512 just generate brown sample images for anyone else? It's because the 512 weights are loading with v2-inference-v.yaml instead of the v2-inference.yaml, you can fix this by duplicating...
@d8ahazard Why you can't just create a checkbox until you've solved how to detect the right .yaml file automatically?
> You need to convert that ckpt to its diffuser version manually to start training. In the train tab there is an option to convert the output model back to...
> Yeah, there is no upstream support yet, will add it if upstream supports it. @smy20011 Nitrosocke has already trained a model using v2.0 https://twitter.com/Nitrosocke/status/1596001111835484161
@smy20011 This repo https://github.com/bmaltais/kohya_ss now has support for SD v2.0 dreambooth
@smy20011 Diffusers now supports dreambooth for SD v2.0 https://twitter.com/psuraj28/status/1597272885026250761
> Should be fixed in v0.1.10. Leave this issue open for 3-4 days for people to test it. @smy20011 Sadly it doesn't work, it's giving the same error you get...