Goodbye to 2 years of hard work and 66 of my LoRA models

I am feeling a bit better. It's a small consolation, but I was able to afford making a few Flux versions of my favorite SDXL 1.0 LoRAs. It will cost more to use and they are almost the same but at least my favorite ones are not gone. @Tyberius @Chris451back @thepisu @Shazamm @givemeallthekittehs @BetaPrime99 @Kebyroth @Wroof @GingerD @JragenaAI

I haven't yet invested much in loras, but I think it's devastating to do that to people who have invested so much time in them, and who turn to the loras for particular results. Annihilating hours of work without any community consultation is really shocking. I'm sorry for your loss - this image really does capture it.

It’s dreadful! I’ll miss all these wonderful LoRas that you and others have spent so much time creating. The Flux LoRas use too many credits in my opinion, and are not half as good as yours. I don’t know what to suggest as to what to write, but many others are putting things on Facebook about this x

I often got surprising results with loras, but got the best results with really low settings like 0.3 I must say there were too many with vague descriptions, wich made it confusing. Better would be to give every lora a separate page, with examples and tips and tricks. By now there could be an encyclopedia with information about the different algorithms, their history, how they were developed and their strong and weak points.

Sad news indeed, That is a lot of LORAs to lose. This whole scene is just evolving so fast, hard to catch my breath!

Fantastic

Perfect image… sad situation…bogus indeed

So frustrating indeed. Here for you as you know. ❤️❤️❤️

This image really show what are you feeling!