Not mine: https://civitai.com/models/881646/synthwave-pony
Author notes(https://civitai.com/user/SickMoonDoe):
About this version
Reverts text encoder training.
This should improve prompt adherence and avoid instances where trigger words led to simple white/gradient backgrounds.
This was baked from my https://civitai.com/models/843858/softie-pony checkpoint and likely works best there; but it works great on the other Pony models I've tried.
Weigh around 0.75 to start; but it's real fun to crank it up high ;)
