Don't know if you guys notice a big difference. But this was made using a second checkpoint model to refine the image.
In theory it should increase the level of details of a single image, at the cost of nearly doubling generation time.
If you think it's a big improvement let me know
edit: also makes it less "anime" and more "realistic"
PsyOpsAI art
2024-06-21 09:21:19 +0000 UTCNoximilien l'Horloger
2024-06-21 09:19:37 +0000 UTC