V4: New dataset. A much wider range of people, V1 - V3 we're almost all solo women. A much larger variety in landscape shots. More character variety including age, group images, and went from 350 tags in the data to over 1600 tags. New lighting environments including sunlight and sunset.
This uses my trained SinGlo lora/checkpoints from Illustrious mixed in with Anima Base!
Works very well with natural language.
YOU NEED:
qwen base encoder & qwen image vae
All gallery images should have metadata so you can drag and drop to see the exact things you need along with prompts and settings.
SETUP:
Steps: 30 - 50
CGF: 4 - 5
Sampler: Eular/Eular A (haven't tested others)
Description
Small update. Changed from using Preview 3 to Base.
FAQ
Comments (4)
Example pics have no metadata in them
Didn't even notice! I gave the images prompts to the best of my knowledge. Their not exact but very close. Thanks for lmk.
What a fantastic checkpoint. Whatever you did to avoid the trap most other checkpoints fall into, of being a static 1girl generator where the characters are front and center and boring and not doing a damn thing and it's always the exact same viewpoint, keep doing it. Lovely results.
Hey thanks! I def try! I hope to improve future versions on all my checkpoints. Appreciate it.