Sinistcha Anima
Another model trained off certain artist's work, huh?
- Capable of multiple characters
- Does not lose control on low configs
Recommended configurations:
Sampling steps: 22 (lowres) - 25 (1024≤) - 29 (If including text in image)
Res: 512x512 for ~16VRAM, can go up to 1920x1920 on good GPUs
Trigger: nyantcha, @nyantcha
(image does not include metadata sadly, screenshot)
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
Model tree for kisserpun/anima-sinistcha
Base model
nvidia/Cosmos-Predict2-2B-Text2Image Finetuned
circlestone-labs/Anima