Trained with 130 images on anima-base-v1.0.
Trigger: yulcinoa_style anime illustration. (Works without a trigger.)
Comfyui workflow is embedded in the images if you need one.
Settings:
Lora weight: 1
Sampler: Euler A, SGM Uniform, 25 - 28 steps
CFG: 4
Description
Trained with 130 images on anima-preview2. The samples were generated with anima-preview3. V3 released while the training was running, but it seems to be compatible with my lora so i just went with it.