my dataset still a bit little, i think about making version v2.0 with character tag, and varoius expression etc, after seeing how gorgeus this is
use it with sampler (euler ancestral and scheduler normal) or (er sde and scheduler sgm uniform) at weight lora around 0.7 to 1, with between around 40-70 step, if you're using turbo just 15 step it's still pretty good
there is no trigger word for this model
####################################
I accept LoRA training commissions for both Anima and Illustrious!
If you're interested, want to make a request, or simply support me, please check my Linktree to access my Ko-Fi:
Even the smallest contribution means a lot and truly motivates me to keep creating models and tools for the community — especially since I'm currently training everything on cloud GPUs. 🤍
Also, feel free to explore my model library — I love making LoRAs, though my page is still small at the moment. With your support, I'll keep making more in the future!