V2: Improved the data set and trained on Illustrious V0.1 for better
consistency across models, enjoy!
This model is intended to assist with generations that involve squatting
Trained on 228 images using Prefect illustrious XL
Tag: squatILL, facing viewer, looking back
** For better results add the position you're NOT using to negative
Enjoy and share results!
Description
FAQ
Comments (1)
Why do squatting poses in LoRA or checkpoints often have the feet up on the toes? Can I generate them with the feet flat instead, like this?
