Qi-paw ฅ ˆ. ̫.ˆ ฅ

This model is a fine-tune of Illustrious-XL SmoothFT, trained to improve overall quality and stability.

• The UNet was fine-tuned on curated data using DPO training with over 10,000 image pairs.

• After training, the text encoder was replaced with the one from Illustrious-XL-v2.0

License:

Same as Illustrious. Feel free to use in your merges.