๐ LoRA is trained purely on anime screencaps
๐ Activation tags are in the right panel
๐ Note: I'm well aware it's the "same" character, but having two separate versions made her eyes & face features much more stable. That's why it's not one LoRA.
๐ธ Always use ADetailer to get detailed eyes
๐ Technical notes:
LoRA is compatible with various styles - Lower the weight & remove "anime screencap" from the prompt, if you are getting too strong anime style
Hi. res fix used on all sample images:
Upscaler: R-ESRGAN 4x+ Anime6B
Steps: 10-15
Denoising: 0.3
๐ฌLeave a like, comment, review but first of all, ENJOY ;)