This is another LoRA I used to fine-tune my training dataset. It was trained using some of my personal colored drawings along with works from various other artists.

I mainly use it for coloring and refining characters when only low-quality, low-resolution, or black-and-white images are available to create a character LoRA.

Do not really need quality-related positive prompts — I’ve found that doing so usually changes the style away from what I’m aiming for. But you can try it, it may work better for you.

When used alongside a character LoRA, it usually enhances details such as the eyes and hair, creates sharper linework, and applies an ink-painted colouring style.

It can also produce more delicate sketches when prompts like "greyscale", "monochrome", "sketch", or "traditional media" are added.

It may be capable of more — feel free to experiment and have fun with it! :)