Ah, Alola!

20241217v2

Selene from POKEMON

Trigger Words: eneles

Trained on Pony Diffusion V6 XL checkpoint.

I have set a moderation on the gallery, so the NSFW content may not appear.

Update

Reorganizing the dataset and adding new hairstyles, outfits. Enhanced the consistency of the visual style.

Issues

Strange visual elements may appear, and sometimes there are cases of distorted limbs or missing body parts.

The learning results for the twin-tail hairstyle and one specific outfit are mediocre.

Usage tips

• The model's output may heavily rely on the checkpoint.

• You could find the example prompts from my post.

• My prompts are basically composed in the order of [character traits] + [style] + [expression] + [clothing] + [camera and action] + [background], and you can delete or modify them as needed.

• Recommended weight: 1~0.6, adjust as needed until the character's appearance meets your requirements.

• Upscale value recommendation is around 1.2~2.0, denoising strength is 0.2

• Facial distortion may easily occur in situations such as full-body shots. If there is facial distortion, consider using ADetailer for repair.

20240808v1

Recommended weight: 1.0~0.6, adjust as needed until the character's appearance meets your requirements.

0.6 is more flexible, the higher the weight, the more fixed the style of character. When using "realistic" and other style tags with lower weights, it might have a surprising effect.

Upscale value recommendation is around 1.5~2.0 , denoising strength is 0.2

Continuous adjustment of the prompt words is needed to achieve a good representation.

The model is currently prone to various phenomena such as bad hands, bad foot and scene fragmentation. It might be worth considering reducing the model's weight to see if it can alleviate the issue.