Type | |
Stats | 362 691 273 |
Reviews | (75) |
Published | Aug 10, 2024 |
Base Model | |
Training | Steps: 1,100 Epochs: 114 |
Trigger Words | hikaridawn akaridawn |
Hash | AutoV2 49F32C2638 |
The training of this model and the images it generates are solely for learning purposes.
You could find the example prompts from the images above.
Trigger words: hikaridawn for Dawn, akaridawn for Akari
Recommended weight: 1.0~0.6, 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
Continuous adjustment of the prompt words is needed to achieve a good representation.
The higher the weight, the more fixed the style of character.
There is a situation of overfitting.
The default clothing tends to cause spots on the screen, which is probably due to the dataset not further distinguishing images of various styles, and to improve this, it might be necessary to tag images of a certain style with corresponding style tags.
For some clothing such as beanie, when not wanted, consider adding them to the negative prompt.