|
--- |
|
license: other |
|
license_name: faipl-1.0-sd |
|
license_link: https://freedevproject.org/faipl-1.0-sd/ |
|
base_model: yodayo-ai/kivotos-xl-2.0 |
|
pipeline_tag: text-to-image |
|
language: |
|
- en |
|
--- |
|
|
|
# Millennium-XL |
|
 |
|
|
|
Modified Kivotos-XL-2.0 with Blue Archive style |
|
|
|
Who wants this??? |
|
|
|
## Prompt Guidelines |
|
Almost same as Animagine-XL-3.1 |
|
|
|
## Recommended Prompt |
|
None(Works good without `masterpiece, best quality, very aesthetic`) |
|
|
|
## Recommended Negative Prompt |
|
`worst quality, low quality, lowres, jpeg artifacts, unfinished` |
|
|
|
## Recommended Settings |
|
Steps: 12-24 |
|
|
|
Sampler: DPM++ 2M(dpmpp_2m) |
|
|
|
Scheduler: Simple or Karras |
|
|
|
Guidance Scale: 4-9 |
|
|
|
## About trinity version |
|
Adjusted for flatter style |
|
|
|
## Training information |
|
Finetuned Kivotos-XL by repeating the training and merging a DoRA 3 times with sd-scripts. |
|
|
|
## Dataset information |
|
Dataset size: 350 |
|
|
|
## Training scripts: |
|
[sd-scripts](https://github.com/kohya-ss/sd-scripts) |
|
|
|
## Notice |
|
This model is licensed under [Fair AI Public License 1.0-SD](https://freedevproject.org/faipl-1.0-sd/) |
|
|
|
If you make modify this model, you must share both your changes and the original license. |
