TAG:
trailblazer_(honkai:_star_rail)
kafka_(honkai:_star_rail)
march_7th_(honkai:_star_rail)
stelle_(honkai:_star_rail)
silver_wolf_(honkai:_star_rail)
dan_heng_(honkai:_star_rail)
caelus_(honkai:_star_rail)
blade_(honkai:_star_rail)
jingliu_(honkai:_star_rail)
himeko_(honkai:_star_rail)
bronya_rand
dan_heng_(imbibitor_lunae)_(honkai:_star_rail)
seele_(honkai:_star_rail)
fu_xuan_(honkai:_star_rail)
herta_(honkai:_star_rail)
jing_yuan
tingyun_(honkai:_star_rail)
ruan_mei_(honkai:_star_rail)
topaz_(honkai:_star_rail)
pom-pom_(honkai:_star_rail)
bailu_(honkai:_star_rail)
clara_(honkai:_star_rail)
asta_(honkai:_star_rail)
black_swan_(honkai:_star_rail)
The dataset was collected on February 7, 2024, and consists of 13k relevant images used for training.
Theoretically, all characters with a significant amount of data in the Honkai: Star Rail dataset prior to that date should be accurately represented, approaching a complete roster.
This model works best with Animagine XL V3 and models trained based on Animagine XL V3 or models that incorporate Animagine XL V3. It provides a higher level of character fidelity.
Tools used for image collection: https://github.com/deepghs/waifuc
Tools used for dataset processing: https://github.com/Eugeoter/sd-dataset-manager
Project used for model training: https://github.com/bmaltais/kohya_ss
Special thanks to GPU Sponsor: nieta.art
Unauthorized reproduction or publication on other platforms is strictly prohibited without the author's permission.
更多关于本模型的信息，如何使用，请看：
https://civitai.com/articles/4057/lora-about-honkai-star-rail-all-in-one-animagine-xl-v3
数据集的收集时间为2024年2月7日，使用了13k张相关图片训练。
对于在此之前星穹铁道数据集较多的角色，理论上都还原，无限接近于真正的全角色，
本模型最好在Animagine XL V3、以及基于Animagine XL V3训练的模型和融合了Animagine XL V3上的模型上使用，角色还原度会比较好。
图片收集工具：https://github.com/deepghs/waifuc
数据集处理工具：https://github.com/Eugeoter/sd-dataset-manager
模型训练使用的项目：https://github.com/bmaltais/kohya_ss
特别鸣谢：算力赞助 / GPU Sponsor: nieta.art
未经作者许可，禁止以任何方式转载或发布到其他平台。