
Hi i learned alot about Loras so here is a new and much better version of the edelgard Lora, the old one was trained on 3000 danbooru images, this one was created using 100 handpicked images and tagging them accordingly.
Best to use at 0.7 weight
To use the model best use:
Edelgard Emperor:
Use (Edelgard Emperor) in the positive prompt
Normal Edelgard:
Use (Edelgard van Hresvle) in the positive prompt and use Edelgard Emperor in the negative prompt
Monastery Edelgard:
Use (Edelgard van Hresvle) and (garreg mach monastery uniform, black uniform) in the positive prompt and use Edelgard Emperor in the negative prompt
描述:
Changed a lot of things thanks to this guide, is now even better then version V2:
https://rentry.org/LazyTrainingGuide
Best to use at 0.65 weight
Outfits
Edelgard Emperor:
Use (Edelgard Emperor) in the positive prompt
Normal Edelgard:
Use (Edelgard van Hresvle) in the positive prompt and !!!use (Edelgard Emperor) in the negative prompt
Monastery Edelgard:
Use (Edelgard van Hresvleg) and (garreg mach monastery uniform, black uniform) in the positive prompt and !!!use (Edelgard Emperor) in the negative prompt
I would recommend using the prompt upper body for best results
All images were created using Anything V4.5
训练词语: edelgard van hresvleg ,edelgard emperor,garreg mach monastery uniform
名称: EdelgardV3Lora.safetensors
大小 (KB): 147576
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success