Kairu

kairunoborugu_v1 (1).safetensors

v1.0

<h1 id="pdxl-ver-ghb7ve32p">PDXL Ver</h1>Probably the best style LoRA I have ever made.I forgot to make an image in the main post for furries so I made you all a whole collection in the gallery below. Sorry!<h1 id="sd-1.5-ver-nhqi0oxio">SD 1.5 Ver</h1>This LoRA disappoints me at the moment. I'll be retraining it soon. ***Update! This model is being a hassle. A V2 may happen eventually, but it won't be soon. If anyone is interested in the dataset to try yourself let me knowWhat is it good for?: It does a good job of adding in some lighting effects to your images. Try mixing it at lower weight with other art style LoRAs to change the lighting. I will include two examples at the end of the demo list.It's also surprisingly good at nsfw.What's wrong with it?: Primarily it's hands.... They are so bad most of the time. You'll likely need to inpaint to get them to turn out well.Other oddities: It really likes snowy mountains. If you don't specify a location that's what you'll probably get. It doesn't need 'kairunoborugu' in the prompt but it does help.As usual: Trained on Anything v3 Demos are on Anything v4.5 Weight for all images except those including my other LoRAs 0.8 Weight for images using my other LoRAs 0.6 with other LoRA at 0.8 Mixed LoRAs Shown: Harada Takehito then metata NSFW examples at the end before the other LoRA exmplesFor the staff: Not sure if I can use multiple LoRAs when showing off a LoRA, but if not then just let me know and I can remove those two images. Thanks!

Files with SHA256: 01f941bdf1c560d8a1db61a172263ee9bfb494875151ad0beb3e599acb2c169b