Lora trained on the artist Wlop (https://www.artstation.com/wlop)

The style gives a digital painting look, it is very good at reflections and small shiny objects.

The artist does mostly fantasy medieval images, as a result the LoRa does well with armors and weapons (swords, lances, fails,etc.).

It works well with the 3:2 ratio in portrait, but anything should work.

It was trained with noobAI v-pred v10 model. I recommend using ΣΙΗ (https://civitai.com/models/1217??5?modelVersionId=1371790) or WAI (https://civitai.com/models/989367?modelVersionId=1202076)

This LoRa was a request of someone on the ProjectAI discord, the dataset was kindly given already cleaned, all I did was tag the images and tune the training settings.

The settings used are inside the .yaml file (it's not a text encoder, thats the only option civitai gives to upload them), open with a notepad to see them.

Trigger word is : wlop

Exemple prompt :
wlop,1girl,:@,glowing red eyes,knight,dagger,long blue hair,tongue out,oni horns,
BREAK
very awa,masterpiece, best quality, amazing quality, very aesthetic, absurdres,