As the name tells, this lora is trained on 170 HQ images on prompting "manhwa_style" in bing

just like this image

Lora Strength-0.8

cfg scale-5 to 7

sampling steps-20 to max

model use/favourable-animagine xl,animegod xl,juggermagine xl (please do not use aam xl-it gives glitching images)