I use Civitai for lora training, image reactions, posting images or tips are very appreciated, helps me create more lora.

Lora made on the style of Nini, an artist that releases a few works over a decade ago and disappeared. this was more of a test given the small training sample size, thus it came out more of a character lora than style. Recreates characters very accurately, but the style comes out blurry unless directly referencing a training image. The style is much stronger when using "monochrome, grayscale, comic," prompts

Trigger word: Nini

Strength: 1.0

Trained prompts: large breasts, small breasts, cowgirl, reverse cowgirl, ?????? press, piledriver, ass, ass view, breast grab, ?????? pinch, ????????, ???????, cooperative ????????, coopbj, double ???????, paizuri, boobjob, doggystyle, front view, looking at viewer, breast sucking, spooning, 1girl, POV, bouncing breasts, girl on top, guy on top, man on top, grabbed from behind, bent over, standing missionary, standing doggystyle, carry sex

Characters:

• Moka, long hair, long bangs, braid, sweater, turtleneck , long sleeves, long skirt, apron

• Choco, animal ears, long twintails, short bangs, sleveless dress, detached sleeves

Use monochrome, grayscale, comic to trigger the doujin stye, it does an excellent job at it.

You can experiment with adding rough ??? to some positions to trigger a more intense version.

It creates good quality color and grayscale images, and is very compatible with other character loras, make sure to lower the weight of the character loras to preserve the style. 0.8 or lower for character loras seems to work great.

Reviews, Hearts, Likes etc. are very appreciated!