Pinterest Version The name says it all xD This version is my personal go to model for generating reference images. I draw traditionally (by hand) and I used to always look out for images to draw on Pinterest, this model captures basically a simple style with balanced colors, linework, and visible details similar to what any image I would choose to draw would have. This version was a mix of various loras, Amanatsu + Indexed and previous models I've uploaded. I have been working for months on merges, testing various styles and giving out variety. I am happy with this outcome, I feel as if I have finally achieved a proper REFERENCE generator model, which was the sole reason why I got into AI and merges. I'd like to thank all the new followers I have been getting, and to those who gift me buzz and download my models, specially to those who share their generations with them. It helps me improve a lot ♥ -------------- My next goal is to make my own Comic model but that will possibly take a while to achieve, I don't have the knowledge nor the resources to TRAIN my own model from scratch or even add more training to an existing model, but hopefully I learn soon, since I would like to make a proper Comic model that carries with it various styles leaning heavily to just comics and not anime.

Aristo

aristo_inkject.safetensors

InkJect

AniWest

Pinterest

maGenta

Aero

Satura92

Comikal

A merge I made using bemixIllustrious + comijc.I tested various mixes, my main goal is to use them with my lora merges, so I can boost the quality, and properly achieve a style for my personal use. (Use my generated images to draw on my spare time manually).Simply using a lora with good base models, I haven't been able to reach my goal. I've managed to land on an iteration that finally satisfies what I envisioned. I will add 2 new lora merges to the section that are recommended with this Checkpoint. Some of the reference images will are made using these loras, so make sure you check under each one. I will also add reference images using just the base model.

Files with SHA256: 309dcd6261aa51c546efac3ee51452b3d8e9b5c1878f96984b516fdf76e731c1