license: creativeml-openrail-m
tags:
- stable-diffusion
- text-to-image
Introduction:
This model was trained on a digital painting style mainly with characters and portraits. The main objective is to train a model to be a tool to help with character design ideas.
V2:
Trained for an additional 5000 steps. Results will be much more stable and major improvement over V1. Don't forget to add the yaml file into your models directory.
V2 checkpoint filename is CharHelper_v2_ SDv2_1_768_step_8500.ckpt
Usage
This model tends to like the higher CFG scale range. 7-15 will bring good results. Images come out well if they are 756X756 resolution size and up.
A good prompt to start with is:
(a cyberpunk rogue), charhelper, ((close up)) portrait, digital painting, artwork by leonardo davinci, high detail, professional, masterpiece, anime, stylized, face, facial expression, inkpunk, professional anatomy, professional hands, anatomically correct, colorful
Negative: ((bad hands)), disfigured, distorted face, mutated, malformed, bad anatomy, mutated feet, bad feet, poorly drawn, ((odd proportions)), noise, blur, missing fingers, missing limbs, long torso, ((ugly)), text, logo, over-exposed, over-saturated, ((bad anatomy)), over-exposed, ((over-saturated)), (((weapon))), long neck, black & white, ((glowing eyes))
Just substitute what's in the beginning parenthesis with your subject. You can also substitute "((close up))" with "((mid range))" as well. These worked best for me, but I'm excited to see what everyone else can do with it.
Examples
Below are some examples of images generated using this model:
A Woman with Los Muertos Skull Facepaint:
Rugged Samurai Man:
Space Girl:
Raver Girl with HeadPhones:
CyberPunk Rogue:
Toon Animal:
Female Astronaut:
Japanese Samurai:
Bell Head:
Native American Chief:
CyberPunk Buddha:
Alien Boy:
Los Muertos Facepaint 2:
Robot Fighter:
Video Game Elf Character:
V1:
Trained for 3500 steps on SD v2.1 using TheLastBen's Fast Dreambooth. Usage: Use CharHelper in prompt to bring out the style. Other prompts that work well are 'Character Art', 'Close-up/Mid-range Character portrait', 'Digital Painting', Digital Illustration', 'Stylized', and 'anime'.
Still needs work with anatomy and full body images may need inpainting to fix faces but there are plans to fine-tune the model further in hopes to improve functionality.