One-Shot Illustrious

This is a personal merge I created with the intention of being easy to use and not require external resources or post-detail work to get decent results, hence the name One-Step. All example images don't use loras and haven't been upscaled, though some use automatic detailing with CLIPSeg (basically the equivalent of ADetailer). Check the suggested resources below for models used in the merge.

DISCLAIMER

While this model can generate SFW images, it is NSFW and is capable of generating explicit material. Use caution in your prompts if you're looking to avoid anything NSFW, but I can't guarantee a random NSFW image won't appear from time to time.

RECOMMENDED SETTINGS

These are the settings I use, but there could be something better. Feel free to experiment on your own.

Resolutions: 832x1216 / 768x1344 / 1024x1536 <- what I use (Caution: Higher resolutions significantly increase the chance of incorrect finger and toe count, as well as other oddities.)

Steps: ~25-35

CFG Scale: 7

Sampler: Euler Ancestral

Scheduler: Karras

Upscaler / Hires. fix (Optional): 0.3 strength, 2x upscale, 4x-Ultrasharp or SwinIR_4x (the other upscalers I tested tend to remove details of the original image)

ADetailer / CLIPSeg (Optional): Use as desired. It's often not needed to generate a decent looking image, but feel free to use if you want that extra bit of detail, are trying to correct errors, etc.

RECOMMENDED BASE PROMPTS

Same as above, this is just what I use. Feel free to experiment to your own preference.

Positive: masterpiece, best quality, amazing quality, detailed eyes

Negative: bad quality, worst quality, worst detail, sketch, censor, bar censor, extra fingers, extra toes, heart-shaped pupils

For a flat anime style, add "anime coloring" to positive and "realistic, photorealistic" to negative. If you're getting a wet-clothes look when you're not trying to or don't want to, add "sweaty clothes" to negative. If you don't want eyes visible in the image, remove "eyes" from prompt. As a reminder, the model will try very hard to include everything mentioned in the positive prompt, so anything you don't want needs to be excluded.