license: creativeml-openrail-m
datasets:
- Nacholmo/controlnet-test-darkest-color
- yuvalkirstain/pexel_images_lots_with_generated_captions
tags:
- controlnet
base_model: runwayml/stable-diffusion-v1-5
New version:
https://huggingface.co/Nacholmo/controlnet-qr-pattern-v2
Controlnet model for use in qr codes
Conditioning only 15% of the pixels closest to black, so as not to affect the luminance of the rest of the image. Also manteing the color to some degree.
Lasted version on the Automatic1111-Compatible folder
Move the .yaml to extensions/sd-webui-controlnet/models
If you want to use it in diffusers (Not automatic111), the recommended checkpoint is the 11500
Recommended parameters
Recommended parameters:
Steps: +50, Size: 920x920,
ControlNet:
preprocessor: none,
weight: 1,
starting/ending: (0, 0.75),
control mode: Balanced
Play arround with the starting step, 0 to 0.25 its the sweetspot, if it start at 0 the qr has priority, the higher you raise them, the stronger the prompt gets
To achieve optimal results with the Hires fix, ensure that your resolution is set to a minimum of 600x600. Additionally, set the denoising strength to a minimum of 0.7 and the hires steps to at least 20.
Example prompt parameters:
1girl, solo, red flower, print dress, bow, year of the ox, food print, short hair, black hair, looking at viewer, bangs, long sleeves, red bow
Negative prompt: (KHFB, AuroraNegative),(Worst Quality, Low Quality:1.4)
Steps: 40, Sampler: DPM++ 2M SDE Karras, CFG scale: 6, Seed: 1463218996, Size: 920x920, Model hash: b42b09ff12, Model: cetusMix_v4, Clip skip: 2, Version: 0ddf613, Parser: Full parser, ControlNet 0: "preprocessor: none, model: 11500_color [c4220211], weight: 1, starting/ending: (0, 0.75), resize mode: Crop and Resize, pixel perfect: False, control mode: Balanced, preprocessor params: (-1, -1, -1)"
QR Pattern and QR Pattern sdxl were created as free community resources by an Argentinian university student. Training AI models requires money, which can be challenging in Argentina's economy. If you find these models helpful and would like to empower an enthusiastic community member to keep creating free open models, I humbly welcome any support you can offer through ko-fi here https://ko-fi.com/nacholmo