After the first lines of config.json file, dedicated to general configuration, you will find the 3 parts of illustration presets.
First one is the preset arrays of style-illustrations.
Each element defines a preset with :
{
"name": "s1",
"description": "Illustration #1 for all styles, A cat",
"metadata": {
"Base Model": "juggernautXL_v8Rundiffusion.safetensors",
"Prompt": "A cat",
"Seed": "314159"
}
},
A preset contains :
- a name
- a description
- a metadata object that describes how the picture must be generated
Metadata part may contain that parameters :
{
"name": "s1",
"description": "Illustration #1 for all styles, A cat",
"metadata": {
"Base Model": "juggernautXL_v8Rundiffusion.safetensors",
"Prompt": "A cat",
"Seed": "314159",
"ADM Guidance": "(1.5, 0.8, 0.3)",
"Sampler": "dpmpp_2m_sde_gpu",
"Scheduler": "karras",
"Performance": "Quality",
"Sharpness": 5,
"Guidance Scale": 2,
"Refiner Model": "juggernautXL_v8Rundiffusion.safetensors",
"Refiner Switch": 0.5,
"Lora 1": "sd_xl_offset_example",
"Weight Lora 1": 1
}
},
Add, edit or remove presets as you like.
You must restart FoooXus app to take into account these file updates
Models and Loras presets below in the file are similar