First attempt at training a Lora.
Dataset of 190 HD images with various positions, clothes and haircuts to guarantee a flexible character, working with all Illustrious models.
NSFW supported, enjoy.
Recommanded strength : 0.8 - 0.9
Thank you for any feedbacks and support !
Description
selected epoch : 10
params of the training:
{
"lr": 0.0005,
"engine": "ai-toolkit",
"epochs": 15,
"ecosystem": "sdxl",
"keepTokens": 1,
"networkDim": 32,
"numRepeats": 2,
"resolution": 1024,
"lrScheduler": "cosine",
"minSnrGamma": 5,
"noiseOffset": 0.1,
"networkAlpha": 32,
"optimizerType": "adafactor",
"shuffleTokens": true,
"textEncoderLr": 0.00005,
"flipAugmentation": true,
"trainTextEncoder": true
}
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.