π‘ Search "huggy" to find all my models on site.
π‘ Models are updated frequently, so be sure to check back and use the latest version.
π‘ DO NOT MIXED USE Illustrious lora with SDXL / PONY / FLUX model.
Settings commendation:
CFG Scale: 5.0
Steps: 29π To be fixed in next update:
change trigger word from hu99y to Takahashi_Ryuichi
Increase training steps from 2058 to 2500
π This lora is part of my goal to create 100 lora.
Description
{
"Image": "39",
"engine": "kohya",
"unetLR": 0.0004,
"clipSkip": 1,
"loraType": "lora",
"keepTokens": 1,
"networkDim": 32,
"numRepeats": 5,
"resolution": 1024,
"lrScheduler": "cosine_with_restarts",
"minSnrGamma": 3,
"noiseOffset": 0.10,
"targetSteps": 2058,
"enableBucket": true,
"networkAlpha": 16,
"optimizerType": "Adafactor",
"textEncoderLR": 0.00005,
"maxTrainEpochs": 42,
"shuffleCaption": true,
"trainBatchSize": 4,
"flipAugmentation": true,
"lrSchedulerNumCycles": 3
}