2n5 / mikenko2222 Artist Style LoRA (Noob Vpred)

slightly larger dataset (40 images) - first autotagged and then manually checked, removing unnecessary tags/adding missing tags to images as needed.

pre-downscaled images before training using custom image downscaling script

• sd-scripts does this already, but I wanted to achieve a higher downscale quality, using a script that downscales the images using a combination of Lanczos (40 %) and LSID (60%) algorithm by blending images from both algorithms together at mentioned percentages.

  • LSID from paper: https://arxiv.org/abs/2510.24334

enabled flipped arguments, as there were no images in dataset where flipping the image would cause symmetrically issues and to diversify dataset

using edm instead of minsnr for timestep correction/training

final lora model is a 50/50 blend of epoch 56 and epoch 60, using the lora-merger in sd-scripts.

resized lora using this script, with a threshold of 2.95, shaving off 30-35 MBs and reducing overfitting:

• https://github.com/elias-gaeros/resize_lora