Welcome to uruti_2388 LoHA repository

This model has been trained using kohya_ss SD scripts via kohya_ss GUI.

This model uses AnythingV3 and Waifu Diffusion's kl-f8-anime2 VAE for training.

Total of 12 Epochs with 84 Images total scraped from Gelbooru and Danbooru at a constant learning rate of 0.0001 unet and 0.00005 text encoder. Buckets were disabled for full image context and flip aug was used.