LoRA text2image fine-tuning - SedatAl/manga-LoRa

These are LoRA adaption weights for stabilityai/stable-diffusion-xl-base-1.0. The weights were fine-tuned on the Chan-Y/Manga-Drawings dataset.

Special VAE used for training: madebyollin/sdxl-vae-fp16-fix.

How to use

resolution=1024 
train_batch_size=1 
gradient_accumulation_steps=4 
max_train_steps=528 
learning_rate=1e-4 
lr_scheduler="cosine" 
lr_warmup_steps=100