Trained on my local Workstation on almost 100k highly curated images. Use Danbooru style tags on inference
Just some Anime finetune I did on SDXL