This is a Junji Ito art style lora. Trigger word for the WAI 1.0 version is junji_ito_uzumaki. Trigger word for the WAI 2.0 version is junji_ito_uzu000. Replace the zeros with any number from 001 to 665. It's recommended that you use a lora loader that can adjust model strength, clip strength, and individual block strength. The lora was trained on the waiNSFWIllustrious_v120 checkpoint, so you should use that model for generating images or else you'll might get bad results. See the example images for the ComfyUI workflow.

Junji Ito Uzumaki Style

This is improved version with recaptioned dataset and more granular training of lora modules. The trigger word is junji_ito_uzu000. Replace the zeros with any number from 001 to 665.

Files with SHA256: 97dcf6f8366aaa5cd7ba63a7eaf0eb0cb3552a256dfbb4e89e091aa42aa810c2