ユーザーLoRA
Training on Deliberate v1.1. Upscale on 4x_NMKD-Siax_200k. No triggers. Generated on Deliberate v2.0. All the defects an details were corrected using INPAINT. Without using Photoshop.
