This model was trained on SDXL 1.0 . Recommended weight 0.7-0.8. Trigger word ShinozakiAI. Don't use for illegal purposes. Trained on 32 1024x1024 images, works fine in other common SDXL resolutions from my testing. Example images made with ZavychromaXL_v31. While the example images are somewhat cherry-picked, i've found it consistent.