This Style LoRA was trained exclusively on the Illustrious model, with high step counts to ensure optimal usage. It is specifically designed for use with Illustrious-based models and may not perform as expected on other checkpoints. For best results, use a LoRA weight between 0.7 and 1.1. Lower weights may reduce stylistic influence, while higher weights can introduce distortions. Higher sampling steps (typically 30 or more) are recommended to fully realize the intended style and detail.

Files with SHA256: 5e5b9a64ac06fcb5ff83a419ed22d0328f7b48a4fca2b68b7ae66a49ad3ee867