Overview

Welcome to my very first Object LoRA! This is an experimental project trained on a single image to recreate the iconic Black Lotus artifact from Magic: The Gathering. Because of its unique, minimal training data, the model can be a bit wild but yields fantastic and creative results with the right prompts.

Model Specifications

• Base Model: Illustrious

• Trigger Word: bl-mtg

• Recommended Strength: 1.0

Known Issues & Prompting Tips

Given the experimental nature of this single-image training, you may encounter a couple of specific quirks:

• Head Placement: The Lotus sometimes tries to spawn on the character's head as a hair accessory.

  • The Fix: To control the object's placement, use precise action tags. Adding holding_flower or other interactive tags (like looking at flower, smelling flower) significantly limits this bug and forces the artifact into the character's hands or the correct context.

• Duplication: The model occasionally generates multiple Black Lotuses in the same scene. This is a known side-effect of the experimental training and cannot be completely fixed by prompting yet.

Let's See Your Creations!

As this is my first foray into training an object LoRA, your interaction means a lot. Please share your generated images, leave a review, and drop your feedback in the comments. I would love to see what kind of magic you can summon with this artifact!