Could the author release a training script for training LoRA or fine-tuning based on Diffusers?

[wangqixun]

Hello, thank you for your great open-source work! We are the author team of InstantID, InstantX, and have noticed the powerful text-to-image generation capabilities of Lumina-T2X. We would like to adapt this model, including other plugins from InstantX.

Could you provide a training script based on diffusers for training LoRA or fine-tuning? With this script, we can quickly adapt, and it will also facilitate developers familiar with diffusers to rapidly develop the Lumina ecosystem.