README.md

metadata

title: Wan LoRA Trainer
sdk: docker

Wan 2.1 LoRA Trainer

This Space provides a ready-to-use environment for training a LoRA adapter for the Wan 2.1 image-to-video model.

How it Works

• Self-Contained: All required models are stored directly within this Space for maximum reliability.
• Automatic Startup: When the Space builds, it automatically executes the start.sh script to begin the training process. You can monitor the progress in the "Logs" section.

How to Train on Your Own Data

This Space is pre-configured to run with a test dataset. To train on your own data, simply upload your dataset files, create a .toml configuration file for it, and update the start.sh script to point to your new config.

Wan 2.1LoRA Trainer

This Space runs the Wan 2.1 LoRA training script. The required models are linked via the repository configuration above.

Simple GUI for Musubi Tuner (Wan 2.1 models only)

How to use GUI

• Download the repository by running in the command line: git clone https://github.com/Kvento/musubi-tuner-wan-gui

• To open the GUI just run Start_Wan_GUI.bat.

• All settings can be saved and loaded using the "Load Settings" and "Save Setting" buttons.

• More info about settings see in Wan2.1 documentation, Advanced Configuration, Dataset configuration guide.