This repository provides tools and scripts for training and fine-tuning Lightricks' LTX-Video (LTXV) model. It enables LoRA training, full fine-tuning, and video-to-video transformation workflows on custom datasets.
All detailed guides and technical documentation have been moved to the docs/
directory:
- ⚡ Quick Start Guide
- 🎬 Dataset Preparation
- 🛠️ Training Modes
- ⚙️ Configuration Reference
- 🚀 Training Guide
- 🔧 Utility Scripts
- 🛡️ Troubleshooting Guide
- 08.07.2025: Added support for training IC-LoRAs (In-Context LoRAs) for advanced video-to-video transformations. See the training modes doc for more details. Pretrained control models: Depth, Pose, Canny.
- 06.05.2025: Added support for LTXV 13B. An example training configuration can be found here.
- Cakeify LoRA: Transforms videos to make objects appear as if they're made of cake. (Dataset)
- Squish LoRA: Creates a playful squishing effect. (Dataset)
- Depth Map Control: Generate videos from depth maps.
- Human Pose Control: Generate videos from pose skeletons.
- Canny Edge Control: Generate videos from Canny edge maps. (Canny Control Dataset)
These examples demonstrate how you can train specialized video effects and control adapters using this trainer. Use these datasets as references for preparing your own training data.
We welcome contributions from the community! Here's how you can help:
- Share Your Work: If you've trained interesting LoRAs or achieved cool results, please share them with the community.
- Report Issues: Found a bug or have a suggestion? Open an issue on GitHub.
- Submit PRs: Help improve the codebase with bug fixes or general improvements.
- Feature Requests: Have ideas for new features? Let us know through GitHub issues.
Have questions, want to share your results, or need real-time help?
Join our community Discord server to connect with other users and the development team!
- Get troubleshooting help
- Share your training results and workflows
- Collaborate on new ideas and features
- Stay up to date with announcements and updates
We look forward to seeing you there!
Parts of this project are inspired by and incorporate ideas from several awesome open-source projects:
Happy training! 🎉