This repository is an old fork and is no longer maintained. Active development has moved to a new, independent repository with an expanded feature set and a focus on a broader AI training ecosystem.
(This repository will be archived and made read-only.)
(Original README content below)
A LoRA training system built on Derrian Distro & Kohya SS with interactive Jupyter widget interfaces. Supports local and cloud deployment on VastAI, RunPod, and similar platforms.
| Python Version | License | Discord | Twitch | Support |
|---|---|---|---|---|
 |
 |
 |
 |
 |
Key Features:
- Widget-based configuration interface
- Async based Uploading so your images don't become dreams.
- Integrated dataset preparation and tagging tools
- Training parameter calculator and optimization
- Multiple LoRA variants and optimizers
- Cross-platform compatibility
- Huggingface Uploads for datasets AND loras.
- Coming eventually when my brain lets me: FULL MODEL TRAINING! It's supported with KohyaSS, but our widget interface is a bit primitive and I need to understand more before bringing that in.
We are STILL in heavy development. New features in theory SHOULD WORK, but are hard to catch.
This branch includes experimental features that are available in the Kohya backend but may not be fully tested in our setup:
- 🔬 FLUX training - Available in Kohya, integration status unknown
- 🧬 SD3/SD3.5 training - Available in Kohya, integration status unknown
- 🌟 Lumina2 training - Available in Kohya, integration status unknown
- 🔧 Latest bug fixes and performance improvements
- ⚡ Enhanced upload widgets (fixed cache issues)
- ⚡ Language Cleanup Cleaned up a lot of marketing speak and started the roadmap to check inconsistencies on missing content. Note: These experimental features exist in the underlying Kohya scripts but haven't been thoroughly tested with our widget system. Use at your own risk and expect possible issues. If they look like they're exposed in our widget setup, there is no saying if they current work due to our unified setup. We're working on fast trying to get functionality quickly. If you have any issues please report them to the issues area.
What You Need
- GPU: Nvidia (For built-in CUDA support) or AMD Cards for ROCm. (Future Support for ARC and otherwise coming)
- Python: Version 3.10+ required
- Platform: Windows or Linux based Operating Systems.
More details on installation can be found here Quick Start Guide or in our Installation Setup.
You will need Git and Python 3.10+. If you don't have python, you can install Python 3.10+ from Python's main website here. Our set up prefers 3.10.6 at a minimum.
Install Git if needed:
- Windows: Download from git-scm.com
- Mac:
xcode-select --installin Terminal - Linux:
sudo apt install git(Ubuntu/Debian)
Main Installation Steps:
# 1. Clone the repository
git clone https://github.com/Ktiseos-Nyx/Lora_Easy_Training_Jupyter.git
cd Lora_Easy_Training_Jupyter
# 2. Run the installer (downloads ~10-15GB)
python ./installer.py
# For detailed installation output (recommended for troubleshooting):
python ./installer.py --verbose
# or: python ./installer.py -v
# Alternative for Mac/Linux:
chmod +x ./jupyter.sh && ./jupyter.sh(If Jupyter is NOT running)
jupyter notebook
# Or: jupyter labThe system uses three specialized notebooks:
Dataset_Maker_Widget.ipynb- Prepare images and captions for trainingUnified_LoRA_Trainer.ipynb- Configure and execute LoRA trainingUtilities_Notebooks.ipynb- Calculate parameters and resize trained models
For detailed workflow instructions, see our Quick Start Guide and Notebook Workflow Guide.
For more help and support please check Troubleshooting this has more comprehensive information. If you're a developer, we're working on our testing notebook, there is one in the wings of the /tests folder, but it has older code and may not match what is current running.
Before asking for help, please review our Support Guidelines. We're happy to assist, but effective troubleshooting requires your participation - this means running the basic diagnostic commands and providing complete error information. Cherry-picking troubleshooting steps won't lead to solutions!
Windows Users: If you encounter Rust compilation errors during safetensors installation, this is not related to our notebook setup. It's a common Python packaging issue on Windows. Feel free to reach out on our Discord for assistance - we're happy to help guide you through the solution!
Getting Help: - ✅ Official Support: GitHub Issues or Our Discord - ❌ No Support: Random discords, Reddit DMs, social media comments, etc. - 📚 Self-Help: Check our comprehensive docs/ folder first - 🎯 Submodule Issues: Feel free to blame us on the original repos (kohya-ss, LyCORIS, etc.)!
- Built on the Shoulders of Giants This project builds upon and integrates the excellent work of:
- Jelosus2's LoRA Easy Training Colab - Original Colab notebook that inspired this adaptation
- Derrian-Distro's LoRA Easy Training Backend - Core training backend and scripts as well as the forked Lycoris Repository and CAME/REX optimization strategies.
- HoloStrawberry's Training Methods - Community wisdom and proven training techniques as well as foundational Google Colab notebooks.
- Kohya-ss SD Scripts - Foundational training scripts and infrastructure
- Linaqruf - Pioneer in accessible LoRA training, creator of influential Colab notebooks and training methods that inspired much of this work
- AndroidXXL, Jelosus2 - Additional Colab notebook contributions that made LoRA training accessible
- ArcEnCiel - Ongoing support and testing as well as Open Source AI Generative Models.
- Civitai - Platform for Open Source AI Content
- LyCORIS Team - Advanced LoRA methods (DoRA, LoKr, etc.)
Special thanks to these creators for making LoRA training accessible to everyone!
Found a security issue? Check our Security Policy for responsible disclosure guidelines.
MIT License - Feel free to use, modify, and distribute. See LICENSE for details.
We welcome contributions! Check out our Contributing Guide for details on how to get involved. Feel free to open issues or submit pull requests on GitHub.
Made with ❤️ by the community, for the community.
