English / 日本語

This repository contains training, generation and utility scripts for Stable Diffusion and other image generation models.

We are grateful to the following companies for their generous sponsorship:

If you find this project helpful, please consider supporting its development via GitHub Sponsors. Your support is greatly appreciated!

• Version 0.10.3 (2026-04-02):

  • Stability when training with fp16 on Anima has been further improved. See PR #2302 for details. We deeply appreciate those who reported the issue.

• Version 0.10.2 (2026-03-30):

  • LECO training for SD/SDXL is now supported. Many thanks to umisetokikaze for PR #2285 and PR #2294.
    • Please refer to the documentation for details.
  • networks/resize_lora.py has been updated to use torch.svd_lowrank, resulting in a significant speedup. Many thanks to woct0rdho for PR #2240 and PR #2296.
    • It is enabled by default. You can specify the number of iterations with the --svd_lowrank_niter option (default is 2, more iterations will improve accuracy). Setting it to 0 will revert to the previous method. Please check --help for details.
  • LoKr/LoHa is now supported for SDXL/Anima. See PR #2275 for details.
    • Please refer to the documentation for details.
  • Multi-resolution datasets (using the same image resized to multiple bucket sizes) are now supported in SD/SDXL training. We also addressed the issue of duplicate images with the same resolution being used in multi-resolution datasets. See PR #2269 and PR #2273 for details.
    • Thanks to woct0rdho for the contribution.
    • Please refer to the English documentation / Japanese documentation for details.
  • Stability when training with fp16 on Anima has been improved. See PR #2297 for details. However, it still seems to be unstable in some cases. If you encounter any issues, please let us know the details via Issues.
  • Other minor bug fixes and improvements were made.

• Version 0.10.1 (2026-02-13):

  • Anima Preview model LoRA training and fine-tuning are now supported. See PR #2260 and PR #2261.
  • Many thanks to CircleStone Labs for releasing this amazing model, and to duongve13112002 for submitting great PR #2260.
  • For details, please refer to the documentation.

• Version 0.10.0 (2026-01-19):

  • sd3 branch is merged to main branch. From this version, FLUX.1 and SD3/SD3.5 etc. are supported in the main branch.
  • There are still some missing parts in the documentation, so please let us know if you find any issues via Issues etc.
  • The sd3 branch will be maintained as a development branch synchronized with dev for the time being.

• Stable Diffusion 1.x/2.x
• SDXL
• SD3/SD3.5
• FLUX.1
• LUMINA
• HunyuanImage-2.1

• LoRA training
• Fine-tuning (native training, DreamBooth): except for HunyuanImage-2.1
• Textual Inversion training: SD/SDXL
• Image generation
• Other utilities such as model conversion, image tagging, LoRA merging, etc.

• LoRA Training Overview
• Dataset config / Japanese version
• Advanced Training
• SDXL Training
• SD3 Training
• FLUX.1 Training
• LUMINA Training
• HunyuanImage-2.1 Training
• Fine-tuning
• Textual Inversion Training
• ControlNet-LLLite Training / Japanese version
• Validation
• Masked Loss Training / Japanese version

• Image generation / Japanese version
• Tagging images with WD14 Tagger / Japanese version

This repository provides recommended instructions to help AI agents like Claude and Gemini understand our project context and coding standards.

To use them, you need to opt-in by creating your own configuration file in the project root.

Quick Setup:

1. Create a CLAUDE.md and/or GEMINI.md file in the project root.

2. Add the following line to your CLAUDE.md to import the repository's recommended prompt:

   @./.ai/claude.prompt.md

   or for Gemini:

   @./.ai/gemini.prompt.md

3. You can now add your own personal instructions below the import line (e.g., Always respond in Japanese.).

This approach ensures that you have full control over the instructions given to your agent while benefiting from the shared project context. Your CLAUDE.md and GEMINI.md are already listed in .gitignore, so they won't be committed to the repository.

Python 3.10.x and Git:

• Python 3.10.x: Download Windows installer (64-bit) from https://www.python.org/downloads/windows/
• git: Download latest installer from https://git-scm.com/download/win

Python 3.11.x, and 3.12.x will work but not tested.

Give unrestricted script access to powershell so venv can work:

• Open an administrator powershell window
• Type Set-ExecutionPolicy Unrestricted and answer A
• Close admin powershell window

Open a regular Powershell terminal and type the following inside:

git clone https://github.com/kohya-ss/sd-scripts.git
cd sd-scripts

python -m venv venv
.\venv\Scripts\activate

pip install torch==2.6.0 torchvision==0.21.0 --index-url https://download.pytorch.org/whl/cu124
pip install --upgrade -r requirements.txt

accelerate config

If python -m venv shows only python, change python to py.

Note: bitsandbytes, prodigyopt and lion-pytorch are included in the requirements.txt. If you'd like to use another version, please install it manually.

This installation is for CUDA 12.4. If you use a different version of CUDA, please install the appropriate version of PyTorch. For example, if you use CUDA 12.1, please install pip install torch==2.6.0 torchvision==0.21.0 --index-url https://download.pytorch.org/whl/cu121.

Answers to accelerate config:

- This machine
- No distributed training
- NO
- NO
- NO
- all
- fp16

If you'd like to use bf16, please answer bf16 to the last question.

Note: Some user reports ValueError: fp16 mixed precision requires a GPU is occurred in training. In this case, answer 0 for the 6th question: What GPU(s) (by id) should be used for training on this machine as a comma-separated list? [all]:

(Single GPU with id 0 will be used.)

The file does not contain requirements for PyTorch. Because the version of PyTorch depends on the environment, it is not included in the file. Please install PyTorch first according to the environment. See installation instructions below.

The scripts are tested with PyTorch 2.6.0. PyTorch 2.6.0 or later is required.

For RTX 50 series GPUs, PyTorch 2.8.0 with CUDA 12.8/12.9 should be used. requirements.txt will work with this version.

To install xformers, run the following command in your activated virtual environment:

pip install xformers --index-url https://download.pytorch.org/whl/cu124

Please change the CUDA version in the URL according to your environment if necessary. xformers may not be available for some GPU architectures.

Linux or WSL2 installation steps are almost the same as Windows. Just change venv\Scripts\activate to source venv/bin/activate.

Note: Please make sure that NVIDIA driver and CUDA toolkit are installed in advance.

To install DeepSpeed, run the following command in your activated virtual environment:

pip install deepspeed==0.16.7 

When a new release comes out you can upgrade your repo with the following command:

cd sd-scripts
git pull
.\venv\Scripts\activate
pip install --use-pep517 --upgrade -r requirements.txt

Once the commands have completed successfully you should be ready to use the new version.

If you want to upgrade PyTorch, you can upgrade it with pip install command in Windows Installation section.

The implementation for LoRA is based on cloneofsimo's repo. Thank you for great work!

The LoRA expansion to Conv2d 3x3 was initially released by cloneofsimo and its effectiveness was demonstrated at LoCon by KohakuBlueleaf. Thank you so much KohakuBlueleaf!

The majority of scripts is licensed under ASL 2.0 (including codes from Diffusers, cloneofsimo's and LoCon), however portions of the project are available under separate license terms:

Memory Efficient Attention Pytorch: MIT

bitsandbytes: MIT

BLIP: BSD-3-Clause