Stable Diffusion and LLMs offline on your own hardware
No web server to run, additional requirements to install or technical knowledge required.
Just download the compiled package and start generating AI Art!
AI Runner is a multi-modal AI application which allows you to run open-source large language models and AI image generators on your own hardware.
Feature | Description |
---|---|
🗣️ Communication | |
✅ Voice-based chatbot conversations | Have conversations with a chatbot using your voice |
✅ Text-to-speech | Convert text to spoken audio |
✅ Speech-to-text | Convert spoken audio to text |
✅ Vision-to-text | Extract text from images |
✅ Text generation with LLMs | Generate text using large language models |
🎨 Image Generation | |
✅ Stable Diffusion (all versions) | Generate images using Stable Diffusion |
🔜 Kandinsky | Generate images using Kandinsky |
✅ Near Real-Time Drawing | Draw and generate images in near real-time |
✅ Text to Image (aka TextToImage / Txt2Img) | Generate images from textual descriptions |
✅ Image to Image (aka ImageToImage / Image2Image) | Generate images based on input images |
🖼️ Image Manipulation | |
✅ Inpaint and Outpaint | Modify parts of an image while maintaining context |
✅ Pix2Pix | Transform images from one domain to another |
✅ Depth to Image (aka DepthToImage / Depth2Img) | Generate images from depth maps |
✅ Controlnet | Control image generation with additional input |
✅ LoRA | Efficiently fine-tune models with LoRA |
✅ Textual Embeddings | Use textual embeddings for image generation control |
🔜 Upscale with GFPGAN | Use textual embeddings for image generation control |
🔧 Utility | |
✅ Run offline, locally | Run on your own hardware without internet |
✅ Fast generation | Generate images in ~2 seconds (RTX 2080s) |
✅ Run multiple models at once | Utilize multiple models simultaneously |
✅ Drawing tools | Built-in tools for drawing and image manipulation |
✅ Image filters | Apply various filters to images |
✅ Dark mode | Comfortable viewing experience in low-light environments |
✅ Infinite scrolling canvas | Seamlessly scroll through generated images |
✅ NSFW filter toggle | Help control the visibility of NSFW content |
✅ NSFW guardrails | Help prevent generation of harmful content |
✅ Standard Stable Diffusion settings | Easily adjust standard Stable Diffusion parameters |
✅ Fast load time, responsive interface | Enjoy a smooth and responsive user experience |
✅ Pure python | No reliance on a webserver, pure python implementation |
For emoji support on Ubuntu install the Noto Color Emoji font:
sudo apt install fonts-noto-color-emoji
Install portaudio
, libxcb-cursor
, xclip
and espeak
sudo apt-get install libportaudio2 libxcb-cursor0 espeak xclip
If you are on Windows, first install the following dependencies:
pip install pypiwin32
pip install airunner
If you want to install AI Runner from source, you can do so using the following command:
git clone -b develop https://github.com/Capsize-Games/airunner.git
cd airunner && pip install -e .
The current controlnet_aux
Huggingface library has a bug which is awaiting a patch (submitted by me). Until it is accepted,
install controlnet manually from my fork:
git clone -b develop https://github.com/w4ffl35/controlnet_aux.git
cd controlnet_aux
pip install -e .
If you install from pypi or source, uninstall opencv-python
(we use opencv-python-headless
instead)
pip uninstall opencv-python
There are many ways to run AI Runner, depending on your operating system and how you installed it.
Unzip the AI Runner zip file which you downloaded from itch.io
cd airunner
./airunner
cd airunner
airunner.exe
Alternatively, you can use the itch.io launcher application which simplifies the process of downloading and running AI Runner.
cd dist/airunner
./airunner
python3 -m airunner
cd src/airunner
python main.py
Linux:
docker-compose up linux
Windows:
docker-compose up windows
See the installation wiki page for more information
Instructions on how to use AI Runner can be found in the wiki
Clone this repository
git clone https://github.com/Capsize-Games/airunner.git
cd airunner
pip install -e .
pip install pyinstaller
bash build.dev.sh
docker-compose run build
docker-compose run devbuild
Run a specific test
python -m unittest src/airunner/tests/test_draggable_pixmap.py
Test coverage is currently low, but the existing tests can be run using the following command:
python -m unittest discover tests
Run tests with coverage tracking:
coverage run --source=src/airunner --omit=__init__.py,*/GFPGAN/*,*/data/*,*/tests/*,*_ui.py,*/enums.py,*/settings.py -m unittest discover src/airunner/tests
To see a report in the terminal, use:
coverage report
For a more detailed HTML report, run:
coverage html
View results in htmlcov/index.html
.