A desktop application for generating and editing image captions, with support for both OpenAI and JoyCaption.
⚠️ Warning: The current releases (as ofv0.2.3-alpha) are unstable and do have bugs.
I am in the middle of a major refactor and many things are currently broken, including:
- Image virtualization
- Large project stability
- Drag and drop support
Current basic functionality includes:
- Individual image captioning
- Batch captioning
- OpenAI API (+JoyCaption via vLLM)
- Manual caption editing
- Import/export kohya dataset
- Custom and preset captioning prompts
- Session persistence
- Image editor
- Support for OpenAI API and JoyCaption alpha two (vLLM)
- Pre-defined prompts for JoyCaption
- Keeps original files unmodified
- Single and batch processing
- Basic image editing
- Download the latest release or build from source (see development docs)
- Configure your captioning model:
- For OpenAI: Add your API key
- For JoyCaption: Set up vLLM (see docs/joycaption-alpha-two.md)
- Select your prompt (optimized for JoyCaption but allows any prompt)
- Import images or a folder
- Caption your images one at a time or in batches:
- Click generate to process the current image; or
- Select multiple images and click caption to process them in a batch
- Export images and txt files when you're done to train SD/Flux
- Basic image editing tools:
- Crop
- Rotate
- Flip
- "Separate Viewed Images" option helps you keep track of your progress when you use the batch captioning feature
- Reset to saved caption: Allows you to revert any caption to what it was when you first opened it
Since the app isn't signed, I recommend you look through the source code and build it on your own machine (see development docs). But that's a little annoying, so there are also installers for major platforms.
You can download the latest release here for your platform (it's built for darwin but compiled for linux and windows).
On Windows, it's easy to bypass the security prompt on the installer by clicking "more info" and then "run anyway":
Windows protected your PC Microsoft Defender SmartScreen prevented an unrecognized app from starting. Running this app might put your PC at risk. App: spacecat sage.exe Publisher: Unknown publisher
On macOS, there are a few more steps to open the app:
Overriding security settings to open an app is the most common way that a Mac gets infected with malware.
On your Mac, choose Apple menu > System Settings, then click Privacy & Security in the sidebar. (You may need to scroll down.)
Go to Security, then click Open.
Click Open Anyway.
This button is available for about an hour after you try to open the app.
Enter your login password, then click OK.
The app is saved as an exception to your security settings, and you can open it in the future by double-clicking it, just as you can for any authorized app.
MIT License for individuals only. Contribute for any other use.
Icon background from @umpontoseisig
Logo by yours truly