Archived

Replaced with a new CLI tool that does its job way better and has more options. See joy-caption-cli

Stable Cheese Interrogator

A minimal app using clip-interrogator and blip for captioning lots of images quickly.

The application served to caption custom images which had only basic names like IMAGE_001.png and such.

This program has an optionated way on captioning images. May open it up more by implementing some more parameters for customizing captioning.

How to use?

Well you can either use it as a CLI tool using following commands:

Installation

py -m venv venv
py -m pip install -r requirements.txt

Output JSON in Terminal for CLI usage

py main.py C:\Users\<username>\OneDrive\Pictures\test_picture_01.png C:\Users\<username>\OneDrive\Pictures\test_picture_02.png # Caption as many images as you like.

Output JSON for images in a folder

py main.py C:\Users\<username>\OneDrive\Pictures\test_picture_01.png C:\Users\<username>\OneDrive\Pictures\test_picture_02.png # Caption as many images as you like.

Output Files in directory

This creates an "outputs" directory with a new subfolder (current date & time) where images get saved with a new captioned filename and their respective extension.

py main.py C:\Users\<username>\OneDrive\Pictures\test_picture_01.png C:\Users\<username>\OneDrive\Pictures\test_picture_02.png --output_files

Or by directly running the gradio app:

py app.py # Starts an gradio instance which can be accessed via browser

App.py

Frontend for hugginface spaces. Simplify CLI tool usage.

• Gallery Image of Sailboat by Michelle McEwen on Unsplash