/pyPromptChecker

A small script for AI images created by stable diffusion webui

Primary LanguagePythonGNU General Public License v3.0GPL-3.0

pyPromptChecker

A small script designed for analyzing AI-generated images created by AUTOMATIC/stable-diffusion-webui.
It extracts incomprehensible strings embedded within image files, formatting them into human-readable.
The formatted data can be exported as JSON and subsequently imported from the JSON output.
Additionally, the script offers in-data search and basic file management functionalities.
All of these features are accessible even without AUTOMATIC/stable-diffusion-webui.

Screenshots

Main

More screenshots here.

Features

  • Extract creation data, formatting and display it.
  • Any number of files can be processed simultaneously (Tested up to 1500 files).
  • Image file move and delete with single click.
  • JSON export and import.
  • Make List of model hash from .safetensors and .ckpt files.
  • Tab navigation with filename or thumbnails.
  • Search with various conditions.

See more details here.

Requirements

pyPromptChecker binary edition no longer has any requirements.

I decided to give up distribute binary editions.
Because very difficult to avoid false positives of Windows Defender.
If you want to continue using you need to install at least python 3.x

  • Python 3.x
  • pillow (PIL)
  • pypng
  • PyQt6
  • pyqtdarktheme

As follows for interrogate features.

  • onnxruntime
  • numpy
  • opencv-python
  • huggingface_hub

Installation

pyPromptChecker binary edition (for Linux and Windows users)

1. Download the binary packages from the 'Releases'.
2. Extract pyPromptChecker directory to any location of your choice.
3. Optionally, desktop files for enable drag-and-drop functionality.
4. Execute pyPromptChecker by double-click.

pyPromptChecker source code edition

For Linux

git clone https://github.com/yui-tan/pyPromptChecker
cd pyPromptChecker/install
./install.sh

For Windows

  1. Make sure you've got Python 3.x hanging around.
  2. Download the packages from the 'Releases' or code > download ZIP.
  3. Extract the packages to a location of your choice.
  4. Run command in PowerShell.
Set-ExecutionPolicy -ExecutionPolicy Unrestricted -Scope Process
cd \path\to\pyPromptChecker\install
./install.ps1
Set-ExecutionPolicy -ExecutionPolicy Undefined -Scope Process

Usage

Source code edition

mikkumiku -a, --ask  
# Open directory choose dialog.
mikkumiku -f [FILEPATH ...], --filepath [FILEPATH ...]  
# Send file or directory paths to the script.

Roadmap

  • Tab navigation with thumbnails
  • Listview
  • Diff mode
  • Filtering and searching tabs various conditions
  • Add support for other image file formats
    • jpeg
    • webp
  • Add support for import JSON
  • Automated LoRa creation
    • Positive prompt export to txt files.
    • Implement interrogate feature.
    • Import configure from files.
    • Linking sd-scripts.
  • Export Data to Stable-Diffusion-Webui via API
  • Get marry to Miku

Author

Yui-tan / Yuiyui

Nutjob who loves Hatsune Miku.
napier2.718281828@gmail.com
https://civitai.com/user/Yui_tan
https://chichi-pui.com/users/yuiyui20170927

Licence

This script created under GPLv3.