GithubHelp home page GithubHelp logo

maagic6 / sdimv Goto Github PK

View Code? Open in Web Editor NEW
16.0 1.0 1.0 104 MB

Standalone program for viewing metadata of Stable Diffusion generated images

License: GNU General Public License v3.0

Python 100.00%
metadata-extraction stable-diffusion image-viewer prompt prompts pyqt6

sdimv's Introduction

SDIMV

Release Downloads License: GPLv3

Standalone program for viewing metadata of stable diffusion generated images

Download (Windows)

Download from releases

Features

  • Supports AUTOMATIC1111 web UI, NovelAI, and ComfyUI* png images
  • Supports AUTOMATIC1111 web UI jpg images
  • Supports PixAI mp4 animations
  • Resizing and rearranging of UI
  • Drag and drop images/folders

*Note: SDIMV has not been tested on many ComfyUI images so expect bugs/crashes

Build (Linux/MacOS) (need to handle platform-specific code + update requirements.txt first)

pip install -r requirements.txt
pip install pyinstaller
pyinstaller SDIMV.spec~~

Executable can be found in /dist after building

Acknowledgements

sdimv's People

Contributors

maagic6 avatar thiagojramos avatar

Stargazers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

Watchers

 avatar

Forkers

thiagojramos

sdimv's Issues

Bug: Lora's are reported twice in the lora field

I didn't try this with a lot of images, but with this prompt:
Cosmic Copse, a small grove of trees under a sky filled with a beautiful cosmic nebula, under the ever-flickering, hazy sky, fostering strangely symmetrical forms, illustration,flat color,Flat style,(Masterpiece),(best quality),<lora:Landscape illustration:0.7>, (Masterpiece:1.3) (best quality:1.2) (high quality:1.1)
I get this in my Lora output:
<lora:Landscape illustration:0.7> <lora:Landscape illustration:0.7>

Oh, also, you may want to change Lora to LoRA. To be technically accurate I guess :)

Feature request: Automatically view image data

If I open the viewer by draggin/dropping files onto the .exe, it doesn't automatically view any item.
It would be nice if it viewed the first item in the list automatically.

The same goes for when dragging/dropping files onto it, or pressing BROWSE to browse for files.
If the list has only one entry, it should always display this entry as if the VIEW button is pressed.

Feature suggestion: SETTINGS menu, + option to maintain aspect ratio of image

I suggest adding a SETTINGS-button to the right of the BROWSE button.
In here, we can start populating with options.

It's great if you can support new functionality as options, in case people have other preferences.

The feature I would suggest adding to start is the option on how the image should be formatted.
Should it always maintain aspect ratio, or should it zoom + crop. Right now it defaults to zoom + crop.

I prefer maintaining it's aspect ratio, but not everyone may.

Feature suggestion: Global arrow hotkeys, or keyboard to select text

Using arrows to select the next/previous image in the list of content works great, it's a good way of navigating the program.

I have 2 suggestions for further improvement to the interface. They are mutually exclusive, so both would not work at the same time (could be an option or mode though).

Option 1:
Arrow keys (up/down) to select images are global. You can always use up/down to go through images. Regardless of if your cursor is selecting text or what you have highlighted in the software.

Option 2:
Make the textboxes and fields (the data), selectable using arrow keys. If you click on the propt and use the mouse to select a section of text, it should activate this text field as a text input field, meaning that you should be able to use STHIFT + ARROW KEYS to manipulate the selected text. This is useful if you make a "sloppy" selection and want to fix it by just adding a letter or word to your selection.

Ideally, other common hotkeys like HOME/END/PGUP/PGDN, CTRL to jump words, and SHIFT to select etc, should also function.

Feature request: Single click to "View" image data

Right now you have to double-click to load an image. Is there a good reason for this?
Why not make it a single-click?
Also, for the arrow keys to activate when you click to move down the list of images.

Basically, automatically activating each image for easier browsing and comparing of data.

Feature request: Browse to folder of image

Minor feature request:
It would be nice with a link/path to the directory of the image, as well as a link to the image itself.

I could see it as 2 clickable links under the "Selected File" column.
Something like this:

Selected file: [FILE PATH.png]
Open Folder
View Image
Positive prompt: [ ]

Alternatively as some buttons or link to the right of the "Selected file path":
Selected file: [FILE PATH.png] [Open Folder] [View Image]

Feature suggestion: Change size of image list and preview image

It would be neat if I could choose how wide I want the image/list of images to be. Maybe I want to see more of the image (good for horizontal images), or maybe I want to see a very long path. Let me choose.

Also, if I would rather have a larger space for my image, and elss space for the text fields below, this would be cool too. Meaning to scale the top part (image list and image preview) flexibly in height.

File List cache/storing should be an option

I think storing the "last images" in the program is a bad idea to have enabled by default.
Ideally I would like to opt out of such image caching.

To me, this program is not an image viewer, or a place where I expect to keep a history of viewed files like that.
It's a tool to give information upon request. And as such its less useful to keep a file history to me. And even a slight integrity risk in my opinion.

I may not want my images/paths exposed and saved like this. Not even in a settings.ini

So please if it's not too much work, could you put this as an option? I don't mind what default value it is, as long as I can control it.

image

Cannot launch since 1.20

Since v1.2.0 I cannot launch the application. It just closes. It's the same with v1.21.
Can't give you much more info to debug as nothing is logged as far as I can see.

Any ideas?

Feature Request: Edit metadata

This program is already better at viewing the metadata than any other program I have tested.

The only "key" feature missing is a way to edit the metadata.

Ideally, also in a batch way for all loaded files. Meaning I would love to be able to batch-insert something into the metadata, or to batch search/replace a string (could be useful if I used a temporary LoRA model file-name, and I wanted to clean it up for release without having to re-generate the images).

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.