GoulartNogueira / sd-model-organizer

model organizer extension for stablediffusion-web-ui

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

SD Model Organizer

SD Model Organizer is a Stable Diffusion WebUI extension that allows to store and manage own model collections, add information, own notes, previews about model. Easily download from any external direct link or shared file link from Google Drive. Share own models collection with json file or use remote Firestore database to make it available and synced on several devices.

logo.png



๐Ÿš€ Core features

  • Sort model records by time added, name, model types, groups and search by name.
  • Downloads model with single click
  • Downloads batch of models regarding to the filter applied on home screen.
  • Downloads models to the model type predefined path or manually selected.
  • Downloads models into subdirectories.
  • Downloads model's image preview to make it visible in additional networks.
  • Bind new record with already existing model file locally.
  • Add image external url as model's preview.
  • Add model description and own notes with rich text editor (TinyMCE)
  • Stores model records in local SQLite database.
  • Stores model records in remote Firebase Firestore.
  • Export/import existing model as json files.
  • Import models from civitai.com url or model id.
  • Import models info from .info file.
  • List local model files.
  • Remove local files related to model.
  • Dark/Light theme support.



Readme file will be updated soon regarding to 1.1 updates.

Use --mo-show-dir-settings param to show models path parameters in seettings menu

1. Home Screen

home.png The main screen of the extension. Displays already added records. May be configured to display content as cards or table layout in the webui settings.



1.1 Controls

home_controls.png

  • Reload - Allows to reload data from the storage manually. Avoid using page refresh - it will load only initial data from Gradio.
  • Download All - Redirects to the download screen and adds all records displayed on the home screen to the download queue.
  • Add - Redirects to the Add Record screen.



1.2 Display Options

home_display_options.png Display options is an accordion that hides sorting and filtering options to be applied for records to show on home screen.

  • Sort By - Allows to sort records by Time Added, Time Added Reversed, Name and Name Reveresed.
  • Download first - Downloaded records will be displayed first in the list if checked.
  • Search by name - Allows to search records by name, not case-sensitive.
  • Model types - Filters records by selected model types.
  • Groups - Filters records by selected record groups.
  • Show downloaded - Shows downloaded records if checked.
  • Show not downloaded - Shows not downloaded records if checked.



1.3 Cards

home_cards.png After mouse hovering above record card, card options will be displayed:

  • Details - Redirects to the details screen of the current record.
  • Download - Redirects to the download screen and adds current records to the download queue. Visible only if models isn't downloaded yet.
  • Edit - Redirects to the record editing screen.
  • Remove - Redirects to the record removal screen.



1.4 Import/Export

home_import_export.png Import/Export Accordion is placed in the bottom of the home screen, click to expand it.

  • Import - Drag and drop .json file with records to import it to the current selected storage. Note: existing records will not be merged with new ones. Imported records will be added as new.
  • Export - Click to export records displayed on the home screen (Only displayed will be exported, regarding to the filters applied). Click on download button to download it from the browser, or navigate to the <your_extensions_dir>/sd-model-organizer/export dir.



2. Add/Edit Record screen

2.1 Input Fields

edit_fields.png

  • Cancel - Click to cancel record adding/editing and return to previous screen.
  • Save - Saves or adds new model and returns to previous screen.
  • Name - Model title to display (Required)
  • Model type - Dropdown to select record's model type. Currently supported Checkpoint, VAE, Lora, Hyper Network, Embedding and other Other. If Download path is not defined in the Download options section, model will be downloaded to the default model's path defined in webui. Also default path might be changed in the extension settings. Note: Other type requires to Download path to be set in the Download Options
  • Download URL - Link to the model file to download (Required). Note: if URL will be changed during editing - SHA256, MD5 and file location will be erased. Remove file manually or via remove-files only option
  • Preview image URL - Link to the image for preview. Image will be downloaded if Download preview options is checked in the extension settings.
  • Model page URL - Link to the model page information.
  • Bind with existing - Textbox appears below Add Groups section when record doesn't have location after downloading model and user selects model type, and it's directory contains files not related to any of records added before.



2.2 Groups

edit_groups.png Groups dropdown allows to add groups to the current record from already existing records or add new ones by expanding Add Groups accordion. Type comma-separated group names in the textbox and press Add Group button to add new groups.



2.3 Download options

edit_download_options.png Download options section allows to configure download destination of model's file.

  • Download Path - Path to the download dir, default if empty. (Required for "Other" model type)
  • Download File Name - Downloaded file name. Default if empty.
  • Subdir - Downloads file into subdirectory. (for example: {webui_path}/models/Stable Diffusion/photo/)



2.4 Prompts

edit_prompts.png Prompts sections allows to note positive and negative prompts of the model record.



2.5 Description

edit_description.png Description is textfield that support Rich Text editing with TinyMCE editor. Supports different text-styles, links, images, videos, tables and pasting text from clipboard with styles.



3. Record Details screen

details.png Screen that displays all record's information available:

  • Name - Record name.
  • Model Type - Model type.
  • Size - Size of model's file in case it was downloaded.
  • MD5 - Model's md5 hash in case it was downloaded.
  • SHA256 - Model's sha256 hash in case it was downloaded.
  • Location - File path in case model was downloaded.
  • Model page - Link to the model's page if it was added.
  • Groups - Model's groups.
  • Download URL - Model's download link.
  • Download path - Download destination directory.
  • Download filename - Download destination file name.
  • Subdir - Download destination subdirectory.
  • Positive Prompts - Positive prompts text if it was added.
  • Negative Prompts - Negative prompts text if it was added.
  • Description - Formatted description text.



4. Download screen

Download screen contains cards with records selected for downloading. Each card contains current state of download progress for each separate record.

download_pending.png download_in_progress.png download_completed.png



5. Record Removal screen

removal.png Removal screen allows to remove record itself or with files if record was downloaded and files are available.



6. Extension settings

settings.png SD Model Organizer's settings in the webui Settings Tab -> Model Organizer Menu.

  • Layout Type - Layout type displayed on home screen. Currently two options available Cards or Table.
  • Card width - Card's width on home screen. Default value will be used if 0.
  • Card height - Card's height on home screen. Default value will be used if 0.
  • Storage Type - Storage type to store model records. SQLite is selected by default and stores data in {sd-webui}/extensions/sd-model-organizer/database.sqlite file. Firebase option allows to store records data in the remove Firestore database, follow instruction in the separate section of this page for setup.
  • Download Preview - Enabled downloading models preview with model. Checked by default.
  • Model directory - Model's directory to download checkpoints, uses default path if empty.
  • VAE directory - VAE directory to download VAE files, uses default path if empty.
  • Lora directory - Lora directory to download Lora files, uses default path if empty.
  • Hypernetworks directory - Hypernetworks directory to download Hypernetworks files, uses default path if empty.
  • Embeddings directory - Embeddings directory to download Embeddings files, uses default path if empty.



7. Firestore setup

  • install pip install firebase-admin==4.5.0 (latest firebase-admin (6.1.0) has an incompatible protobuf (4.22.0) version dependency. But tensorboard requires 3.20.0 Firebase version that should work fine is 4.5.0.

firestore_setup.png

  • Go to https://console.firebase.google.com/
  • Create a new project
  • Open a project and navigate to the Build -> Firestore Database in the left-side menu.
  • Enable Firestore Database.
  • Go to Project settings, Service Accounts Tab, Select Firebase Admin SDk, click on Generate new private key.
  • Download json file, move it to the {sd-webui}/extensions/sd-model-organizer directory and rename to service-account-file.json
  • Restart webui.

Note: Keep this file safe and don't use service account keys from projects that contains any sensitive data.



๐Ÿ“ฆ Installation

The SD Model Organizer can be installed from the Stable Diffusion WebUI. Simply navigate to the Extensions tab, click on Install from URL, paste link to this repository, click Install button and reload webui.

Manually:

Clone repository into your extensions folder, install bs4 and firebase-admin==4.5.0 if you are going to use Firebase storage type packages with pip, restart webui if needed.



๐Ÿค Contributing

Contributions, issues, and feature requests are welcome! Feel free to check out the issues page if you want to contribute. Please keep in mind I'm not a Python/JS/Gradio developer. I'm new to these technologies stack and my overall experience with it is about one month only.



๐Ÿ“ License

This project is licensed under the MIT License. Feel free to use, modify, and distribute it as you wish.



Models on screenshots

About

model organizer extension for stablediffusion-web-ui


Languages

Language:JavaScript 80.8%Language:CSS 14.7%Language:Python 4.4%