Overview of Plugins in Ango Hub
Last updated
Last updated
Plugins in Ango Hub empower you to enhance and customize the platform’s functionality to suit your specific needs. With plugins, you can seamlessly add new export formats, upload assets, integrate machine learning models, and much more.
Ango Hub offers five distinct plugin types, each designed to meet different requirements:
Export Plugins: Customize and extend data export options, enabling support for new formats or adapting existing ones.
Model Plugins: Integrate your own machine learning models to automate tasks such as pre-labeling, interactive annotation, and quality assurance.
Batch Model Plugins: Handle large-scale model processing efficiently by performing batch predictions across multiple assets.
File Explorer Plugins: Extend the file management capabilities of Ango Hub, enabling custom file browsing and interaction features.
Markdown Generator Plugins: Automatically generate documentation or reports in markdown format based on your project data.
You can install plugins from our centralized plugin repository, the Directory, and quickly add them to your project. More on the Installing Plugins here.
Ango Hub provides several ways to run plugins. Follow the instructions below to run plugins from different parts of the platform.
To run a plugin directly from the Plugins page:
Navigate to Your Project > Settings > Plugins.
Select the plugin you want to run and click Open.
Configure the plugin as needed by adjusting the Plugin Settings.
Ensure the plugin’s status is Live to run it properly.
Click the Run button to run the plugin.
To use a model plugin within the labeling editor:
Go to the Labeling Editor.
Ensure the plugin’s preset is prepared and set as the default. For more details, visit the Plugin Setting Presets page.
Click the Plugin icon in the labeling editor’s toolbar.
Click the name of your plugin from the list and run it.
To add a model plugin to your workflow:
Go to the Workflow tab.
Add a Plugin Stage to the workflow.
Connect the plugin stage to other workflow stages as needed.
Select your desired model plugin.
Load an existing preset or configure the plugin manually, including settings such as overwrite options, class mappings, and the config JSON.
Save the workflow.
Once configured, the plugin will automatically trigger when assets enter this stage, running the associated code for each asset. Make sure the plugin’s status is set to Live to ensure it runs correctly.
If the plugin stage stops running for any reason, you can restart it using the Re-run button.
You can develop your own plugins and choose to use them just in your own organization, or to publish them to the Directory for everyone to use. Documentation on creating plugins can be found here.
You can develop plugins of five types, each with its own documentation page:
We also have a page detailing how to deploy your plugin to the cloud to ensure maximum uptime.
