3.1 KiB
| title |
|---|
| Streamlit Launcher |
:::info Enterprise Feature The Streamlit Launcher App is available under the ClearML Enterprise plan. :::
Streamlit is a framework for creating visual web interfaces for your models. The ClearML Streamlit launcher application spins up a Streamlit execution environment and serves your Streamlit app on a machine of your choice. Once you start a Streamlit launcher instance, you will be provided with an externally accessible link to your Streamlit app, where you can make use of your models.
The Streamlit launcher monitors the Streamlit app activity and shuts down if it is inactive for a specified maximum idle time.
:::important Task Traffic Router The Streamlit Launcher relies on the ClearML Traffic Router which facilitates user authentication, and redirects requests to the IP/port served by the Streamlit app. :::
Once you start a Streamlit launcher instance, you can view the following information in its dashboard:
- Streamlit App status indicator
- App is running and is actively in use
- App is setting up
- App is idle
- App is stopped
- Idle time
- Streamlit App - Externally accessible link to your Streamlit app. You can send this link to your colleagues, so they can
access the app. Click
to copy link - Streamlit Git repo - Repository that holds the Streamlit app script
- Live preview of the Streamlit app
- Console Log - The console log shows the launcher instance's activity, including server setup progress, server status changes
Streamlit Launcher Instance Configuration
- Streamlit Instance Name - Name for the Streamlit launcher instance. This will appear in the instance list
- Project name - Project where your Streamlit launcher app instance task will be stored
- Task name - Name of task for your Streamlit launcher app instance
- Git Repository - Git repository containing the Streamlit script
- Git Branch - Git branch containing the Streamlit script
- Streamlit Script Name - Name of Streamlit script to be executed
- Queue - The ClearML Queue to which the Streamlit launcher app instance task will be enqueued (make sure an agent is assigned to that queue)
- Docker Image - Docker image the ClearML Agent will use for running the Streamlit app
- Arguments - Arguments to be passed to the script
- Idle Time Limit (Hours) - Maximum idle time (Period in which no requests are received by the Streamlit app) after which the Streamlit app and the launcher instance will shut down.