ClearML/clearml-docs
1
0
Fork 0
mirror of https://github.com/clearml/clearml-docs synced 2025-03-20 03:58:46 +00:00
Code Issues Actions 1 Packages Projects Releases Wiki Activity

Add Building Bots usecase

Browse Source
This commit is contained in:
revital 2025-03-03 11:41:18 +02:00
parent 41d483c51d
commit f1a6bea551
1 changed files with 21 additions and 52 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

73
docs/getting_started/building_bots.md
View File

@ -2,64 +2,33 @@
title: Building Bots title: Building Bots
--- ---
You can use ClearML to build bots monitor tasks, sending notifications and alerts based on specific events or conditions. ClearML can be used to build bots that monitor tasks and send notifications based on specific events or conditions.
The base `Monitor` class provides essential functionalities that enable developers to implement task monitoring logic The base [`Monitor`](https://github.com/clearml/clearml/blob/master/clearml/automation/monitor.py) class provides
tailored to their requirements. The Monitor class can be extended to create custom bot logic. the functionalities for implementing custom task monitoring logic.
The `Monitor` class can be extended to introduce custom bot logic. Developers can inherit from this class and override
methods to introduce custom bot logic. For example, by overriding the `process_task` method, you
can:
* Send notifications via Slack or other channels
* Log task statuses to a database
* Trigger automated responses when a task reaches a specific status
This is useful for: This is useful for:
* Real-Time Task Monitoring: Automate the tracking of task statuses (e.g., completed, failed) to streamline workflows * Real-Time Task Monitoring: Automate the tracking of task statuses
and reduce manual log-checking. * Custom Notification Logic: Implement specific filtering criteria based on task names, task projects, or other
* Custom Notification Logic: Implement specific filtering criteria based on task names, project identifiers, or other criteria to receive relevant alerts.
criteria to receive alerts that are relevant to your needs.
* Resource Efficiency: Utilize a customizable polling mechanism to minimize resource usage while continuously monitoring
tasks.
The Monitor class can be extended to introduce custom bot logic. For example, by overriding the `process_task` method, you ## SlackBot for ClearML Task Monitoring
can define specific actions to be taken when a task reaches a certain status—whether it’s sending notifications through
various channels (like Slack, email, etc.), logging to a database, or triggering automated responses.
ClearML provides an example Slack bot built using [`clearml.automation.Monitor`](https://github.com/clearml/clearml/blob/master/clearml/automation/monitor.py). The bot does the following:
Using a Slack Bot for ClearML Task Monitoring * Monitors task status (completion, failure, etc.).
* Sends notifications to a specified Slack channel.
* Retrieves task details such as status, console logs, and links to the ClearML Web UI.
Overview You can configure filters for task types and projects to reduce unnecessary notifications.
Integrating a Slack bot with ClearML allows teams to stay informed about task statuses in real time. This is especially
useful for monitoring machine learning experiments and automation workflows. By leveraging the ClearML automation
monitor, you can receive Slack notifications for completed or failed tasks, helping teams react quickly and maintain
smooth operations.
$$$$$$$By using a ClearML Slack bot, teams can streamline their workflow, stay informed, and quickly address critical task updates without manually checking logs or dashboards.
Use Cases and Benefits
2. Real-Time Monitoring - Get instant Slack notifications for completed or failed ClearML tasks. Reduce the need for manual log-checking by receiving automated updates.
3. Customizable Alerts - Configure alerts to report only failed tasks, completed tasks, or both.
Set filters for specific projects, users, or iteration thresholds to avoid unnecessary notifications.
## SlackBot example
This integration is powered by `clearml.automation.monitor`, which allows for efficient task tracking. The `SlackMonitor`
class extends the Monitor class from ClearML and enables customized Slack notifications based on task events. The bot
retrieves task details such as:
* Task status (completed or failed)
* Console output logs
* ClearML Web UI links
The ClearML monitor is a highly flexible object that allows deep visibility into the ClearML ecosystem. It can provide
insights into task completions, failures, and various other ClearML-generated events. One powerful use case is leveraging
this monitor to create a Slack bot that keeps users informed about task progress. The Slack bot is built using a set of
Slack APIs around the monitor, providing real-time notifications about task statuses.
The Slack bot is fully equipped with numerous configurable arguments, making it versatile for different workflows. While
the provided example can be used as a command-line tool, users can also customize it to fit their specific needs. The
bot can be added to a Slack workspace and configured to report only specific types of tasks, such as failed tasks,
completed tasks, or those from designated projects. This eliminates the need for manual monitoring and allows users to
stay informed effortlessly.
For a full example and detailed setup, refer to the Slack Alerts Example.
For a full example and detailed setup, see the [Slack Alerts Example](../guides/services/slack_alerts.md).
![Slack Alerts](../img/examples_slack_alerts.png)