HPO does not complete after timeout is reached

[d13g0]

Describe the bug

HPO does not complete after timeout is reached

To reproduce

Here is my code: https://gist.github.com/d13g0/76f3d092313d261308528297da21bbba

I am checking on timeouts and the number of active tasks in the while loop.

Also, once the optimization is completed (zero active tasks left...) shouldn't that trigger the completion of the HPO task itself?

Thanks.

Expected behaviour

The HPO task completes when the timeout is reached

Environment

• Server type self hosted
• ClearML SDK Version 1.12.1
• ClearML Server Version (Only for self hosted): WebApp: 1.12.1-397 • Server: 1.12.1-397 • API: 2.26
• Python Version: 3.10
• OS (Windows \ Linux \ Macos): the worker running the clones is Ubuntu Linux and the computer running the HPO task is my Mac Ventura M1.

[jkhenning]

Hi @d13g0, in order to use the time limit, you need to specify it using set_time_limit(), otherwise the feature will be inactive

[jkhenning]

I'm sorry, I missed the use of optimization_time_limit 🙂