Hello,

Is there some way to use clearml with "spawn" multiprocessing?

By "spawn" multiprocessing I refer to:
import multiprocessing as mp mp.set_start_method('spawn')
Any attempt I made to use "spawn" multiprocessing with clearml failed.

One failure is easily reproducible, just run:
import multiprocessing as mp mp.set_start_method('spawn') from clearml import Task task = Task.init("OMD", task_name="bla") task.set_base_docker("/home/rdekel/anaconda3/envs/P1") task.close()
And you will see that clearml fails, with the following console output:
ClearML Task: overwriting (reusing) task id=1f49354479cc490881f98ae8cbde278c ClearML results page: 2021-09-20 12:31:37,961 - clearml.Task - INFO - Waiting for repository detection and full package requirement analysis ClearML Monitor: Could not detect iteration reporting, falling back to iterations as seconds-from-start 2021-09-20 12:36:37,971 - clearml.Task - INFO - Repository and package analysis timed out (300.0 sec), giving up
With other code clearml halts indefinitely instead of timing out after 300 seconds, but I suspect this very simple example illustrates the more general problem.

I am using clearml version 1.0.4.

I would appreciate any help.

Thanks a lot,
(And Hag Sameach,)
Ron