Examples: query, "exact match", wildcard*, wild?ard, wild*rd
Fuzzy search: cake~ (finds cakes, bake)
Term boost: "red velvet"^4, chocolate^2
Field grouping: tags:(+work -"fun-stuff")
Escaping: Escape characters +-&|!(){}[]^"~*?:\ with \, e.g. \+
Range search: properties.timestamp:[1587729413488 TO *] (inclusive), properties.title:{A TO Z}(excluding A and Z)
Combinations: chocolate AND vanilla, chocolate OR vanilla, (chocolate OR vanilla) NOT "vanilla pudding"
Field search: properties.title:"The Title" AND text
Answered
Hi, Thank You For This Great Tool. We Have An Issue When Resuming A Task When A Machine Is Being Preempted. In Some Cases (Not Always), The Scalars Are Not Logged Properly Anymore When Resuming The Task. The State Still Is Changed To Resume, But The Scala

Hi, thank you for this great tool.
We have an issue when resuming a task when a machine is being preempted.
In some cases (not always), the scalars are not logged properly anymore when resuming the task. The state still is changed to resume, but the scalars stop showing.

Code looks similar to this
task = Task.get_task(project_name, task_name) task.reload() task.started() # always works task.report_scalar() # does not always work task.completed()Any ideas? 🙂 thank you

  
  
Posted 2 years ago
Votes Newest

Answers 7


Hi DefiantShark80 ,

task.report_scalar() # does not always workwhat do you mean? report_scalar not sending the info or raising an error?

  
  
Posted 2 years ago

report_scalar not sending the info

  
  
Posted 2 years ago

Works more robustly now, thanks!

  
  
Posted 2 years ago

the reloading is not necessary?

  
  
Posted 2 years ago

not in that case

  
  
Posted 2 years ago

Can you try something like

` from clearml import Task

task = Task.get_task(task_id="TASK ID")
task.started()
task.get_logger().report_scalar("Some title", series="series", value=2, iteration=0)
task.flush()
task.completed() `This works for me, I can get the report

BTW whats the status of the task before running this code?

  
  
Posted 2 years ago

I will try that, thanks 🙂
it's usually aborted, in that case we already do task.started(force=True) if that's what you meant 🙂

  
  
Posted 2 years ago
560 Views
7 Answers
2 years ago
one year ago
Tags