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
Hey Everyone, Not Sure Exactly When It Started To Happen But I Get Unexpected Behavior With

Hey everyone,
not sure exactly when it started to happen but I get unexpected behavior with task.init
jobs with same project/tasks name are not overwritten - new task is created with identical name, I use the default params, i.e
task = Task.init(project_name="my-root-project/lions_test", task_name="task_num1")3 jobs submitted and in my server I see 3 jobs with the same name

one thing worth to mention that I'm using clearml-server 1.4, bit outdated and will upgrade in the next couple weeks

  
  
Posted 2 years ago
Votes Newest

Answers 18


okie so this works only if jobs run in parallel
first job create new task id
second job (initiated immediately after first job) do the reuse properly

if I wait for first job to finish - then run again new second job with same name, it will not do reuse

is this expected?

  
  
Posted 2 years ago

Well, the data regarding reuse is usually stored in the home folder under .clearml/.session.json I think

  
  
Posted 2 years ago

didn't do that test
I usually wait for first job to finish before I start new one

  
  
Posted 2 years ago

But it will also consider the file name from which this was run

  
  
Posted 2 years ago

I have another instance with clearml-server 1.7 and I got same behavior
as I missing anything? I was under the assumption that jobs with same project/task names should be overwritten and not duplicated

  
  
Posted 2 years ago

and are there times when you try to run it in parallel?

  
  
Posted 2 years ago

from clearml import Task task = Task.init(project_name="Inbar2022/LanguageFactoryDanish/lions_test", task_name="lions3")
python main.py --cuda --epoch 1

  
  
Posted 2 years ago

nope... 😕

  
  
Posted 2 years ago

the behaviour PleasantOwl46 the overwrite behaviour is only if you run it again on the same workstation - the SDK stored a cache and will overwrite if all details are identical and only if less then 72 hours have passed (see sdk.development.task_reuse_time_window_in_hours configuration setting)

  
  
Posted 2 years ago

Another question - is there any specific reason for wanting to reuse the task?

  
  
Posted 2 years ago

same basic job not gets overwritten, but created new one every time

  
  
Posted 2 years ago

SuccessfulKoala55 any clue?

  
  
Posted 2 years ago

compare between both of the tasks

  
  
Posted 2 years ago

SuccessfulKoala55 where does the SDK stores the cache?
we have an cluster with shared storage so all computes nodes that is running the jobs has same storage
should I assume it will use the cache and overwrites identical jobs?
trying to reproduce this but still every new and same jobs gets new task ID

  
  
Posted 2 years ago

got it, I don't really understand why it happens, quite certain I didn't see this in the past

  
  
Posted 2 years ago

How did you run it exactly?

  
  
Posted 2 years ago

not really - I can try to run these in parallel

  
  
Posted 2 years ago

tried it from single workstation, but I get same unexpected behavior
same project/task names, same workstation running the job, anything else I should check to confirm those are identical jobs?
every new identical job start with ClearML Task: created new task id=...

  
  
Posted 2 years ago
1K Views
18 Answers
2 years ago
one year ago
Tags