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 All! I Am Trying To Use The Clearml-Agent To Execute Tasks Scheduled In The Clearml Webui, However I Have Some Issues And I Was Wondering How The Agent Works Internally To Recreate An Image. In The Console Logs For Example:

Hey all! I am trying to use the clearml-agent to execute tasks scheduled in the ClearML WebUI, however I have some issues and I was wondering how the Agent works internally to recreate an image. In the console logs for example:
ModuleNotFoundError: No module named 'albumentations' and of course the experiment fails. I'm using explicit task execution to run tasks pulled using the ClearML API client, so I was wondering what's the best way to resolve this issue?

  • Do projects have to be containerized manually before using the clearml agent? Can the ClearML agent even use custom images?Further, different projects use different versions of python frequently. I am using a script that uses the ClearML API ( from clearml.backend_api.session.client import APIClient), to read tasks in the queue. The clearml package is just installed in a conda environment with python=3.9, nothing else. For some reason, projects are trying to use this interpreter for executing the pulled tasks. For example, a project that uses a 3.8 binary returns interpreter not found kind of error, and it forces execution using 3.9.
  • Is there a way to pull the required version of python for a task? Maybe some API endpoint? This way, execution could be parameterized.Thanks in advance!
  
  
Posted 8 months ago
Votes Newest

Answers 3


Regarding the python version, the version used to run the local task is recorded on the task , and the agent will try to locate as close as possible python version when running the task remotely (from the python versions installed on the remote machine/docker image)

  
  
Posted 8 months ago

Can you share logs or different runs and explain what the code does in each? Are all your scripts running remotely using an agent?

  
  
Posted 8 months ago

Hi @<1600661428556009472:profile|HighCoyote66> , what exactly is the flow you're currently using? In ClearML, experiments are created on the server by running local code (allowing the ClearML SDK to capture their properties and dependencies) or alternatively using the clearml-task CLI to create the task by specifying all properties and dependencies, than these experiments can be cloned and queued for remote execution - at this point, the agent running the task remotely can set up an environment and install all required dependencies.
From your description, I assume some dependencies are not installed, so I wonder how you've created the task and how you are running the agent to run the tasks remotely

  
  
Posted 8 months ago
506 Views
3 Answers
8 months ago
7 months ago
Tags
Similar posts