ClearML FAQ | Why There Is No

Answered

Why There Is No

Why there is no auto_connect_frameworks argument in Task.create but only in Task.init ?

  				
Posted 
	one year ago

					More
				  		
  Report
		
					GaudySnake67
				
					0
					 × 1

Votes Newest

Answers 3

Hi @<1618056041293942784:profile|GaudySnake67>
Task.create is designed to create an External task not from the current running process.
Task.init is for creating a Task from your current code, and this is why you have all the auto_connect parameters. Does that make sense ?

  				
Posted 
	one year ago

					More
				  		
  Report
		
					AgitatedDove14
				
					0
					 × 1

Sure, but why wouldn't I want the ability to control auto_connect_parameters even in an external task?

  				
Posted 
	one year ago

					More
				  		
  Report
		
					GaudySnake67
				
					0
					 × 1

By default the remote link (i..e the Task you are creating with Task.create will have all the auto logging turned on)
For finer control we kind of assume you have Task.init inside your remote script, and then just pass add_task_init_call=False does that make sense ?
Do you think we should have a way to configure those auto_connect args when creating the Task?

  				
Posted 
	one year ago

					More
				  		
  Report
		
					AgitatedDove14
				
					0
					 × 1

Write your answer

793 Views

3 Answers

one year ago

one year ago

Tags

Similar posts

Hello, Deploy Clearml In Docker And Train A Yolov8 Model, Finish Train, In Plots No Pics Show,So Why

Hi! I Am Currently Testing The Function Of Tags, Using Dataset.Add_Tag(). Why Do Tags Only Show On A Version Level, But Not On The Dataset-Level? (See Images) The Tag “Blaa” Has Been Added Manually In The Web-Ui. Why Do “Kophenhangen”, “Paris” Or “Vis” No

Hi, I'Ve Noticed That In A Few Experiments That Got Stuck With No Indication That Something Went Wrong And It Kept On "Running" Until I Manually Aborted The Experiments, Is There A Way To Create A Timeout For If The Server Is Stuck? For Clarification I'M

Hi All. I'M Trying To Start Docker Container In Agent Not As A Root. Read, That There Is Clearml_Agent_Exec_User Environment Variable For That. But Is Seems To Be Ignored, When I Pass It As With Docker'S -E Argument. What'S The Correct Way To Make Agent N

Hello! Is There A Way To Override The Configuration Vault Parameters Of A Pipeline Step With The Add_Function_Step Method? I See In The Docs That Add_Step Method Has The Option To Override The Vault With The Configuration_Overrides Argument, But Not Add_F

Good Afternoon, Please Tell Me, Are There Plans To Make Clearml A Scalable Solution, So That Information About Several Experiments Is Not Recorded In One Index And So That Deletion Is Real, And Not Just Marked That This Data Is No Longer Available?

Hello!, Does Anyone Know Why When Running A Task In A Specific Docker Image, There Are Some Python Packages Whose Versions Doesn'T Match? For Example, I Have Installed The Package Ultralytics On 8.1.25 On The Docker Image But It'S On 8.1.11 In The Install

Hi! Is There Any Way To Add Git-Like Ignore File For Versioning Clearml Data? I Saw In Docs A Wildcard Argument When Files Are Added To A Dataset. How Can I Specify Ignoring Of Some File Types? For Example, I Want To Ignore Ipynb Checkpoints. How Can I Do

Is There A Way I Can Upload An Ml Model (Raw, No Training Task) To S3 And Log It In The Models Section Of Clearml. I Succeeded With Storage Manager But Dont Know How To Make It Visible In Models Section. Also My Model Is Sometimes A Single File And Someti

Hello, We Use Clearml Agents In Daemon Mode. This Writes The Logs To /Tmp. However When The Agent Dies, It Deletes The Logs As Well And We Dont Know Why It Died. Is There Any Setting To Override The Log Deletion?