ClearML FAQ | I Have A Script In Which I Added

Answered

I Have A Script In Which I Added

i have a script in which i added Task.init() hook, so when i run the script, it creates the experiment and i can its logs there etc.. but init() function doesn't allow me to put the task in the server queue.. only Task.enque() function does... i am wondering the queueing functionality is mainly used from the UI or its ok to use from running training script manually as well

  				
Posted 
	3 years ago

					More
				  		
  Report
		
					PompousParrot44
				
					0
					 × 1

Votes Newest

Answers 2

PompousParrot44
Check out the task.execute_remotely()
You can call it right after the task init, and it will enqueue your running Task, and leave the process (if you want).
https://github.com/allegroai/trains/blob/65a4aa7aa90fc867993cf0d5e36c214e6c044270/trains/task.py#L1437

  				
Posted 
	3 years ago

					More
				  		
  Report
		
					AgitatedDove14
				
					0
					 × 1

thanks

  				
Posted 
	3 years ago

					More
				  		
  Report
		
					PompousParrot44
				
					0
					 × 1

Write your answer

559 Views

2 Answers

3 years ago

one year ago

Tags

Similar posts

Hi, Is There A Way To Know When Some Experiment Change Status To Published? How I Can Catch The Change ? Is It Same With Slack Monitoring Example Script? I Want To Do Automatic Deployment When Its Published Tia

Hi Team, My Company Asked Me To Research About Clearml. So I Tried To Use Clearml Its Good.. But Some Isssue I Faced Here. I Want Create Clearml & Clearml Agent In Single Terraform Script. Script I Can Add In Userdata, Issue I Faced I Cant Create Clearml_

Hi. I Use Set_Base_Docker For Remote Execute. Its My Custom Image From Gitlab Registry. And My Question Is How I Can Use Script And Etc From My Image? On My Local Machine I’Ve Got In Docker Container Folder With Code From That Image If I Made It Like From

Hi All! I'M Running A Pipeline From Functions. I Have One Task In The Pipeline Where I Want To Call A Specific Function From Commend Line So I'Ve Put It Into A Bash Script And While Running Pipeline Locally It Worked Fine Because Bash File Was Available O

1St: Is It Possible To Make A Pipeline Component Call Another Pipeline Component (As A Substep)? Or Only The Controller Can Do It? 2Nd: I Am Trying To Call A Function Defined In The Same Script, But Unable To Import It. I Passing The Repo Parameter To The

Hi, How Can I Easily Start A Shell Script From Within An Experiment And Have Its Logs (Stdin/Err) Logged In Clearml?

I Have A Function That Runs Normally. Its Job Is To Monitor A Specific Folder, And When I Execute The Script Locally It Works Fine. When I Make A Taskscheduler And Pass That Function To The Scheduler, Then Run Remotely, Then I Get An Error: Clearml Resul

Hi, How Can I Query On Plots And Log Added For A Specific Task Programmatically? (As In, Besides The Clearml Ui And Asumming I Have Its Task Id)

Hi, I Have A Question Regarding Trains And Git. If I Make Changes To A File And Run It Using Trains (Without Pushing The Changes To Github), What Version Of The Script Is Running? It Seems That Every Change I Make Need To Be Added, Committed And Pushed Be

Hello, Does Anybody Know What Triggers A New Model To Be Added In A Project (Working In Pytorch) ? I'M New To Trains And Adding It To My Script Generated A Huge Amount Of Models (Almost 1 Per Datapoint I Would Say) And It Would Also Prompt