CostlyOstrich36

0 Questions, 4213 Answers

Active since 10 January 2023

Last activity 2 years ago

Reputation

Answers 4213

0 Hi, I’M Running An Agent In K8S Using The Glue. I’M Trying To Set The Setup Shell Script To Install Some Packages Through Apt But I Don’T Think It Even Runs. I Dont See Anything In The Console Logs That Says It Does, At Least. Is There Anyway I Can Debug

Can you add a full log of an experiment?

one year ago

0 Hi Team, I Have Been Trying To Setup Pipeline Using Clearml. I Set Up An Agent In Colab And Also In Ec2. However In Both Cases, The Pipeline Is Pending. I Checked The Logs In The Agent And It Is As Below. Can Someone Help Me Understand What Is Wrong? Fy

Hi @<1552101447716311040:profile|SteadySeahorse58> , if the experiment is still in pending mode it means that it wasn't picked up by any worker. Please note that in a pipeline you have the controller that usually runs on the services queue and then you have the steps where they all can run on different queues - depending on what you set

2 years ago

0 Hi! I’Ve Been Using Clearml For Some Time In My Previous Company, But I Have A New Job Now, And Naturally I Want To Continue Using It And Hopefully Convince My Team Mates! But I Couldn’T Set The Clearml-Server Docker. We Are Discouraged From Using

Hi! Good to see another ClearML user that carries it with them between companies ^^

Also, did you make sure to give the required permissions to the clearml folders in /opt/clearml/ ?

3 years ago

0 Another Question: How Can I Make Clearml-Agent Use Pre-Installed Version From The Nvidia/Pytorch (

Oh LOL 😛

3 years ago

0 Is There A Way To Use The Scalar Smoothing When Opening The "Maximize Graph"?

BoredPigeon26 , it's a feature in our next release 🙂

4 years ago

0 I'M Trying To Understand The Difference Between Tasks And Pipelines. I Get The Big Picture But The Thing I Struggle To Understand Is The Interplay In The Code Itself. If You Set Up A Pipeline You Wouldn'T Include A Task Init? Or Is There A Reason To Use B

Hi @<1545216070686609408:profile|EnthusiasticCow4> , generally speaking, pipelines are a special type of task. When you write steps using decorators you don't have to add the task init. However you can also build pipelines using existing tasks in the system, where those were created with task.init

2 years ago

0 Hello Guys, Can You Tell Me Where The Console Outputs Are Stored? For Some Reason, All Outputs Have Disappeared From All My Pipelines. Any Explanation, Does Anyone Have An Idea What Might Have Happened?

Hi @<1702492411105644544:profile|YummyGrasshopper29> , console logs are saved in Elastic. I would check on the status of your container

one year ago

0 Hi, It Seems That Autorefresh Stopped Working For Me. When Checked It Used To Automatically Update Charts In The Scalars Tab And Now It Is Not Happening (Even Though The Box Is Checked) And I Need To Manually Refresh The Page For The Scalars To Update. As

Hi @<1566596960691949568:profile|UpsetWalrus59> , it seems you are correct. Can you please open a github issue to follow up on this? I'm sure it should be fixed fairly quickly 🙂

2 years ago

0 Hi, I Am Trying To Use The Clearml-Agent In Docker Mode To Run An Experiment, But It Seems To Fail Passing The Clearml.Conf File To The Docker Container:

And you made sure to run clearml-agent init on the machine or to implement the clearml.conf manually?

2 years ago

0 Hi

Hi BattyLizard6 , can you please open a GitHub issue to follow the issue

3 years ago

0 [Errors When Migrating Clearml Server From Aws To Gcp]

Hi @<1523702496097210368:profile|ScantChimpanzee51> , your steps look ok but the error pretty much indicates that there is a folder permissions issue. Please navigate manually to /opt/clearml/data folder and check "ls -al" command what are the user and permissions for the "elastic_7" folder and then enter the elastic_7 folder and check the same for its "nodes" subfolder. If the permissions are correct try restarting the docker and checking if it helps.

2 years ago

0 How To Open The Dashboard (Local Server) I.E. While Running It Give Me A Local Host Port, But How To Lunch It Without Running? Tanks,

How do you run docker compose? If you run it with the -d in the end it should stay and be persistent even after restart, if I'm not mistaken

3 years ago

0 Hi Folks, Is It Possible To Disable Automatic Package Detection And Just Use Requirements.Txt? I Have Different Local And Remote (Agent) Environments So I'D Like To Keep Those Two Separate. I'Ve Tried Using Task.Set_Packages("Requirements_Remote.Txt") Aft

You can also just delete the installed packages section from the webUI and it will force it to use the requirements.txt

one year ago

0 Hi, Can I Ask How I Can Make Clearml-Datasets In Comparison With Pytorch Datasets/Dataloader? In Particular, Pytorch Dataloaders Would Be Able To Batch Pull And Then Preprocess Data Using Multi-Cpus, Feed It Into The Training Loop And Achieve As High Util

SubstantialElk6 , I think this is what you're looking for:
https://clear.ml/docs/latest/docs/references/sdk/dataset#get_local_copy
Dataset.get_local_copy(..., part=X)

3 years ago

0 Hi Clearmlers, I'M Trying To Create A Dataset With Tagged Batches Of Data. I Firstly Create An Empty Dataset With Dataset_Name = 'Name_Dataset', And Then Create A Another Tagged Dataset With The First Batch And With Parent_Datasets=['Name_Dataset']. It'S

Can you add a code snippet that reproduces this for you please?

one year ago

0 Hi There All! Is There A Way To Set Empty Requirements When Creating A Remote Task? I Have Tried To Set Packages To None Or Empty Array, But It Doesn’T Work, Which Is Corroborated By The Documentation Of That Function

Hi @<1659005876989595648:profile|ExcitedMouse44> , you can simply configure the agent not to install anything and just use the existing environment 🙂

The relevant env variables for this are: CLEARML_AGENT_SKIP_PIP_VENV_INSTALL
CLEARML_AGENT_SKIP_PYTHON_ENV_INSTALL
None

one year ago

0 Hello There! After Updating Clearml Server To The Latest Version I'M Not Able To Download Old Datasets. I Got An Error

Maybe SmugDolphin23 or AppetizingMouse58 might have some insight into this since behavior of Datasets changed in the last year I think

2 years ago

0 I Am Trying To Implement A Service Task Which Will Do Some Orchestration Of Jobs, Queues, Vm Instances, Etc. I'M Using Clearml.Backend_Api.Session.Client.Apiclient To Fetch The List Of Queues And Tasks With Their Statuses. What I Also Need And Cannot Figu

Hi @<1631102016807768064:profile|ZanySealion18> , I would suggest using the web UI as a reference. Open developer tools and check what is being sent/received when looking at the workers/queues pages

one year ago

0 Hi Everyone! I Use Clearml Pipelines And I Have Too Much Parameters In It So I Want To Use Configuration File. How Could I Connect Configuration File (Like `Task.Connect_Configuration_File`) But In Pipeline With Ui Interface?

Hi @<1569496075083976704:profile|SweetShells3> , and how do you expect to control the contents of the file? Via the UI or to upload it and then run the pipeline?

2 years ago

0 Hi, I’M Getting This Error When I Try To Run Task On A Remote Agent With Docker Mode Web Ui:

Logs shows me that key is mounted to the docker container

How are you mounting the credentials?
What version of ClearML-Agent are you using?

3 years ago

0 Hi There, It Seems Like There Is A Bug With The Visualization Of Debug Samples On The Ui (Server V1.2.0, Self-Hosted): When Clicking On A Debug Sample Then On The Download Button, If The Sample Is Stored In S3, The Download Button Opens A Blank Page With

JitteryCoyote63 , let me take a look if it happens to me as well 🙂

3 years ago

0 Hi, I Have Started Learning Clearml And How To Use It. I Deployed Clearml And Clearml-Agent In A Kubernetes Cluster Using Helm Chart. Can Anyone Please Guide Me On How To Run An Ml Experiment Using Clearml-Agent Deployed In The Cluster? I Have Set The Arg

None

2 years ago

0 Hey Team I Can See Last Clearml-Server Released On August, When New Release Going Public? I'M Going To Upgrade Our Env And Prefer To Update After Upcoming New Release

Hi @<1523701842515595264:profile|PleasantOwl46> , the version is released, thus public. Not sure what you mean, can you please elaborate?

one year ago

0 Hello, Is It Possible To Use Clearml-Serving To Serve A Model Built With Scikit-Learn'S Gridsearchcv Pipeline? I'M Asking This Question Because When I Test It, I Get An Error.

Hi @<1673501397007470592:profile|RelievedDuck3> , should be possible. What errors are u getting?

one year ago

0 Hey There, Is There Any Way I Can Tell The Task Not To Set A Random Seed? I'M Setting Up Reproducibility Myself But When I Call Task.Init() The Seed Is Changed. Is It Possible To Tell Clearml Not To Initialize Any Rng? It Appears That Task.Set_Random_Seed

Hi TartBear70 ,

You can use the following method:
https://clear.ml/docs/latest/docs/references/sdk/task/#taskset_random_seed
Please note you need to set it before running Task.init()
If you set it to None this will cancel any random seed override performed by ClearML.

Tell me if this helps 🙂

3 years ago

0 Hi, I'Ve Run Into A Problem And Would Appreciate Some Help. I Installed Clearml Locally. When I Run A New Task On A Remote Server And In The Python Training Code I Set It To Only Train On One Gpu. Everything Works Fine And I See All The Scalars Automatica

Also, what GPUs are you running on that machine?

11 months ago

0 Hello Everyone. I Was Following The Docs About Creating Pipelines Using Decorators (As It Looked Really Neat). Everything Worked Well For My Example, But I'Ve Noticed That Much Time For Building The Container Was Taken By Installing Packages That Weren'

VividDucks43 , I think I might have misunderstood you a bit - For a single pipeline you would have the same requirements.txt, why would you need many?

3 years ago

0 I Have The User/Pass Set Up On The

wait.

11 months ago

0 Hi Everyone, I'M Having An Issue With Clearml Datasets And Would Like To Know If This Is Possible. I Have A Task That Is Executed Repeatedly. This Task May Require Data To Be Loaded And Updated From A Dataset. My Question Is: Is There A Way To Append To A

In that case you are correct. If you want to have a 'central' source of data then Datasets would be the suggested approach. Regarding your question on adding data, you would always have to create a new child version and append new data to the child.

Also maybe squashing the dataset might be relevant to you - None

4 months ago

0 Hi All. In Self Hosted Clearml Suddenly Scalars Stopped To Be Shown In Scalars Tab. Any Ideas?

Having the latest versions is always a good practice

2 years ago

Show more results