AgitatedDove14

49 Questions, 8124 Answers

Active since 10 January 2023

Last activity one year ago

Reputation

Badges 1

25 × Eureka!

Answers 8124

0 Hi I Saw This On The Clearml-Agent Docs But Other Than The Docker Image, I'M Not Sure How To Integrate This With Clearml Py And Clearml-Server. Please Advise.

SubstantialElk6 whats the command line you are using ?

4 years ago

0 Crazy Idea:

s like the

would be a really good starting place.

This is actually JS (typescript) ... not python, not sure on how to continue from there 😞

one year ago

0 Hi There!

Would it also be possible to query based on

multiple

user properties

multiple key/value I think are currently not that easy to query,
but multiple tags are quite easy to do

tags=["__$all", "tag1", "tag2],

2 years ago

0 It Seems Like Clearml Agent Does Not Support Arparse Subparsers, Right?

@<1527459125401751552:profile|CloudyArcticwolf80> what are you seeing in the Args section ?
what exactly is not working ?

2 years ago

0 Any Guy Know Who Is The Admin Of This Workspace?

I believe AnxiousSeal95 is.
ElatedFish50 any specific reason for the question?

3 years ago

0 I'M Experiencing Some Weird Behavior From The Automatic Logging Iterations. It Seem To Be Capped At The Number Of Batches Rather Than The Epochs. How Can I Control Which Variable The Logging Mechanism Tracks?

Interesting... TrickyRaccoon92 could it be the validation phase was creating a new Tensorboard file ?

4 years ago

0 Hi, I Am Using Clearml With Pytorch-Ignite And Its Earlystopping Handler. I Would Like To Log The Counter Of The Patience Of This Handler, How Can I Do That?

Hi JitteryCoyote63 , is there a callback for that?

4 years ago

0 Hello, We Have A Self Hosted Clearml Server Connected To Different Queues And Use It To Launch Remote Experiments (Clearml==1.9.3, Clearml-Agent==1.5.2Rc0). It Is Working Really Well For Us Unless One Workflow :) We Would Like To Abort An Experiment And E

Woot woot, will do!

2 years ago

0 Hi, I Have A Question Regarding The Aws_Autoscaler: It Usually Takes ~Hours To Get A Gpu Instance Nowadays. I Was Thinking, It Would Be Much More Interesting To Stop The Instances (Clearml-Agents) Instead Of Terminating Them Once They Are Inactive, So Tha

instead of terminating them once they are inactive, so that they could be available immediately when they are needed.

JitteryCoyote63 I think you can increase the IDLE timeout on the autoscaler, and achive the same behavior, no ?

3 years ago

0 Second: Is There A Way To Take Internally Tracked Training Runs And Publish Them Publicly, E.G. For A Research Paper? "Appendix A: Training Runs Can Be Found Here, Feel Free To Explore Them And Look At The Loss Curves"? For Example

Hi SmallDeer34
On the SaaS you can right click on an experimenter and publish it 🙂
This will make the link available for everyone, would that help?

3 years ago

0 Hello, Is There Any Option To Run A Docker Image On A Clearml-Agent Without Clearml Or Python? I Have A Colleague Who Just Wants To Execute His Docker On One Of My Clearml-Agents, But Who Does Not Want To Use Clearml. Right Now I Would Tell Him To Just Us

ReassuredTiger98 if this user passes to the task as docker args the following, it might work:
'-e CLEARML_AGENT_SKIP_PYTHON_ENV_INSTALL=1'

3 years ago

0 Hello All! Quick Question, Do Any Of You Know Of A Clean Way To Access The Clearml Logger Inside Of A

That being said it returns none for me when I reload a task but it's probably something on my side.

MistakenDragonfly51 just making sure, you did call Task.init, correct ?
What dues
from clearml import Task task = Task.current_task()returns ?

Notice that you need to create the Task before actually calling Logger.current_logger() or Task.current_task()

2 years ago

0 Hi Everyone! Is There A Way To Specify The Working Directory In A Pipeline Component? I’M Using Pipelines From Decorators, I Can Set The Repo Url Just Fine, But I’M Running Everything From A Subfolder, And The Working Dir Is Set To

Hmm yes, @<1570220858075516928:profile|SlipperySheep79> I think you are right in your case it make sense to do add this option.
Could you add GH issue with the feature request? it should be fairly easy to add and we use GH to make sure we track those requests
wdyt?

one year ago

0 Hello, Is It Possible For The Clearml-Agent In Docker Mode To Not Pull A Specific Docker Image, But To Build One From The Experiment Repository Using The Dockerfile And .Dockerignore Of The Experiment Repository?

BTW: dockerhub is free and relatively cheap to upgrade 🙂
(GitHub also offers docker registry)

3 years ago

0 Hi Guys, With The New Venv Caching Available In Clearml, I Have The Following Problem: I Force My Pip Requirements To Be:

JitteryCoyote63 I found it 🙂
Are you working in docker mode or venv mode ?

4 years ago

0 Hi Everyone, I'M Getting Really Weird Issue With Clearml Installation. Basically, I Have An Environment Where I Already Configured Clearml Via Clearml-Init, But Inside Code I'M Running On Slurm My Task.Init() Raises "Missingconfigerror". Worth Mentioning,

Nice 👍

one year ago

0 Hello, I'M A Bit Lost In The Docs For The Mlops, I Have Script Which Already Integrate Clearml Logging, Should I Use Clearml-Task To Launch It On An Agent ? (I Already Have A Clearml-Server And A Clearml-Agent Running).

You can change it the CWD folder, if you put . in working dir it will be the root git repo, but you can do any subfolder, obviously you need to change the script path to match the folder, e.g. ./folder/script.py etc.

4 years ago

0 Hi. Inside A Notebook When I Cerate A New Clearml Task And Then Run Sklearn Gridsearchcv , Clearml Uploads A Lot Of Model. Is There A Way To Force Clearml Not To Upload These Models? Related Question Is What Are These Models Anyway? Their Name Only Contai

Hmm apparently it is not passed, but it could be.
Would the object itslef be enough to get the values? wouldn't it make sense to get them from outside somehow? (I'm assuming there is one set of args used at any certain moment?)

2 years ago

0 Hi, I Created Two Workers And I Want To Remove Them Now. How To Remove The Created Workers? Cli Or Api?

Hi ThoughtfulBadger56
Just add --stop to the clearml-agent
(the exact same command as you used to spin it, just add --stop at the end and it will stop it, or just do clearml-agent daemon --stop and it will iteratively close them)

2 years ago

0 I'M Using Tensorboard Summarywriter To Add Scalar Metrics For The Experiment. If Experiment Crashed, And I Want To Continue It From Checkpoint, For Some Reason It Plots Metrics In A Really Weird Way. Even Though I Pass Global_Step=Epoch To The Summarywrit

I think we should open a GitHub Issue and get some more feedback, maybe we should just add support in the backend side ?

3 years ago

0 <image>

No sure what O'm seeing here

4 years ago

0 Hello! How Can I Use "Report_Scatter2D" In Order To Report Timestamp In The X-Axis?

SweetGiraffe8

4 years ago

0 Hello All, I'M Trying To Adapt Clearml With My Workflow. I Installed A Server At My Server, With Workers Attached To It. I'M Trying To Execute A Task From My Local Within One Of My Workers. Trying To Use Docker Mode And A Custom Image. I Also Have A Local

ZanyPig66 this should have worked, any chance you can send the full execution log (in the UI "results -> console" download full log) and attach it here? (you can also DM it so it is not public)

3 years ago

0 Please Tell Me, When Migrating A Local Server, We Have Problems That The Saved Images Are Not Displayed, It Says "Unable To Load Image" And Links To The Old Server If You Click "Copy Image Url" Or "Open Image". The Migration Was Done According To Backup'

Nice!!

3 years ago

0 Hi, I Need Your Help Setting Up An Trains Agent Running In Docker. I Have An Python Script Calling Wget As System Command Which Runs Fine On My Dev Engine. When Cloning The Experiment And Scheduling It Into The Services Queue I Get An Error That The Call

Nice!!!

4 years ago

0 Hi, Is There A Way To Create A Draft Experiment Manually? That Is - Give It A Some File To Run, Or, Better Yet, A Function To Run Which Will Be The Start Of The Experiment? In W&B, For Example It Is Possible To Simply Write (Their

OddAlligator72 just so I'm sure I understand your suggestion:
pickle the entire locals() on current machine.
On remote machine, create a mock entry point python, restore the "locals()" and execute the function ?

BTW:
Making this actually work regardless on a machine is some major magic in motion ... 😉

4 years ago

0 Is There A Quicker Way To Abort All Running Experiments In A Project? I Have Over A Thousand Running Anonymous Data Tasks In A Specific Project And I Want To Abort Them Before Debugging Them.

Is there a quicker way to abort all running experiments in a project? I have over a thousand running anonymous data tasks in a specific project and I want to abort them before debugging them.

We are adding "select" all in the next UI version to do that as quickly as possible 🙂

3 years ago

0 Hey, Just Trying Out Clearml-Serving And Getting The Following Error

RobustRat47 what's the Triton container you are using ?
BTW, the Triton error is:
model_repository_manager.cc:1152] failed to load 'test_model_pytorch' version 1: Internal: unable to create stream: the provided PTX was compiled with an unsupported toolchain.https://github.com/triton-inference-server/server/issues/3877

3 years ago

0 Hi, I Am Running Clearml Agent Using Sdk. When I Run A Remote Job On This Clearml Agent, The Venv Setup Is Totally Based On My Requirements.Txt Instead Of Adding On To What The Image Has Before. Why?

Alternatively I understand I can also run the agent using...

No you should not if you are running the agent inside a container it cannot work in docker mode and spin its own containers
Bottom line use clearml-agent daemon

2 years ago

0 Hey There, I Would Like To Increase The

Sure thing 🙂

4 years ago

Show more results