PungentLouse55 could you test again with the latest from the GitHub? I think the issue should be solved:
pip install git+

Can you see all the agent in the UI (that basically means they are configured correctly and can connect to the server)

JitteryCoyote63 Should be quite safe, there is no major change that I'm aware of on the ClearML side that can effect it.
That said, wait for after the weekend, we are releasing a new ClearML package, I remember there was something with the model logging, it might not directly have something to do with ignite, but worth testing on the latest version.

Hi RipeGoose2
You can also report_table them? what do you think?
https://github.com/allegroai/clearml/blob/master/examples/reporting/pandas_reporting.py
https://github.com/allegroai/clearml/blob/9ff52a8699266fec1cca486b239efa5ff1f681bc/clearml/logger.py#L277

Hmm check if this one works:
optimizer._get_child_tasks_ids( parent_task_id=optimizer._job_parent_id or optimizer._base_task_id, order_by=optimizer._objective_metric._get_last_metrics_encode_field(), additional_filters={'page_size': int(top_k), 'page': 0})If it does, let's PR it as a dedicated function

Thanks again for crazyfrogspb & Tim for helping to find the extra edge case

command line to the arg parser should be passed via the "Args" section in the Configuration tab.
What is the working directory on the experiment ?

Hi @<1668427971179843584:profile|GrumpySeahorse51>
Could you provide the full stack log?
this erros seems to originate from psutil (which is used) but it lacks the clearml-session context

Out of curiosity, if Task flush worked, when did you get the error, at the end of the process ?

👍

using the cleanup service

Wait FlutteringWorm14 , the cleanup service , or task.delete call ? (these are not the same)

Hi @<1578555761724755968:profile|GrievingKoala83>

Is it possible to overrided the parameters through the configuration file when restarting the pipeline from ui?

The parameters of the Pipeline are overridden from the UI, not the pipeline components,
you can to use the pipeline parameters as is as the pipeline components parameters

Is your pipeline built from Tasks, or decorators over functions ?

Hi @<1562610699555835904:profile|VirtuousHedgehong97>
I think you need to upgrade your self-hosted clearml-server, could that be the case?

@<1524922424720625664:profile|TartLeopard58> @<1545216070686609408:profile|EnthusiasticCow4>
Notice that when you are spinning multiple agents on the same GPU, the Tasks should request the "correct" fractional GPU container, i.e. if they pick a "regular" no mem limit.
So something like

CLEARML_WORKER_NAME=host-gpu0a clearml-agent daemon --gpus 0 clearml/fractional-gpu:u22-cu12.3-2gb
CLEARML_WORKER_NAME=host-gpu0b clearml-agent daemon --gpus 0 clearml/fractional-gpu:u22-cu12.3-2gb
```...

Woot woot! 🎊

LazyLeopard18 nice. maybe we should add it in the FAQ / Install. Could you send the exact docker-compose you used and command line, I'll ask the guys to add it 🙂

Hi JealousParrot68

do tasks that are created through create_function_task run the entry_script again instead of just the pure function

Basically they will run the code until the "create_function_task" call, but never after. We are working on adding a decorator to a function, making it a "standalone" script, is this what you actually need ?

and about a month later for some reason the initial iteration seems to have changed to 0

Hmm, I see your point. Just so I fully understand, your are not saying Old experiments were changed, but new experiments (running the same code-ish) have a totally different max iterations value. Is this correct ?

Hi ResponsiveCamel97
Let me explain how it works, essentially it creates a new venv inside the docker, inheriting all the packages form the main system packages.
This allows it to use the installed packages if the version match, and upgrade/change if you need, all without the need to rebuild a new container. Make sense ?

LazyLeopard18 could you explain some more on the specific use case you have in mind?

I just assumed it should only be triggered by dataset related things but after a lot of experimenting i realized its also triggered by tasks...

VexedCat68 I think you are correct, and it should only be triggered by "Dataset" Tasks, that said maybe there is a bug , in which case if there are no additional filters it will get triggered on Any change in the project. This will explain how adding the tags filter solved the issue.
wdyt?

Check here:
https://github.com/allegroai/trains/blob/master/docs/trains.conf#L78

You can configure credentials based on the bucket name. Should work for Azure as well

Wiki edit anyone ?

LOL I keep typing clara without noticing (maybe it's the nvidia thing I keep thinking about)
Carla makes much more sense 😄

Btw I sometimes get a gzip error when I am accessing artefacts via the '.get()' part.

Hmm this is odd, is this a download issue? if this is reproducible maybe we should investigate further...

Hi ShinyRabbit94

system_site_packages: true

This is set automatically when running in "docker mode" no need to worry 🙂
What is exactly the error you are getting ?
Could it be the container itself has the python packages installed in a venv not as "system packages" ?

Yes, that makes sense. Then you would need to use wither the AWS vault features, or the ClearML vault features ...

but could you try with the latest RC?

I love the new docs layout!

Thank you and thank docusaurus, they rock!

Hi UnevenDolphin73

Is there an easy way to add a link to one of the tasks panels? (as an artifact, configuration, info, etc)?

You can add a link as an artifact, that is probably the easiest:
tasl.upload_artifact(name="just link", artifact_object=" ")

EDIT: And follow up regarding the dataset. As discussed somewhere previously, the datasets are now automatically moved to a hidden "sub-project" prefixed with

.datasets

. This creates several annoyances that I...