But, if you like, you can connect a remote interpreter and debug with PyCharm, locally, without clearml-agent

Can you check the running user has all the permissions?

👍

I can not see google package, can you try clone and add it manually? You can always add any package you like to any task with Task.add_requirements('package name', 'package version') ,
Do you have a toy example so I can reproduce it my side (using google.cloud but package is not listed in task)?

ok, I think I missed something on the way then.

you need to have some diffs, because

Applying uncommitted changes Executing: ('git', 'apply', '--unidiff-zero'): b"<stdin>:11: trailing whitespace.\n task = Task.init(project_name='MNIST', \n<stdin>:12: trailing whitespace.\n task_name='Pytorch Standard', \nwarning: 2 lines add whitespace errors.\n"
can you re-run this task from your local machine again? you shouldn’t have anything under UNCOMMITTED CHANGES this time (as we ...

Hi LethalCentipede31

You can report plotly with task.get_logger().report_plotly , like in https://github.com/allegroai/clearml/blob/master/examples/reporting/plotly_reporting.py
For seaborn, once you use plt.show it will be in the UI (example https://github.com/allegroai/clearml/blob/master/examples/frameworks/matplotlib/matplotlib_example.py#L48 )

Hi SmarmySeaurchin8 ,

The trains-agent default uses the ~trains.conf file for credentials, can you verify the api section in this file?

Hi MysteriousBee56 .

What trains-agent version are you running? Do you run it docker mode (e.g.
trains-agent daemon --queue <your queue name> --docker?

Can you check the api version?
from trains.backend_api import Session print(Session.api_version)

Hi RoughTiger69 , when you click on the app “3 dots” link, you can open the configuration and the View details button will open you the original task.

it has like 10 fields of json configurations

under configuration objects, you can find the pipeline configuration.

CostlyOstrich36 did you succeed to reproduce such issue? RoughTiger69 have you made any changes in your workspace (share with someone? remove sharing?)?

Hi UnevenDolphin73 ,

If the ec2 instance is up and running but no clearml-agent is running, something in the user data script failed.

Can you share the logs from the instance (you can send in DM if you like)?

With this scenario, your data should be updated when running the pipeline

thanks SmugTurtle78 , checking it

Hi GhastlySquirrel83 ,

You can specify the repository ( repo=None, repo_branch=None, repo_commit=None parameters) in the add_function_step for connecting the specific repo into the step, you can view all the options with some examples here - https://clear.ml/docs/latest/docs/references/sdk/automation_controller_pipelinecontroller#add_function_step

in the agent’s clearml.conf file, set agent.docker_force_pull to true .
You can also try in the machine running the ClearML agent to run:
docker pull nvidia/cuda:10.1-runtime-ubuntu18.04

can you try
print(task.data.hyperparams)instead of the last line?

UnevenDolphin73 I cant reproduce this issue on my side 😞 can you give me some hints how to?

I can help you with that 🙂

task.set_base_docker("dockerrepo/mydocker:custom --env GIT_SSL_NO_VERIFY=true")

Not sure getting that, if you are loading the last dataset task in your experiment task code, it should take the most updated one.

Hi ProudChicken98 ,

You want the agent NOT to create a new env?

according to the logs, the issue is when installing inplace-abn packages. let me check the error

Hi MinuteWalrus85 .

Good news about fastai , the integration in almost done and a version will be release in the coming days :)

resources configuration, so you have subnet ID and the security group ID and it failed with it?

In the self-hosted we do not have user permissions, so every user sees all the data.

yes, you could also use the container’s SETUP SHELL SCRIPT and run command to install your python version (e.g.
sudo apt install python3.8for example)

So running the docker with ‘“device=0,1”’ works? We will check that

you can this description as the preview, can this help?

task.upload_artifact(name='artifact name', artifact_object=artifact_object, preview="object description")

The preview will show text 😕

Is this image for debugging? Can debug samples section help with that? Or it used for data?

When you are not using the StorageManager you don’t get the OSError: [Errno 9] Bad file descriptor errors?

Unfortunately, it is not possible to delete an experiment using the UI. You can run the script as a service like in the example or run it with job scheduler (crontab for example in linux) to execute it.

Can this do the trick?

not yet, will update here once this is fixed