Is it possible to ask an agent to use a specified existing python environment instead of building one from scratch?

and an explicit question: When using Task.get_tasks how do I filter by specific parameters? For example I have a script that accepts as a command line argume...

Why would the Mongo4 migration scripts (for clearml-server 1.2) try to chown 1000:1000?

Question about the storage manager. Assuming I have an object that updates frequently and always saved at the same path (e.g. my_bucket/my_data.csv ) and I w...

How should I edit the Cleanup Service so I can filter deleted tasks on project / task name? https://github.com/allegroai/clearml/blob/master/examples/service...

How come I use logger.report_matplotlib_figure and get as a result a plotly figure??? O_O

How do I restart trains-agents? How do I stop them?

Question about out-of-Python dependencies. For example NLTK requires apt-get install sqlite-devel how can we manage dependencies like that on the train agents?

Is there a robust way (using the SDK and not the UI) to add tags to Task regardless of where it is executed?

Any documentation (code examples) on how to log custom plotly plots?

Is it possible to view the actual code of a task? As in the script that created the task?

Question about pipelines - so the default for pipeline tasks that are executed remotely is to execute on the services queue. Assuming I want only ONE agent l...

I've been working a bit with trains-agent, having them deployed on different machines listening to queues (docker mode) and it's been working good so far. My...

Trying to enqueue a task through the UI, getting this error - what could it be? (Running on AWS, on the official trains AMI) Error 100 : General data error: ...

What is the correct way of setting environment variables for agents which DON'T RUN in docker mode? Do they simply inherit from the environment they were exe...

What's going on with the documentation site?

Is there a way to run a pipeline ( PipelineController ) without using a clearml-agent ? I can't get it to work with the agent, since the task requirements is...

When using PipelineControler is there a way to execute locally? Or must I use a queue?

How come PipelineController has changed in a very backwards-incompatible fashion but we are still on the same major version? O_O Also the documentation is no...

What is being stored exactly in tmp by the clearml agent? I have a process that cleans the /tmp each day, and I now get messages like clearml_agent: ERROR: C...

Assuming I have a trains server installed on Machine A, a trains agent installed on Machine B and I'm developing on Machine C. Is it possibe to launch a task...

I have a task which I reported a table to - how do I access this table through code? The clearml Task object attribute "artifacts" is empty

I'm looking to utilize the Trains AWS autoscaler functionality, but after going through its docs a few times I still don't get it. Ultimately, my setup is th...

Got something very wrong with my pipeline, the pipeline plot shows it has an experiment in_progress but when going to the experiment itself it is already com...

I have a production inference pipeline which I want to continuously test on my Github to make sure it doesn't break as we move forward. The ideal scenario fo...

Executed from within a PipelineController task, what possible reason does current_task() have to return None ? 2021-11-07T11:32:43.7366887Z clearml.Task.curr...

Question about the file server. Currently, we have a machine with MINIO installed, and all file communication is made using the MINIO SDK client. [MINIO is j...

A suggestion. Sometimes newcomers that join an existing project that uses ClearML forget to configure their ClearML for the organization's server resulting i...

Am I doing something wrong or is should I open an issue about it (bug?)? I'm using the logger.report_plotly method to visualize a custom plot. Using the exac...

Question about the usage of trains agents. In our company we have 3 HPC servers, two of them have multiple GPUs, one is CPU only. I saw in the docs the multi...