Hello, . I'm trying to build Pipeline from tasks like on this video None only with one exception - my files has .ipynb extension, noy .py, and i have error (...

Hi ! Our Evangelist GrumpyPenguin23 will give a quick webinar about: “Chicken or the Egg” in MLOPs: The Case of the DIY Feature Store. During the webinar, Ar...

Currently, to provide SSH access to the docker images for a task, .ssh is mounted into the docker. Wouldn't a solution like this https://gist.github.com/d11w...

Hi all, I have a quick question regarding passing secrets to a docker container. Our set up is the following: we use GCP Autoscaler to spawn "workers" where ...

I am curious if Clear has any sort of multi tenant capabilities. I have just found you guys and plan on spinning up an instance to play around, but wanted to...

Hi guys, what is the correct channel to do a technical question?

I'm looking into a way to hint git repository location as I'm using a launcher to start python script. Currently the dashboard shows some unrelated uncommite...

I have some old training jobs that I logged with tensorboard, is it possible to add them to clearml?

GCP Autoscaler Feature Request: Currently, one has to provide the exact machine image to use in the 'Machine Image' field. However, I'd like to use the lates...

@<1537605940121964544:profile|EnthusiasticShrimp49> @<1523701205467926528:profile|AgitatedDove14>

Hi there, I am hosting ClearML server (deployed into a k3s cluster via helm) and whenever pods in a replica set are restarted, they always enter an error sta...

Hi, ,release notes time! Trains 0.16.4 and Agent 0.16.2 are out! 👏 👏 👏 👏 👏 👏 👏 👏 👏 👏 👏 👏 👏 👏 First we would like to thank all of you that starr...

Hello community, I'm planning to deploy a self-hosted ClearML server on AWS. Our usage pattern is to run it intensively once a week, with a few additional se...

How do people generally handle moving from experimental mode with notebooks and then running pipelines for production training and beyond?

Hello everyone, We’re encountering a persistent issue with our autoscaler setup and could really use some help. Despite having the autoscaler running and the...

Hi, I am fairly new to clearml and would appreciate some help with this problem. My team currently uses the virtual environment mode to run the experiments. ...

I just set up the newest version of clearml server on some new hardware. I am trying to run a service job (on service queue on clearml-services agent) and I ...

Is there a way to set access levels per-user on the trains web app? (I'm basically looking to add a READONLY user role)

I wonder if there is any way to sync data on cloud storage (e.g S3 or MinIO) using clearml-data sync command? Because I don’t see any place in the documentat...

Hi! I have question about data managment part of ClearML. Does ClearML support data versioning like in LakeFS ?) Is it similar ? maybe there is some interest...

Hi, lately while trying to create a new dataset we encounter the following error: my code: from clearml import Dataset, Task dataset_1 = Dataset.create(datas...

I am struggling a bit to understand the use case of a pipeline: Let say you have step1 -> step2 -> step3 What is the point to use pipeline feature versus hav...

Hi I have one self hosted service,which is accessible only when connected to vpn. I am unable to access the url of self hosted service in github actions CI r...

I am having an issue publishing a completed model training. The model has been deployed on remote compute, using a docker image, and the datasets have been s...

Let's say I have a project call Proj1 to store datasets with Type "Data Process".. What is the best practice to get the latest datasets ? Example, I start th...

Hi all! Question about pipelines using decorators. The first step of my pipeline uses a specific repo, specified using PipelineDecorator.component(repo='repo...

Hey all, is there a way to upload a fiftyone dataset as an artifact in a clearml pipeline? I am getting the following error when I try to upload it Can't pic...

I noticed that Task.get_tasks(project_name=project) is quite slow in my project of over 300 experiments. Is there a faster way of receiving task objects?

Hey! does anyone know if I can use different ports for my clearml UI server?

Hey again 😁 I am migrating my trains-server to AWS and I would like now to have secure accounts (with password). But I don't want to loose the current users...