Hi, I recently updated my clearml to 1.1.2 and a code that was working before now behaves completely differently: I am using the following to log debug sampl...

Hi, is it possible to pass temporary IAM role to the web app could access?

Hi, I have a question regarding the aws-autoscaler: am I understanding correctly that: max_idle_time_min=5 max_spin_up_time_min=10 polling_interval_time_min=...

Hi Guys, I had several times now the following errors poping in agents while executing a task: trains_agent: ERROR: Failed applying git diff: I attached the ...

Hello, in the following context: controller_task = Task.init(...) # This will clone the parent task, enqueue and wait for finished status data_processing_tas...

Hi, I would like to bring awareness on this issue , this impacts my work as I cannot install the older version of torch (1.11.0)

Does trains 0.16 supports pip >=20.2?

Hi there! Is there an easy way to retrieve the site-package directory that was created by an agent from inside a task? Eg. task = Task.init(...) task.add_req...

Hi, coming back with the venv caching: with the following setting: I call Task._update_requirements(["."]) setup.py has the following install_requires=["my-p...

Hey there, I would like to increase the ulimit for the number of files opened at the same time in a ec2 instance. According to this https://stackoverflow.com...

Hey, I have one question regarding the cleanup_service task in the DevOps project: Does it assume that the agent in services mode is in the trains-server mac...

Hi, in the AWS AutoScaler, I am getting the following warning: Warning! exception occurred: APIError: code 400/1004: Worker is not registered: worker=aws:A10...

Is it possible to run an agent, listen to the services queue without using docker?

Hi, is there a way to get some stats about the use of workers? I would like to know, over the past 3 months: Number of training hours per user Number of trai...

Hi there, I have a problem with PyJWT: I am using trains==0.16.4 and trains-agent==0.16.3 in my agents. I installed PyJWT==1.7.1 in the agent (through extra_...

Hi there, I am running a clearml-agent in services mode (with docker) on a machine with two disks: one with the OS (8Go, 91% space used) and one for the data...

Are the various task types available in 0.15? I am getting > 2020-06-09 12:58:53,287 - trains.Task - WARNING - Retrying, previous request failed : 'custom' i...

Hi, I encounter a weird behavior: I have a task A that schedules a task B. Task B is executed on an agent, but with an old commit 🤔 although the branch is p...

Hi all, Would it be possible to make the aws autoscaler log each scale in/out operation in the console to help debugging/understanding the course of events?

The “Manage queue” option in the right tab on a queued experiment is broken in v1.0 (it does nothing)

Hi ClearML team members! Is there any progress made on the clearml-serving repo? I’d love to start using it but I lack a straightforward get started example....

Hi, I would like to use pytorch3d==0.5.0 with torch==1.9.1 on cuda version 110, locally it works, but the clearml agent fails setting up the environment with...

Hi, I would like to switch from the elastic-search service in the docker-compose of the clearml-server to an externally managed, scalable elastic-search clus...

Hi, are the experiments logs stored in s3 or in the trains-server? (When using s3 as artifact storage)

Hi, in one of my agents with CUDA Version: 11.1 (from nvidia-smi), clearml agent 0.17.1 detects version 100 (I can see from experiments logs: agent.cuda_vers...

Hi, a small bug (not really a bug) in the autoscaler: I have p3.2xlarge instances that take a long time to shutdown. With polling_interval_time_min=1 , the a...

Hi, I deleted all archived experiments in a project and I just realized all experiments of all projects were deleted (clearml server v1.0.0) 🤔

Not very important, but small suggestion for the web UI: under the QUEUES tab, in the queues wait time graph, would it be possible to switch from seconds to ...

Hi, I see that there is a new parameter in aws autoscaler: max_spin_up_time_min - What is the difference with max_idle_time_min ?

Is there a way to report a simple series with X and Y coords, X and Y being two lists of same length?