Yes I mean the python packages, I would prefer to have all the installation once and if a new package is used just to install the new package not every time the same packages get installed many times.

I am running the agent by (clearml-agent daemon) so I thing it is running using virtual environment mode not docker

yes Just i want to make sure it will work as intended so just queue 3 will take any idle GPU when there is a task to pull if there is no idle GPU will be 2 GPUs only

I am storing it here by the below code

is it this? I can copy a URL from here

ok it works thank you

I will check with another model to see

I am running this part of code

No thing in the uncommitted changes

yes thank you

do I need to change USER (with my device name) or it is a keyword

I tried to run this command but still the same

where to find it?

i run the code again then enqueue the task and for now it takes more than 5 min and still pending in the queue, I will see when it will run 😓

is it normal that it's slower than my device even though the agent is much more powerful than my device? or because it is just a simple code

I don't understand what you mean by that

Is it related to the python versions? When I run the same code on my first device through clearML to the agent it works fine (the version of the first device is 3.10.6) but I think it is not related to the version

3.10.7

how to put or handle this configuration and where?

I am trying to access the dataset by the below code

ok the problem solved by "pip install virtualenv" on the agent 😁

how to check it?

thanks alot it run successfully

I posted another question if anyone is free to help and thanks a lot for you all for your kind help

clearml==1.7.0

I run it in my local device on vs code

after installing nbconvert all the code is visible on the uncommitted changed

so how to get the path for the file itself not all the files? 😓