python 3.9.12 / clearml 1.13.1 / MacOS Ventura 13.0.1 / torch 2.1.0
Does it wok if you remove the Task.init call?
Is this reproducible? I tried to run the same example code on my machine, and it started training ...
Do you have issues with other pytorch examples? Could you try simple reporting example:
https://github.com/allegroai/clearml/blob/master/examples/reporting/scalar_reporting.py
Hi @<1625303806923247616:profile|ItchyCow80>
Could you add some prints ? Is it working without the Task.init call? the code looks okay and the - No repository found, message basically says it logs it as a standalone script (which makes sense)
Yes, there are no other messages. And it was recorded on the ClearML server through the above training code. But now it doesn’t work with the message I mentioned.
Hi @<1625303806923247616:profile|ItchyCow80> , can you please describe how you're running it? Is it inside a jupyter nokebook? Do you have a code sample?
Hmm what's the clearml version? Whats the python version, whats the OS? And pytorch version?
Okay. thanks, but this code is already calling it within main().
and then, when remove clearml.Logger module, the code works well
Could you verify the Task.init call is inside the main function and Not the global scope? We have noticed some issues with global scope calls in some cases
when remove Task.init, i found error message “AttributeError: ‘NoneType’ object has no attribute ‘report_scalar’ ” in clearml.Logger module