ok so, idk why it helped, but setting base_task_id
instead of base_task_name in the pipe.add_step
function, seems to overcome this

only sometimes, the pipeline runs using local machines

Can you post a minimal example here? Does this always happen or only sometimes? Also how is the pipeline run? Using autoscaler or local machines?

WebApp: 3.16.3-949 • Server: 3.16.1-974 • API: 2.24

You can see the version if you go to settings page. It is in the bottom right of the screen 🙂

?

i can send you our pipeline file and task

we use the clearml hosted server, so i don't know the version

Hi @<1544853695869489152:profile|NonchalantOx99> , how are you running the pipeline? What are the clearml & server versions?

Do you have a snippet that reproduces this?