AgitatedDove14 After checking, I discovered that apparently it doesn't matter if each pipeline is executed by a different worker, the error persists. Honestly this has me puzzled. I'm really looking forward to getting this functionality right because it's an aspect that would make ClearML shine even more.