I have been trying to initiate 2 projects that involve using auto ml time series forecasting. I am using python software package to make the calls.
Usually they take some time to train and run inference but yesterday, I was noticing the timing was taking very long. Not only that, after waiting for over 30 minutes, I saw on the GUI that my jobs were failing at calculating feature importances. This is puzzling because I have ran the same projects (same data, different date) just to test things and this has never happened. Today I am waiting 40+ minutes and the calls are still "running". It never takes this long for only 2 projects. I am not at my limit for max workers either.