After I Finish Training A Model, I Want To Call Logger.Report_Scalars To Help Monitor Inferencing Status (We Do A Lot Of Batch) But After The Model Finishes Training, Scalars Are No Longer Accepted By The Task As It Is Considered Completed. Help!

task = Task.get_task("task id here")
task.mark_started(force=True)
task.get_logger().report_scalar(...)
task.mark_completed(force=True)