Follow-up to #220 : now that we're outputting with the new handler, we're getting duplicate log entries. Example: [job]
2024-07-11 14:45:25.745 PDT Ran prediction in 3.0 s; success: True
2024-07-11 14:45:25.745 PDT INFO:deepcell_imaging:Ran prediction in 3.0 s; success: True
2024-07-11 14:45:25.745 PDT Saving raw predictions output to gs://deepcell-batch-jobs_us-central1/job-runs/j3d36718b-a686-4bd5-9035-fb6c59e6c68a/raw_predictions.npz.gz
2024-07-11 14:45:25.745 PDT INFO:deepcell_imaging:Saving raw predictions output to gs://deepcell-batch-jobs_us-central1/job-runs/j3d36718b-a686-4bd5-9035-fb6c59e6c68a/raw_predictions.npz.gz
2024-07-11 14:45:26.251 PDT Saved output in 0.51 s
2024-07-11 14:45:26.251 PDT INFO:deepcell_imaging:Saved output in 0.51 s
Screenshot:
We have 2 options:
Remove the handler from the default, top-level logger
Change the deepcell logger to not propagate
I lean toward option 1, with the further change of adding the GCP-formatted handler to the top-level logger so all logs get handled properly.
Follow-up to #220 : now that we're outputting with the new handler, we're getting duplicate log entries. Example: [job]
Screenshot:
We have 2 options:
I lean toward option 1, with the further change of adding the GCP-formatted handler to the top-level logger so all logs get handled properly.