Sage maker model monitoring data quality error 'FailureReason': 'Job inputs had no data'

0

I am trying to set up a data quality job for a batch transfer model. All inputs look right to me but the job does not kick off and displays above error when job is described.

my_default_monitor.create_monitoring_schedule(
    monitor_schedule_name=mon_schedule_name,    
     batch_transform_input=BatchTransformInput(
        data_captured_destination_s3_uri=s3_data_capture_path,
        destination="/opt/ml/processing/input",
        dataset_format=MonitoringDatasetFormat.csv(header=False)
    ),   
    output_s3_uri=s3_reports_path,
    statistics= statistics_path,
    constraints = constraints_path,    
    schedule_cron_expression=  CronExpressionGenerator.daily(hour=19),
    enable_cloudwatch_metrics=True
)

the file in 's3_data_capture_path' is a .csv file. It is created every month format it being same as training input file for model monitoring baseline job.

Trouble shooting, I did so far as below ... still same error. The error message is not at all helpful!

  1. flip s3_data_capture_path to relative path to path to actual file copied from aws console.
  2. try both daily cron job and hourly cron job

Questions Other than SMS notebook, can I trace more log details elsewhere? It does not seem to kick off and seems to fail in validation... Is it not actually finding the file (which is unlikely as I verified this in different ways) or is it looking for a specific format? Could it be possible that the daily run or hourly looks for values in the last hour or last day?

Please suggest!

No hay respuestas

No has iniciado sesión. Iniciar sesión para publicar una respuesta.

Una buena respuesta responde claramente a la pregunta, proporciona comentarios constructivos y fomenta el crecimiento profesional en la persona que hace la pregunta.

Pautas para responder preguntas