MonitoringSchedule ​
The MonitoringSchedule resource lets you manage AWS SageMaker MonitoringSchedules for monitoring model performance and data quality over time.
Minimal Example ​
Create a basic monitoring schedule with required properties and one optional property.
ts
import AWS from "alchemy/aws/control";
const basicMonitoringSchedule = await AWS.SageMaker.MonitoringSchedule("basicMonitoringSchedule", {
MonitoringScheduleName: "ModelPerformanceMonitoring",
MonitoringScheduleConfig: {
// Example configuration for monitoring
MonitoringJobDefinition: {
// Define the monitoring job here
BaselineConfig: {
BaselineParameters: {
// Parameters for baseline
},
Statistics: {
// Statistics for monitoring
}
},
// Additional configurations...
}
},
EndpointName: "my-endpoint" // Optional
});
Advanced Configuration ​
Configure a monitoring schedule with detailed settings and tags for resource management.
ts
const advancedMonitoringSchedule = await AWS.SageMaker.MonitoringSchedule("advancedMonitoringSchedule", {
MonitoringScheduleName: "AdvancedModelMonitoring",
MonitoringScheduleConfig: {
MonitoringJobDefinition: {
BaselineConfig: {
BaselineParameters: {
// Define baseline parameters
},
Statistics: {
// Define statistics
}
},
// Additional configurations...
}
},
Tags: [
{
Key: "Environment",
Value: "Production"
},
{
Key: "Project",
Value: "ModelMonitoring"
}
]
});
Error Handling ​
Create a monitoring schedule with options to handle failure scenarios.
ts
const errorHandlingMonitoringSchedule = await AWS.SageMaker.MonitoringSchedule("errorHandlingMonitoringSchedule", {
MonitoringScheduleName: "ErrorHandlingMonitoring",
MonitoringScheduleConfig: {
MonitoringJobDefinition: {
BaselineConfig: {
BaselineParameters: {
// Baseline parameters
},
Statistics: {
// Statistics
}
},
// Additional configurations...
}
},
FailureReason: "Invalid configuration provided" // Optional, to log the failure reason
});