EndpointConfigurationProductionVariant class
Constructors
-
EndpointConfigurationProductionVariant({Input<String?>? acceleratorType, Input<EndpointConfigurationProductionVariantCapacityReservationConfig?>? capacityReservationConfig, Input<int?>? containerStartupHealthCheckTimeoutInSeconds, Input<EndpointConfigurationProductionVariantCoreDumpConfig?>? coreDumpConfig, Input<bool?>? enableSsmAccess, Input<String?>? inferenceAmiVersion, Input<int?>? initialInstanceCount, Input<double?>? initialVariantWeight, Input<String?>? instanceType, Input<EndpointConfigurationProductionVariantManagedInstanceScaling?>? managedInstanceScaling, Input<int?>? modelDataDownloadTimeoutInSeconds, Input<String?>? modelName, Input<List<EndpointConfigurationProductionVariantRoutingConfig>?>? routingConfigs, Input<EndpointConfigurationProductionVariantServerlessConfig?>? serverlessConfig, Input<String?>? variantName, Input<int?>? volumeSizeInGb})
-
Creates a new EndpointConfigurationProductionVariant.
acceleratorType Size of the Elastic Inference (EI) instance to use for the production variant.
capacityReservationConfig Settings for the capacity reservation for the compute instances that SageMaker AI reserves for an endpoint. See capacityReservationConfig below.
containerStartupHealthCheckTimeoutInSeconds Timeout value, in seconds, for your inference container to pass health check by SageMaker AI Hosting. For more information about health check, see How Your Container Should Respond to Health Check (Ping) Requests. Valid values between 60 and 3600.
coreDumpConfig Core dump configuration from the model container when the process crashes. Fields are documented below.
enableSsmAccess Whether to turn on native AWS SSM access for a production variant behind an endpoint. By default, SSM access is disabled for all production variants behind endpoints. Ignored if modelName is not set (Inference Components endpoint).
inferenceAmiVersion Option from a collection of preconfigured AMI images. Each image is configured by AWS with a set of software and driver versions. AWS optimizes these configurations for different machine learning workloads.
initialInstanceCount Initial number of instances used for auto-scaling.
initialVariantWeight Initial traffic distribution among all of the models that you specify in the endpoint configuration. If unspecified, defaults to 1.0. Ignored if modelName is not set (Inference Components endpoint).
instanceType Type of instance to start.
managedInstanceScaling Control the range in the number of instances that the endpoint provisions as it scales up or down to accommodate traffic.
modelDataDownloadTimeoutInSeconds Timeout value, in seconds, to download and extract the model that you want to host from S3 to the individual inference instance associated with this production variant. Valid values between 60 and 3600.
modelName Name of the model to use. Required unless using Inference Components (in which case executionRoleArn must be specified at the endpoint configuration level).
routingConfigs How the endpoint routes incoming traffic. See routingConfig below.
serverlessConfig How an endpoint performs asynchronous inference.
variantName Name of the variant. If omitted, the provider will assign a random, unique name.
volumeSizeInGb Size, in GB, of the ML storage volume attached to individual inference instance associated with the production variant. Valid values between 1 and 512.
const
-
EndpointConfigurationProductionVariant.fromMap(Map<String, dynamic> map)
-
factory
Properties
-
acceleratorType
→ Input<String?>?
-
Size of the Elastic Inference (EI) instance to use for the production variant.
final
-
capacityReservationConfig
→ Input<EndpointConfigurationProductionVariantCapacityReservationConfig?>?
-
Settings for the capacity reservation for the compute instances that SageMaker AI reserves for an endpoint. See capacityReservationConfig below.
final
-
containerStartupHealthCheckTimeoutInSeconds
→ Input<int?>?
-
Timeout value, in seconds, for your inference container to pass health check by SageMaker AI Hosting. For more information about health check, see How Your Container Should Respond to Health Check (Ping) Requests. Valid values between
60 and 3600.
final
-
coreDumpConfig
→ Input<EndpointConfigurationProductionVariantCoreDumpConfig?>?
-
Core dump configuration from the model container when the process crashes. Fields are documented below.
final
-
enableSsmAccess
→ Input<bool?>?
-
Whether to turn on native AWS SSM access for a production variant behind an endpoint. By default, SSM access is disabled for all production variants behind endpoints. Ignored if
modelName is not set (Inference Components endpoint).
final
-
hashCode
→ int
-
The hash code for this object.
no setterinherited
-
inferenceAmiVersion
→ Input<String?>?
-
Option from a collection of preconfigured AMI images. Each image is configured by AWS with a set of software and driver versions. AWS optimizes these configurations for different machine learning workloads.
final
-
initialInstanceCount
→ Input<int?>?
-
Initial number of instances used for auto-scaling.
final
-
initialVariantWeight
→ Input<double?>?
-
Initial traffic distribution among all of the models that you specify in the endpoint configuration. If unspecified, defaults to
1.0. Ignored if modelName is not set (Inference Components endpoint).
final
-
instanceType
→ Input<String?>?
-
Type of instance to start.
final
-
managedInstanceScaling
→ Input<EndpointConfigurationProductionVariantManagedInstanceScaling?>?
-
Control the range in the number of instances that the endpoint provisions as it scales up or down to accommodate traffic.
final
-
modelDataDownloadTimeoutInSeconds
→ Input<int?>?
-
Timeout value, in seconds, to download and extract the model that you want to host from S3 to the individual inference instance associated with this production variant. Valid values between
60 and 3600.
final
-
modelName
→ Input<String?>?
-
Name of the model to use. Required unless using Inference Components (in which case
executionRoleArn must be specified at the endpoint configuration level).
final
-
routingConfigs
→ Input<List<EndpointConfigurationProductionVariantRoutingConfig>?>?
-
How the endpoint routes incoming traffic. See routingConfig below.
final
-
runtimeType
→ Type
-
A representation of the runtime type of the object.
no setterinherited
-
serverlessConfig
→ Input<EndpointConfigurationProductionVariantServerlessConfig?>?
-
How an endpoint performs asynchronous inference.
final
-
variantName
→ Input<String?>?
-
Name of the variant. If omitted, the provider will assign a random, unique name.
final
-
volumeSizeInGb
→ Input<int?>?
-
Size, in GB, of the ML storage volume attached to individual inference instance associated with the production variant. Valid values between
1 and 512.
final
