libZSservicesZSamazonka-sagemakerZSamazonka-sagemaker
Copyright(c) 2013-2021 Brendan Hay
LicenseMozilla Public License, v. 2.0.
MaintainerBrendan Hay <brendan.g.hay+amazonka@gmail.com>
Stabilityauto-generated
Portabilitynon-portable (GHC extensions)
Safe HaskellNone

Amazonka.SageMaker.Types.TrainingSpecification

Description

 
Synopsis

Documentation

data TrainingSpecification Source #

Defines how the algorithm is used for a training job.

See: newTrainingSpecification smart constructor.

Constructors

TrainingSpecification' 

Fields

Instances

Instances details
Eq TrainingSpecification Source # 
Instance details

Defined in Amazonka.SageMaker.Types.TrainingSpecification

Read TrainingSpecification Source # 
Instance details

Defined in Amazonka.SageMaker.Types.TrainingSpecification

Show TrainingSpecification Source # 
Instance details

Defined in Amazonka.SageMaker.Types.TrainingSpecification

Generic TrainingSpecification Source # 
Instance details

Defined in Amazonka.SageMaker.Types.TrainingSpecification

Associated Types

type Rep TrainingSpecification :: Type -> Type #

NFData TrainingSpecification Source # 
Instance details

Defined in Amazonka.SageMaker.Types.TrainingSpecification

Methods

rnf :: TrainingSpecification -> () #

Hashable TrainingSpecification Source # 
Instance details

Defined in Amazonka.SageMaker.Types.TrainingSpecification

ToJSON TrainingSpecification Source # 
Instance details

Defined in Amazonka.SageMaker.Types.TrainingSpecification

FromJSON TrainingSpecification Source # 
Instance details

Defined in Amazonka.SageMaker.Types.TrainingSpecification

type Rep TrainingSpecification Source # 
Instance details

Defined in Amazonka.SageMaker.Types.TrainingSpecification

type Rep TrainingSpecification = D1 ('MetaData "TrainingSpecification" "Amazonka.SageMaker.Types.TrainingSpecification" "libZSservicesZSamazonka-sagemakerZSamazonka-sagemaker" 'False) (C1 ('MetaCons "TrainingSpecification'" 'PrefixI 'True) (((S1 ('MetaSel ('Just "trainingImageDigest") 'NoSourceUnpackedness 'NoSourceStrictness 'DecidedStrict) (Rec0 (Maybe Text)) :*: S1 ('MetaSel ('Just "supportsDistributedTraining") 'NoSourceUnpackedness 'NoSourceStrictness 'DecidedStrict) (Rec0 (Maybe Bool))) :*: (S1 ('MetaSel ('Just "supportedHyperParameters") 'NoSourceUnpackedness 'NoSourceStrictness 'DecidedStrict) (Rec0 (Maybe [HyperParameterSpecification])) :*: S1 ('MetaSel ('Just "supportedTuningJobObjectiveMetrics") 'NoSourceUnpackedness 'NoSourceStrictness 'DecidedStrict) (Rec0 (Maybe [HyperParameterTuningJobObjective])))) :*: ((S1 ('MetaSel ('Just "metricDefinitions") 'NoSourceUnpackedness 'NoSourceStrictness 'DecidedStrict) (Rec0 (Maybe [MetricDefinition])) :*: S1 ('MetaSel ('Just "trainingImage") 'NoSourceUnpackedness 'NoSourceStrictness 'DecidedStrict) (Rec0 Text)) :*: (S1 ('MetaSel ('Just "supportedTrainingInstanceTypes") 'NoSourceUnpackedness 'NoSourceStrictness 'DecidedStrict) (Rec0 [TrainingInstanceType]) :*: S1 ('MetaSel ('Just "trainingChannels") 'NoSourceUnpackedness 'NoSourceStrictness 'DecidedStrict) (Rec0 (NonEmpty ChannelSpecification))))))

newTrainingSpecification Source #

Create a value of TrainingSpecification with all optional fields omitted.

Use generic-lens or optics to modify other optional fields.

The following record fields are available, with the corresponding lenses provided for backwards compatibility:

$sel:trainingImageDigest:TrainingSpecification', trainingSpecification_trainingImageDigest - An MD5 hash of the training algorithm that identifies the Docker image used for training.

$sel:supportsDistributedTraining:TrainingSpecification', trainingSpecification_supportsDistributedTraining - Indicates whether the algorithm supports distributed training. If set to false, buyers can't request more than one instance during training.

$sel:supportedHyperParameters:TrainingSpecification', trainingSpecification_supportedHyperParameters - A list of the HyperParameterSpecification objects, that define the supported hyperparameters. This is required if the algorithm supports automatic model tuning.>

$sel:supportedTuningJobObjectiveMetrics:TrainingSpecification', trainingSpecification_supportedTuningJobObjectiveMetrics - A list of the metrics that the algorithm emits that can be used as the objective metric in a hyperparameter tuning job.

$sel:metricDefinitions:TrainingSpecification', trainingSpecification_metricDefinitions - A list of MetricDefinition objects, which are used for parsing metrics generated by the algorithm.

$sel:trainingImage:TrainingSpecification', trainingSpecification_trainingImage - The Amazon ECR registry path of the Docker image that contains the training algorithm.

$sel:supportedTrainingInstanceTypes:TrainingSpecification', trainingSpecification_supportedTrainingInstanceTypes - A list of the instance types that this algorithm can use for training.

$sel:trainingChannels:TrainingSpecification', trainingSpecification_trainingChannels - A list of ChannelSpecification objects, which specify the input sources to be used by the algorithm.

trainingSpecification_trainingImageDigest :: Lens' TrainingSpecification (Maybe Text) Source #

An MD5 hash of the training algorithm that identifies the Docker image used for training.

trainingSpecification_supportsDistributedTraining :: Lens' TrainingSpecification (Maybe Bool) Source #

Indicates whether the algorithm supports distributed training. If set to false, buyers can't request more than one instance during training.

trainingSpecification_supportedHyperParameters :: Lens' TrainingSpecification (Maybe [HyperParameterSpecification]) Source #

A list of the HyperParameterSpecification objects, that define the supported hyperparameters. This is required if the algorithm supports automatic model tuning.>

trainingSpecification_supportedTuningJobObjectiveMetrics :: Lens' TrainingSpecification (Maybe [HyperParameterTuningJobObjective]) Source #

A list of the metrics that the algorithm emits that can be used as the objective metric in a hyperparameter tuning job.

trainingSpecification_metricDefinitions :: Lens' TrainingSpecification (Maybe [MetricDefinition]) Source #

A list of MetricDefinition objects, which are used for parsing metrics generated by the algorithm.

trainingSpecification_trainingImage :: Lens' TrainingSpecification Text Source #

The Amazon ECR registry path of the Docker image that contains the training algorithm.

trainingSpecification_supportedTrainingInstanceTypes :: Lens' TrainingSpecification [TrainingInstanceType] Source #

A list of the instance types that this algorithm can use for training.

trainingSpecification_trainingChannels :: Lens' TrainingSpecification (NonEmpty ChannelSpecification) Source #

A list of ChannelSpecification objects, which specify the input sources to be used by the algorithm.