| Column Name | Column Data Type |
|---|
algorithm_name Required Input Column
The name of the algorithm being described. | VARCHAR |
_aws_profile Input Column
The AWS profile defines the AWS identity used. It can be defined via credentials or by assuming a IAM role. | STRUCT(
"type" VARCHAR,
"name" VARCHAR,
"account_id" VARCHAR,
"via_profile_name" VARCHAR,
"assumed_role_arn" VARCHAR,
"organization" STRUCT(
"account_name" VARCHAR,
"id" VARCHAR,
"tags" STRUCT(
"key" VARCHAR,
"value" VARCHAR
)[],
"master_account" STRUCT(
"id" VARCHAR,
"email" VARCHAR
),
"parents" STRUCT(
"type" VARCHAR,
"id" VARCHAR,
"name" VARCHAR,
"tags" STRUCT(
"key" VARCHAR,
"value" VARCHAR
)[]
)[]
)
) |
Show child fields- _aws_profile.account_id
The AWS account id - _aws_profile.assumed_role_arn
The ARN of the assumed role - _aws_profile.name
The unique name of the profile. - _aws_profile.organization
Information about this profile's membership in the AWS organization. Show child fields- _aws_profile.organization.account_name
The name of account speciifed by the organization - _aws_profile.organization.id
The organization id - _aws_profile.organization.master_account
Show child fields- _aws_profile.organization.master_account.email
The organization master account email address - _aws_profile.organization.master_account.id
The organization master account id
- _aws_profile.organization.parents[]
Show child fields- _aws_profile.organization.parents[].id
The id of the parent - _aws_profile.organization.parents[].name
The name of the parent - _aws_profile.organization.parents[].tags[]
Show child fields- _aws_profile.organization.parents[].tags[].key
- _aws_profile.organization.parents[].tags[].value
- _aws_profile.organization.parents[].type
The type of parent can be an organization unit or a root
- _aws_profile.organization.tags[]
Show child fields- _aws_profile.organization.tags[].key
- _aws_profile.organization.tags[].value
- _aws_profile.type
The type of profile, either 'credentials' or 'assumed_role' - _aws_profile.via_profile_name
This IAM role for this profile is assumed by first utilizing another profile with this name to obtain credentials.
|
algorithm_arn
The Amazon Resource Name (ARN) of the algorithm. | VARCHAR |
algorithm_description
A brief summary about the algorithm. | VARCHAR |
algorithm_status
The current status of the algorithm. | VARCHAR |
algorithm_status_details
Details about the current status of the algorithm. | STRUCT(
"validation_statuses" STRUCT(
"name" VARCHAR,
"status" VARCHAR,
"failure_reason" VARCHAR
)[],
"image_scan_statuses" STRUCT(
"name" VARCHAR,
"status" VARCHAR,
"failure_reason" VARCHAR
)[]
) |
Show child fields- algorithm_status_details.image_scan_statuses[]
Show child fields- algorithm_status_details.image_scan_statuses[].failure_reason
if the overall status is Failed, the reason for the failure. - algorithm_status_details.image_scan_statuses[].name
The name of the algorithm for which the overall status is being reported. - algorithm_status_details.image_scan_statuses[].status
The current status.
- algorithm_status_details.validation_statuses[]
Show child fields- algorithm_status_details.validation_statuses[].failure_reason
if the overall status is Failed, the reason for the failure. - algorithm_status_details.validation_statuses[].name
The name of the algorithm for which the overall status is being reported. - algorithm_status_details.validation_statuses[].status
The current status.
|
certify_for_marketplace
Whether the algorithm is certified to be listed in Amazon Web Services Marketplace. | BOOLEAN |
creation_time
A timestamp specifying when the algorithm was created. | TIMESTAMP_S |
inference_specification
Details about inference jobs that the algorithm runs. | STRUCT(
"containers" STRUCT(
"container_hostname" VARCHAR,
"image" VARCHAR,
"image_digest" VARCHAR,
"model_data_url" VARCHAR,
"model_data_source" STRUCT(
"s3_data_source" STRUCT(
"s3_uri" VARCHAR,
"s3_data_type" VARCHAR,
"compression_type" VARCHAR,
"model_access_config" STRUCT(
"accept_eula" BOOLEAN
),
"hub_access_config" STRUCT(
"hub_content_arn" VARCHAR
)
)
),
"product_id" VARCHAR,
"environment" MAP(VARCHAR, VARCHAR),
"model_input" STRUCT(
"data_input_config" VARCHAR
),
"framework" VARCHAR,
"framework_version" VARCHAR,
"nearest_model_name" VARCHAR,
"additional_s3_data_source" STRUCT(
"s3_data_type" VARCHAR,
"s3_uri" VARCHAR,
"compression_type" VARCHAR
)
)[],
"supported_transform_instance_types" VARCHAR[],
"supported_realtime_inference_instance_types" VARCHAR[],
"supported_content_types" VARCHAR[],
"supported_response_mime_types" VARCHAR[]
) |
Show child fields- inference_specification.containers[]
Show child fields- inference_specification.containers[].additional_s3_data_source
The additional data source that is used during inference in the Docker container for your model package. Show child fields- inference_specification.containers[].additional_s3_data_source.compression_type
The type of compression used for an additional data source used in inference or training. Specify None if your additional data source is not compressed. - inference_specification.containers[].additional_s3_data_source.s3_data_type
The data type of the additional data source that you specify for use in inference or training. - inference_specification.containers[].additional_s3_data_source.s3_uri
The uniform resource identifier (URI) used to identify an additional data source used in inference or training.
- inference_specification.containers[].container_hostname
The DNS host name for the Docker container. - inference_specification.containers[].environment
The environment variables to set in the Docker container. Each key and value in the Environment string to string map can have length of up to 1024. We support up to 16 entries in the map. - inference_specification.containers[].framework
The machine learning framework of the model package container image. - inference_specification.containers[].framework_version
The framework version of the Model Package Container Image. - inference_specification.containers[].image
The Amazon EC2 Container Registry (Amazon ECR) path where inference code is stored. If you are using your own custom algorithm instead of an algorithm provided by SageMaker, the inference code must meet SageMaker requirements. SageMaker supports both registry/repository[:tag] and registry/repository[@digest] image path formats. For more information, see Using Your Own Algorithms with Amazon SageMaker. - inference_specification.containers[].image_digest
An MD5 hash of the training algorithm that identifies the Docker image used for training. - inference_specification.containers[].model_data_source
Specifies the location of ML model data to deploy during endpoint creation. Show child fields- inference_specification.containers[].model_data_source.s3_data_source
Specifies the S3 location of ML model data to deploy. Show child fields- inference_specification.containers[].model_data_source.s3_data_source.compression_type
Specifies how the ML model data is prepared. If you choose Gzip and choose S3Object as the value of S3DataType, S3Uri identifies an object that is a gzip-compressed TAR archive. SageMaker will attempt to decompress and untar the object during model deployment. If you choose None and chooose S3Object as the value of S3DataType, S3Uri identifies an object that represents an uncompressed ML model to deploy. If you choose None and choose S3Prefix as the value of S3DataType, S3Uri identifies a key name prefix, under which all objects represents the uncompressed ML model to deploy. If you choose None, then SageMaker will follow rules below when creating model data files under /opt/ml/model directory for use by your inference code: -
If you choose S3Object as the value of S3DataType, then SageMaker will split the key of the S3 object referenced by S3Uri by slash (/), and use the last part as the filename of the file holding the content of the S3 object. -
If you choose S3Prefix as the value of S3DataType, then for each S3 object under the key name pefix referenced by S3Uri, SageMaker will trim its key by the prefix, and use the remainder as the path (relative to /opt/ml/model) of the file holding the content of the S3 object. SageMaker will split the remainder by slash (/), using intermediate parts as directory names and the last part as filename of the file holding the content of the S3 object. -
Do not use any of the following as file names or directory names: -
Ambiguous file names will result in model deployment failure. For example, if your uncompressed ML model consists of two S3 objects s3://mybucket/model/weights and s3://mybucket/model/weights/part1 and you specify s3://mybucket/model/ as the value of S3Uri and S3Prefix as the value of S3DataType, then it will result in name clash between /opt/ml/model/weights (a regular file) and /opt/ml/model/weights/ (a directory). -
Do not organize the model artifacts in S3 console using folders. When you create a folder in S3 console, S3 creates a 0-byte object with a key set to the folder name you provide. They key of the 0-byte object ends with a slash (/) which violates SageMaker restrictions on model artifact file names, leading to model deployment failure. - inference_specification.containers[].model_data_source.s3_data_source.hub_access_config
Configuration information for hub access. Show child fields- inference_specification.containers[].model_data_source.s3_data_source.hub_access_config.hub_content_arn
The ARN of the hub content for which deployment access is allowed.
- inference_specification.containers[].model_data_source.s3_data_source.model_access_config
Specifies the access configuration file for the ML model. You can explicitly accept the model end-user license agreement (EULA) within the ModelAccessConfig. You are responsible for reviewing and complying with any applicable license terms and making sure they are acceptable for your use case before downloading or using a model. Show child fields- inference_specification.containers[].model_data_source.s3_data_source.model_access_config.accept_eula
Specifies agreement to the model end-user license agreement (EULA). The AcceptEula value must be explicitly defined as True in order to accept the EULA that this model requires. You are responsible for reviewing and complying with any applicable license terms and making sure they are acceptable for your use case before downloading or using a model.
- inference_specification.containers[].model_data_source.s3_data_source.s3_data_type
Specifies the type of ML model data to deploy. If you choose S3Prefix, S3Uri identifies a key name prefix. SageMaker uses all objects that match the specified key name prefix as part of the ML model data to deploy. A valid key name prefix identified by S3Uri always ends with a forward slash (/). If you choose S3Object, S3Uri identifies an object that is the ML model data to deploy. - inference_specification.containers[].model_data_source.s3_data_source.s3_uri
Specifies the S3 path of ML model data to deploy.
- inference_specification.containers[].model_data_url
The Amazon S3 path where the model artifacts, which result from model training, are stored. This path must point to a single gzip compressed tar archive (.tar.gz suffix). The model artifacts must be in an S3 bucket that is in the same region as the model package. - inference_specification.containers[].model_input
A structure with Model Input details. Show child fields- inference_specification.containers[].model_input.data_input_config
The input configuration object for the model.
- inference_specification.containers[].nearest_model_name
The name of a pre-trained machine learning benchmarked by Amazon SageMaker Inference Recommender model that matches your model. You can find a list of benchmarked models by calling ListModelMetadata. - inference_specification.containers[].product_id
The Amazon Web Services Marketplace product ID of the model package.
- inference_specification.supported_content_types[]
- inference_specification.supported_realtime_inference_instance_types[]
- inference_specification.supported_response_mime_types[]
- inference_specification.supported_transform_instance_types[]
|
product_id
The product identifier of the algorithm. | VARCHAR |
training_specification
Details about training jobs run by this algorithm. | STRUCT(
"training_image" VARCHAR,
"training_image_digest" VARCHAR,
"supported_hyper_parameters" STRUCT(
"name" VARCHAR,
"description" VARCHAR,
"type" VARCHAR,
"range" STRUCT(
"integer_parameter_range_specification" STRUCT(
"min_value" VARCHAR,
"max_value" VARCHAR
),
"continuous_parameter_range_specification" STRUCT(
"min_value" VARCHAR,
"max_value" VARCHAR
),
"categorical_parameter_range_specification" STRUCT(
"values" VARCHAR[]
)
),
"is_tunable" BOOLEAN,
"is_required" BOOLEAN,
"default_value" VARCHAR
)[],
"supported_training_instance_types" VARCHAR[],
"supports_distributed_training" BOOLEAN,
"metric_definitions" STRUCT(
"name" VARCHAR,
"regex" VARCHAR
)[],
"training_channels" STRUCT(
"name" VARCHAR,
"description" VARCHAR,
"is_required" BOOLEAN,
"supported_content_types" VARCHAR[],
"supported_compression_types" VARCHAR[],
"supported_input_modes" VARCHAR[]
)[],
"supported_tuning_job_objective_metrics" STRUCT(
"type" VARCHAR,
"metric_name" VARCHAR
)[],
"additional_s3_data_source" STRUCT(
"s3_data_type" VARCHAR,
"s3_uri" VARCHAR,
"compression_type" VARCHAR
)
) |
Show child fields- training_specification.additional_s3_data_source
The additional data source used during the training job. Show child fields- training_specification.additional_s3_data_source.compression_type
The type of compression used for an additional data source used in inference or training. Specify None if your additional data source is not compressed. - training_specification.additional_s3_data_source.s3_data_type
The data type of the additional data source that you specify for use in inference or training. - training_specification.additional_s3_data_source.s3_uri
The uniform resource identifier (URI) used to identify an additional data source used in inference or training.
- training_specification.metric_definitions[]
Show child fields- training_specification.metric_definitions[].name
The name of the metric. - training_specification.metric_definitions[].regex
A regular expression that searches the output of a training job and gets the value of the metric. For more information about using regular expressions to define metrics, see Defining metrics and environment variables.
- training_specification.supported_hyper_parameters[]
Show child fields- training_specification.supported_hyper_parameters[].default_value
The default value for this hyperparameter. If a default value is specified, a hyperparameter cannot be required. - training_specification.supported_hyper_parameters[].description
A brief description of the hyperparameter. - training_specification.supported_hyper_parameters[].is_required
Indicates whether this hyperparameter is required. - training_specification.supported_hyper_parameters[].is_tunable
Indicates whether this hyperparameter is tunable in a hyperparameter tuning job. - training_specification.supported_hyper_parameters[].name
The name of this hyperparameter. The name must be unique. - training_specification.supported_hyper_parameters[].range
The allowed range for this hyperparameter. Show child fields- training_specification.supported_hyper_parameters[].range.categorical_parameter_range_specification
A CategoricalParameterRangeSpecification object that defines the possible values for a categorical hyperparameter. Show child fields- training_specification.supported_hyper_parameters[].range.categorical_parameter_range_specification.values[]
- training_specification.supported_hyper_parameters[].range.continuous_parameter_range_specification
A ContinuousParameterRangeSpecification object that defines the possible values for a continuous hyperparameter. Show child fields- training_specification.supported_hyper_parameters[].range.continuous_parameter_range_specification.max_value
The maximum floating-point value allowed. - training_specification.supported_hyper_parameters[].range.continuous_parameter_range_specification.min_value
The minimum floating-point value allowed.
- training_specification.supported_hyper_parameters[].range.integer_parameter_range_specification
A IntegerParameterRangeSpecification object that defines the possible values for an integer hyperparameter. Show child fields- training_specification.supported_hyper_parameters[].range.integer_parameter_range_specification.max_value
The maximum integer value allowed. - training_specification.supported_hyper_parameters[].range.integer_parameter_range_specification.min_value
The minimum integer value allowed.
- training_specification.supported_hyper_parameters[].type
The type of this hyperparameter. The valid types are Integer, Continuous, Categorical, and FreeText.
- training_specification.supported_training_instance_types[]
- training_specification.supported_tuning_job_objective_metrics[]
Show child fields- training_specification.supported_tuning_job_objective_metrics[].metric_name
The name of the metric to use for the objective metric. - training_specification.supported_tuning_job_objective_metrics[].type
Whether to minimize or maximize the objective metric.
- training_specification.supports_distributed_training
Indicates whether the algorithm supports distributed training. If set to false, buyers can't request more than one instance during training. - training_specification.training_channels[]
Show child fields- training_specification.training_channels[].description
A brief description of the channel. - training_specification.training_channels[].is_required
Indicates whether the channel is required by the algorithm. - training_specification.training_channels[].name
The name of the channel. - training_specification.training_channels[].supported_compression_types[]
- training_specification.training_channels[].supported_content_types[]
- training_specification.training_channels[].supported_input_modes[]
- training_specification.training_image
The Amazon ECR registry path of the Docker image that contains the training algorithm. - training_specification.training_image_digest
An MD5 hash of the training algorithm that identifies the Docker image used for training.
|
validation_specification
Details about configurations for one or more training jobs that SageMaker runs to test the algorithm. | STRUCT(
"validation_role" VARCHAR,
"validation_profiles" STRUCT(
"profile_name" VARCHAR,
"training_job_definition" STRUCT(
"training_input_mode" VARCHAR,
"hyper_parameters" MAP(VARCHAR, VARCHAR),
"input_data_config" STRUCT(
"channel_name" VARCHAR,
"data_source" STRUCT(
"s3_data_source" STRUCT(
"s3_data_type" VARCHAR,
"s3_uri" VARCHAR,
"s3_data_distribution_type" VARCHAR,
"attribute_names" VARCHAR[],
"instance_group_names" VARCHAR[]
),
"file_system_data_source" STRUCT(
"file_system_id" VARCHAR,
"file_system_access_mode" VARCHAR,
"file_system_type" VARCHAR,
"directory_path" VARCHAR
)
),
"content_type" VARCHAR,
"compression_type" VARCHAR,
"record_wrapper_type" VARCHAR,
"input_mode" VARCHAR,
"shuffle_config" STRUCT(
"seed" BIGINT
)
)[],
"output_data_config" STRUCT(
"kms_key_id" VARCHAR,
"s3_output_path" VARCHAR,
"compression_type" VARCHAR
),
"resource_config" STRUCT(
"instance_type" VARCHAR,
"instance_count" BIGINT,
"volume_size_in_gb" BIGINT,
"volume_kms_key_id" VARCHAR,
"keep_alive_period_in_seconds" BIGINT,
"instance_groups" STRUCT(
"instance_type" VARCHAR,
"instance_count" BIGINT,
"instance_group_name" VARCHAR
)[]
),
"stopping_condition" STRUCT(
"max_runtime_in_seconds" BIGINT,
"max_wait_time_in_seconds" BIGINT,
"max_pending_time_in_seconds" BIGINT
)
),
"transform_job_definition" STRUCT(
"max_concurrent_transforms" BIGINT,
"max_payload_in_mb" BIGINT,
"batch_strategy" VARCHAR,
"environment" MAP(VARCHAR, VARCHAR),
"transform_input" STRUCT(
"data_source" STRUCT(
"s3_data_source" STRUCT(
"s3_data_type" VARCHAR,
"s3_uri" VARCHAR
)
),
"content_type" VARCHAR,
"compression_type" VARCHAR,
"split_type" VARCHAR
),
"transform_output" STRUCT(
"s3_output_path" VARCHAR,
"accept" VARCHAR,
"assemble_with" VARCHAR,
"kms_key_id" VARCHAR
),
"transform_resources" STRUCT(
"instance_type" VARCHAR,
"instance_count" BIGINT,
"volume_kms_key_id" VARCHAR
)
)
)[]
) |
Show child fields- validation_specification.validation_profiles[]
Show child fields- validation_specification.validation_profiles[].profile_name
The name of the profile for the algorithm. The name must have 1 to 63 characters. Valid characters are a-z, A-Z, 0-9, and - (hyphen). - validation_specification.validation_profiles[].training_job_definition
The TrainingJobDefinition object that describes the training job that SageMaker runs to validate your algorithm. Show child fields- validation_specification.validation_profiles[].training_job_definition.hyper_parameters
The hyperparameters used for the training job. - validation_specification.validation_profiles[].training_job_definition.input_data_config[]
Show child fields- validation_specification.validation_profiles[].training_job_definition.input_data_config[].channel_name
The name of the channel. - validation_specification.validation_profiles[].training_job_definition.input_data_config[].compression_type
If training data is compressed, the compression type. The default value is None. CompressionType is used only in Pipe input mode. In File mode, leave this field unset or set it to None. - validation_specification.validation_profiles[].training_job_definition.input_data_config[].content_type
The MIME type of the data. - validation_specification.validation_profiles[].training_job_definition.input_data_config[].data_source
The location of the channel data. Show child fields- validation_specification.validation_profiles[].training_job_definition.input_data_config[].data_source.file_system_data_source
The file system that is associated with a channel. Show child fields- validation_specification.validation_profiles[].training_job_definition.input_data_config[].data_source.file_system_data_source.directory_path
The full path to the directory to associate with the channel. - validation_specification.validation_profiles[].training_job_definition.input_data_config[].data_source.file_system_data_source.file_system_access_mode
The access mode of the mount of the directory associated with the channel. A directory can be mounted either in ro (read-only) or rw (read-write) mode. - validation_specification.validation_profiles[].training_job_definition.input_data_config[].data_source.file_system_data_source.file_system_id
The file system id. - validation_specification.validation_profiles[].training_job_definition.input_data_config[].data_source.file_system_data_source.file_system_type
The file system type.
- validation_specification.validation_profiles[].training_job_definition.input_data_config[].data_source.s3_data_source
The S3 location of the data source that is associated with a channel. Show child fields- validation_specification.validation_profiles[].training_job_definition.input_data_config[].data_source.s3_data_source.attribute_names[]
- validation_specification.validation_profiles[].training_job_definition.input_data_config[].data_source.s3_data_source.instance_group_names[]
- validation_specification.validation_profiles[].training_job_definition.input_data_config[].data_source.s3_data_source.s3_data_distribution_type
If you want SageMaker to replicate the entire dataset on each ML compute instance that is launched for model training, specify FullyReplicated. If you want SageMaker to replicate a subset of data on each ML compute instance that is launched for model training, specify ShardedByS3Key. If there are n ML compute instances launched for a training job, each instance gets approximately 1/n of the number of S3 objects. In this case, model training on each machine uses only the subset of training data. Don't choose more ML compute instances for training than available S3 objects. If you do, some nodes won't get any data and you will pay for nodes that aren't getting any training data. This applies in both File and Pipe modes. Keep this in mind when developing algorithms. In distributed training, where you use multiple ML compute EC2 instances, you might choose ShardedByS3Key. If the algorithm requires copying training data to the ML storage volume (when TrainingInputMode is set to File), this copies 1/n of the number of objects. - validation_specification.validation_profiles[].training_job_definition.input_data_config[].data_source.s3_data_source.s3_data_type
If you choose S3Prefix, S3Uri identifies a key name prefix. SageMaker uses all objects that match the specified key name prefix for model training. If you choose ManifestFile, S3Uri identifies an object that is a manifest file containing a list of object keys that you want SageMaker to use for model training. If you choose AugmentedManifestFile, S3Uri identifies an object that is an augmented manifest file in JSON lines format. This file contains the data you want to use for model training. AugmentedManifestFile can only be used if the Channel's input mode is Pipe. - validation_specification.validation_profiles[].training_job_definition.input_data_config[].data_source.s3_data_source.s3_uri
Depending on the value specified for the S3DataType, identifies either a key name prefix or a manifest. For example: -
A key name prefix might look like this: s3://bucketname/exampleprefix/ -
A manifest might look like this: s3://bucketname/example.manifest A manifest is an S3 object which is a JSON file consisting of an array of elements. The first element is a prefix which is followed by one or more suffixes. SageMaker appends the suffix elements to the prefix to get a full set of S3Uri. Note that the prefix must be a valid non-empty S3Uri that precludes users from specifying a manifest whose individual S3Uri is sourced from different S3 buckets. The following code example shows a valid manifest format: [ {"prefix": "s3://customer_bucket/some/prefix/"}, "relative/path/to/custdata-1", "relative/path/custdata-2", ... "relative/path/custdata-N" ] This JSON is equivalent to the following S3Uri list: s3://customer_bucket/some/prefix/relative/path/to/custdata-1 s3://customer_bucket/some/prefix/relative/path/custdata-2 ... s3://customer_bucket/some/prefix/relative/path/custdata-N The complete set of S3Uri in this manifest is the input data for the channel for this data source. The object that each S3Uri points to must be readable by the IAM role that SageMaker uses to perform tasks on your behalf. Your input bucket must be located in same Amazon Web Services region as your training job.
- validation_specification.validation_profiles[].training_job_definition.input_data_config[].input_mode
(Optional) The input mode to use for the data channel in a training job. If you don't set a value for InputMode, SageMaker uses the value set for TrainingInputMode. Use this parameter to override the TrainingInputMode setting in a AlgorithmSpecification request when you have a channel that needs a different input mode from the training job's general setting. To download the data from Amazon Simple Storage Service (Amazon S3) to the provisioned ML storage volume, and mount the directory to a Docker volume, use File input mode. To stream data directly from Amazon S3 to the container, choose Pipe input mode. To use a model for incremental training, choose File input model. - validation_specification.validation_profiles[].training_job_definition.input_data_config[].record_wrapper_type
-
Specify RecordIO as the value when input data is in raw format but the training algorithm requires the RecordIO format. In this case, SageMaker wraps each individual S3 object in a RecordIO record. If the input data is already in RecordIO format, you don't need to set this attribute. For more information, see Create a Dataset Using RecordIO. In File mode, leave this field unset or set it to None.
- validation_specification.validation_profiles[].training_job_definition.input_data_config[].shuffle_config
A configuration for a shuffle option for input data in a channel. If you use S3Prefix for S3DataType, this shuffles the results of the S3 key prefix matches. If you use ManifestFile, the order of the S3 object references in the ManifestFile is shuffled. If you use AugmentedManifestFile, the order of the JSON lines in the AugmentedManifestFile is shuffled. The shuffling order is determined using the Seed value. For Pipe input mode, shuffling is done at the start of every epoch. With large datasets this ensures that the order of the training data is different for each epoch, it helps reduce bias and possible overfitting. In a multi-node training job when ShuffleConfig is combined with S3DataDistributionType of ShardedByS3Key, the data is shuffled across nodes so that the content sent to a particular node on the first epoch might be sent to a different node on the second epoch. Show child fields- validation_specification.validation_profiles[].training_job_definition.input_data_config[].shuffle_config.seed
Determines the shuffling order in ShuffleConfig value.
- validation_specification.validation_profiles[].training_job_definition.output_data_config
the path to the S3 bucket where you want to store model artifacts. SageMaker creates subfolders for the artifacts. Show child fields- validation_specification.validation_profiles[].training_job_definition.output_data_config.compression_type
The model output compression type. Select None to output an uncompressed model, recommended for large model outputs. Defaults to gzip. - validation_specification.validation_profiles[].training_job_definition.output_data_config.kms_key_id
The Amazon Web Services Key Management Service (Amazon Web Services KMS) key that SageMaker uses to encrypt the model artifacts at rest using Amazon S3 server-side encryption. The KmsKeyId can be any of the following formats: -
// KMS Key ID "1234abcd-12ab-34cd-56ef-1234567890ab" -
// Amazon Resource Name (ARN) of a KMS Key "arn:aws:kms:us-west-2:111122223333:key/1234abcd-12ab-34cd-56ef-1234567890ab" -
// KMS Key Alias "alias/ExampleAlias" -
// Amazon Resource Name (ARN) of a KMS Key Alias "arn:aws:kms:us-west-2:111122223333:alias/ExampleAlias" If you use a KMS key ID or an alias of your KMS key, the SageMaker execution role must include permissions to call kms:Encrypt. If you don't provide a KMS key ID, SageMaker uses the default KMS key for Amazon S3 for your role's account. For more information, see KMS-Managed Encryption Keys in the Amazon Simple Storage Service Developer Guide. If the output data is stored in Amazon S3 Express One Zone, it is encrypted with server-side encryption with Amazon S3 managed keys (SSE-S3). KMS key is not supported for Amazon S3 Express One Zone The KMS key policy must grant permission to the IAM role that you specify in your CreateTrainingJob, CreateTransformJob, or CreateHyperParameterTuningJob requests. For more information, see Using Key Policies in Amazon Web Services KMS in the Amazon Web Services Key Management Service Developer Guide. - validation_specification.validation_profiles[].training_job_definition.output_data_config.s3_output_path
Identifies the S3 path where you want SageMaker to store the model artifacts. For example, s3://bucket-name/key-name-prefix.
- validation_specification.validation_profiles[].training_job_definition.resource_config
The resources, including the ML compute instances and ML storage volumes, to use for model training. Show child fields- validation_specification.validation_profiles[].training_job_definition.resource_config.instance_count
The number of ML compute instances to use. For distributed training, provide a value greater than 1. - validation_specification.validation_profiles[].training_job_definition.resource_config.instance_groups[]
Show child fields- validation_specification.validation_profiles[].training_job_definition.resource_config.instance_groups[].instance_count
Specifies the number of instances of the instance group. - validation_specification.validation_profiles[].training_job_definition.resource_config.instance_groups[].instance_group_name
Specifies the name of the instance group. - validation_specification.validation_profiles[].training_job_definition.resource_config.instance_groups[].instance_type
Specifies the instance type of the instance group.
- validation_specification.validation_profiles[].training_job_definition.resource_config.instance_type
The ML compute instance type. SageMaker Training on Amazon Elastic Compute Cloud (EC2) P4de instances is in preview release starting December 9th, 2022. Amazon EC2 P4de instances (currently in preview) are powered by 8 NVIDIA A100 GPUs with 80GB high-performance HBM2e GPU memory, which accelerate the speed of training ML models that need to be trained on large datasets of high-resolution data. In this preview release, Amazon SageMaker supports ML training jobs on P4de instances (ml.p4de.24xlarge) to reduce model training time. The ml.p4de.24xlarge instances are available in the following Amazon Web Services Regions. To request quota limit increase and start using P4de instances, contact the SageMaker Training service team through your account team. - validation_specification.validation_profiles[].training_job_definition.resource_config.keep_alive_period_in_seconds
The duration of time in seconds to retain configured resources in a warm pool for subsequent training jobs. - validation_specification.validation_profiles[].training_job_definition.resource_config.volume_kms_key_id
The Amazon Web Services KMS key that SageMaker uses to encrypt data on the storage volume attached to the ML compute instance(s) that run the training job. Certain Nitro-based instances include local storage, dependent on the instance type. Local storage volumes are encrypted using a hardware module on the instance. You can't request a VolumeKmsKeyId when using an instance type with local storage. For a list of instance types that support local instance storage, see Instance Store Volumes. For more information about local instance storage encryption, see SSD Instance Store Volumes. The VolumeKmsKeyId can be in any of the following formats: -
// KMS Key ID "1234abcd-12ab-34cd-56ef-1234567890ab" -
// Amazon Resource Name (ARN) of a KMS Key "arn:aws:kms:us-west-2:111122223333:key/1234abcd-12ab-34cd-56ef-1234567890ab" - validation_specification.validation_profiles[].training_job_definition.resource_config.volume_size_in_gb
The size of the ML storage volume that you want to provision. ML storage volumes store model artifacts and incremental states. Training algorithms might also use the ML storage volume for scratch space. If you want to store the training data in the ML storage volume, choose File as the TrainingInputMode in the algorithm specification. When using an ML instance with NVMe SSD volumes, SageMaker doesn't provision Amazon EBS General Purpose SSD (gp2) storage. Available storage is fixed to the NVMe-type instance's storage capacity. SageMaker configures storage paths for training datasets, checkpoints, model artifacts, and outputs to use the entire capacity of the instance storage. For example, ML instance families with the NVMe-type instance storage include ml.p4d, ml.g4dn, and ml.g5. When using an ML instance with the EBS-only storage option and without instance storage, you must define the size of EBS volume through VolumeSizeInGB in the ResourceConfig API. For example, ML instance families that use EBS volumes include ml.c5 and ml.p2. To look up instance types and their instance storage types and volumes, see Amazon EC2 Instance Types. To find the default local paths defined by the SageMaker training platform, see Amazon SageMaker Training Storage Folders for Training Datasets, Checkpoints, Model Artifacts, and Outputs.
- validation_specification.validation_profiles[].training_job_definition.stopping_condition
Specifies a limit to how long a model training job can run. It also specifies how long a managed Spot training job has to complete. When the job reaches the time limit, SageMaker ends the training job. Use this API to cap model training costs. To stop a job, SageMaker sends the algorithm the SIGTERM signal, which delays job termination for 120 seconds. Algorithms can use this 120-second window to save the model artifacts. Show child fields- validation_specification.validation_profiles[].training_job_definition.stopping_condition.max_pending_time_in_seconds
The maximum length of time, in seconds, that a training or compilation job can be pending before it is stopped. - validation_specification.validation_profiles[].training_job_definition.stopping_condition.max_runtime_in_seconds
The maximum length of time, in seconds, that a training or compilation job can run before it is stopped. For compilation jobs, if the job does not complete during this time, a TimeOut error is generated. We recommend starting with 900 seconds and increasing as necessary based on your model. For all other jobs, if the job does not complete during this time, SageMaker ends the job. When RetryStrategy is specified in the job request, MaxRuntimeInSeconds specifies the maximum time for all of the attempts in total, not each individual attempt. The default value is 1 day. The maximum value is 28 days. The maximum time that a TrainingJob can run in total, including any time spent publishing metrics or archiving and uploading models after it has been stopped, is 30 days. - validation_specification.validation_profiles[].training_job_definition.stopping_condition.max_wait_time_in_seconds
The maximum length of time, in seconds, that a managed Spot training job has to complete. It is the amount of time spent waiting for Spot capacity plus the amount of time the job can run. It must be equal to or greater than MaxRuntimeInSeconds. If the job does not complete during this time, SageMaker ends the job. When RetryStrategy is specified in the job request, MaxWaitTimeInSeconds specifies the maximum time for all of the attempts in total, not each individual attempt.
- validation_specification.validation_profiles[].training_job_definition.training_input_mode
The training input mode that the algorithm supports. For more information about input modes, see Algorithms. Pipe mode If an algorithm supports Pipe mode, Amazon SageMaker streams data directly from Amazon S3 to the container. File mode If an algorithm supports File mode, SageMaker downloads the training data from S3 to the provisioned ML storage volume, and mounts the directory to the Docker volume for the training container. You must provision the ML storage volume with sufficient capacity to accommodate the data downloaded from S3. In addition to the training data, the ML storage volume also stores the output model. The algorithm container uses the ML storage volume to also store intermediate information, if any. For distributed algorithms, training data is distributed uniformly. Your training duration is predictable if the input data objects sizes are approximately the same. SageMaker does not split the files any further for model training. If the object sizes are skewed, training won't be optimal as the data distribution is also skewed when one host in a training cluster is overloaded, thus becoming a bottleneck in training. FastFile mode If an algorithm supports FastFile mode, SageMaker streams data directly from S3 to the container with no code changes, and provides file system access to the data. Users can author their training script to interact with these files as if they were stored on disk. FastFile mode works best when the data is read sequentially. Augmented manifest files aren't supported. The startup time is lower when there are fewer files in the S3 bucket provided.
- validation_specification.validation_profiles[].transform_job_definition
The TransformJobDefinition object that describes the transform job that SageMaker runs to validate your algorithm. Show child fields- validation_specification.validation_profiles[].transform_job_definition.batch_strategy
A string that determines the number of records included in a single mini-batch. SingleRecord means only one record is used per mini-batch. MultiRecord means a mini-batch is set to contain as many records that can fit within the MaxPayloadInMB limit. - validation_specification.validation_profiles[].transform_job_definition.environment
The environment variables to set in the Docker container. We support up to 16 key and values entries in the map. - validation_specification.validation_profiles[].transform_job_definition.max_concurrent_transforms
The maximum number of parallel requests that can be sent to each instance in a transform job. The default value is 1. - validation_specification.validation_profiles[].transform_job_definition.max_payload_in_mb
The maximum payload size allowed, in MB. A payload is the data portion of a record (without metadata). - validation_specification.validation_profiles[].transform_job_definition.transform_input
A description of the input source and the way the transform job consumes it. Show child fields- validation_specification.validation_profiles[].transform_job_definition.transform_input.compression_type
If your transform data is compressed, specify the compression type. Amazon SageMaker automatically decompresses the data for the transform job accordingly. The default value is None. - validation_specification.validation_profiles[].transform_job_definition.transform_input.content_type
The multipurpose internet mail extension (MIME) type of the data. Amazon SageMaker uses the MIME type with each http call to transfer data to the transform job. - validation_specification.validation_profiles[].transform_job_definition.transform_input.data_source
Describes the location of the channel data, which is, the S3 location of the input data that the model can consume. Show child fields- validation_specification.validation_profiles[].transform_job_definition.transform_input.data_source.s3_data_source
The S3 location of the data source that is associated with a channel. Show child fields- validation_specification.validation_profiles[].transform_job_definition.transform_input.data_source.s3_data_source.s3_data_type
If you choose S3Prefix, S3Uri identifies a key name prefix. Amazon SageMaker uses all objects with the specified key name prefix for batch transform. If you choose ManifestFile, S3Uri identifies an object that is a manifest file containing a list of object keys that you want Amazon SageMaker to use for batch transform. The following values are compatible: ManifestFile, S3Prefix The following value is not compatible: AugmentedManifestFile - validation_specification.validation_profiles[].transform_job_definition.transform_input.data_source.s3_data_source.s3_uri
Depending on the value specified for the S3DataType, identifies either a key name prefix or a manifest. For example: -
A key name prefix might look like this: s3://bucketname/exampleprefix/. -
A manifest might look like this: s3://bucketname/example.manifest The manifest is an S3 object which is a JSON file with the following format: [ {"prefix": "s3://customer_bucket/some/prefix/"}, "relative/path/to/custdata-1", "relative/path/custdata-2", ... "relative/path/custdata-N" ] The preceding JSON matches the following S3Uris: s3://customer_bucket/some/prefix/relative/path/to/custdata-1 s3://customer_bucket/some/prefix/relative/path/custdata-2 ... s3://customer_bucket/some/prefix/relative/path/custdata-N The complete set of S3Uris in this manifest constitutes the input data for the channel for this datasource. The object that each S3Uris points to must be readable by the IAM role that Amazon SageMaker uses to perform tasks on your behalf.
- validation_specification.validation_profiles[].transform_job_definition.transform_input.split_type
The method to use to split the transform job's data files into smaller batches. Splitting is necessary when the total size of each object is too large to fit in a single request. You can also use data splitting to improve performance by processing multiple concurrent mini-batches. The default value for SplitType is None, which indicates that input data files are not split, and request payloads contain the entire contents of an input object. Set the value of this parameter to Line to split records on a newline character boundary. SplitType also supports a number of record-oriented binary data formats. Currently, the supported record formats are: When splitting is enabled, the size of a mini-batch depends on the values of the BatchStrategy and MaxPayloadInMB parameters. When the value of BatchStrategy is MultiRecord, Amazon SageMaker sends the maximum number of records in each request, up to the MaxPayloadInMB limit. If the value of BatchStrategy is SingleRecord, Amazon SageMaker sends individual records in each request. Some data formats represent a record as a binary payload wrapped with extra padding bytes. When splitting is applied to a binary data format, padding is removed if the value of BatchStrategy is set to SingleRecord. Padding is not removed if the value of BatchStrategy is set to MultiRecord. For more information about RecordIO, see Create a Dataset Using RecordIO in the MXNet documentation. For more information about TFRecord, see Consuming TFRecord data in the TensorFlow documentation.
- validation_specification.validation_profiles[].transform_job_definition.transform_output
Identifies the Amazon S3 location where you want Amazon SageMaker to save the results from the transform job. Show child fields- validation_specification.validation_profiles[].transform_job_definition.transform_output.accept
The MIME type used to specify the output data. Amazon SageMaker uses the MIME type with each http call to transfer data from the transform job. - validation_specification.validation_profiles[].transform_job_definition.transform_output.assemble_with
Defines how to assemble the results of the transform job as a single S3 object. Choose a format that is most convenient to you. To concatenate the results in binary format, specify None. To add a newline character at the end of every transformed record, specify Line. - validation_specification.validation_profiles[].transform_job_definition.transform_output.kms_key_id
The Amazon Web Services Key Management Service (Amazon Web Services KMS) key that Amazon SageMaker uses to encrypt the model artifacts at rest using Amazon S3 server-side encryption. The KmsKeyId can be any of the following formats: -
Key ID: 1234abcd-12ab-34cd-56ef-1234567890ab -
Key ARN: arn:aws:kms:us-west-2:111122223333:key/1234abcd-12ab-34cd-56ef-1234567890ab -
Alias name: alias/ExampleAlias -
Alias name ARN: arn:aws:kms:us-west-2:111122223333:alias/ExampleAlias If you don't provide a KMS key ID, Amazon SageMaker uses the default KMS key for Amazon S3 for your role's account. For more information, see KMS-Managed Encryption Keys in the Amazon Simple Storage Service Developer Guide. The KMS key policy must grant permission to the IAM role that you specify in your CreateModel request. For more information, see Using Key Policies in Amazon Web Services KMS in the Amazon Web Services Key Management Service Developer Guide. - validation_specification.validation_profiles[].transform_job_definition.transform_output.s3_output_path
The Amazon S3 path where you want Amazon SageMaker to store the results of the transform job. For example, s3://bucket-name/key-name-prefix. For every S3 object used as input for the transform job, batch transform stores the transformed data with an .out suffix in a corresponding subfolder in the location in the output prefix. For example, for the input data stored at s3://bucket-name/input-name-prefix/dataset01/data.csv, batch transform stores the transformed data at s3://bucket-name/output-name-prefix/input-name-prefix/data.csv.out. Batch transform doesn't upload partially processed objects. For an input S3 object that contains multiple records, it creates an .out file only if the transform job succeeds on the entire file. When the input contains multiple S3 objects, the batch transform job processes the listed S3 objects and uploads only the output for successfully processed objects. If any object fails in the transform job batch transform marks the job as failed to prompt investigation.
- validation_specification.validation_profiles[].transform_job_definition.transform_resources
Identifies the ML compute instances for the transform job. Show child fields- validation_specification.validation_profiles[].transform_job_definition.transform_resources.instance_count
The number of ML compute instances to use in the transform job. The default value is 1, and the maximum is 100. For distributed transform jobs, specify a value greater than 1. - validation_specification.validation_profiles[].transform_job_definition.transform_resources.instance_type
The ML compute instance type for the transform job. If you are using built-in algorithms to transform moderately sized datasets, we recommend using ml.m4.xlarge or ml.m5.largeinstance types. - validation_specification.validation_profiles[].transform_job_definition.transform_resources.volume_kms_key_id
The Amazon Web Services Key Management Service (Amazon Web Services KMS) key that Amazon SageMaker uses to encrypt model data on the storage volume attached to the ML compute instance(s) that run the batch transform job. Certain Nitro-based instances include local storage, dependent on the instance type. Local storage volumes are encrypted using a hardware module on the instance. You can't request a VolumeKmsKeyId when using an instance type with local storage. For a list of instance types that support local instance storage, see Instance Store Volumes. For more information about local instance storage encryption, see SSD Instance Store Volumes. The VolumeKmsKeyId can be any of the following formats: -
Key ID: 1234abcd-12ab-34cd-56ef-1234567890ab -
Key ARN: arn:aws:kms:us-west-2:111122223333:key/1234abcd-12ab-34cd-56ef-1234567890ab -
Alias name: alias/ExampleAlias -
Alias name ARN: arn:aws:kms:us-west-2:111122223333:alias/ExampleAlias
- validation_specification.validation_role
The IAM roles that SageMaker uses to run the training jobs.
|