oci_ai_document_model

This resource provides the Model resource in Oracle Cloud Infrastructure Ai Document service. Api doc link for the resource: https://docs.oracle.com/iaas/api/#/en/document-understanding/latest/Model

Example terraform configs related to the resource : https://github.com/oracle/terraform-provider-oci/tree/master/examples/aiDocument

Create a new model.

Updates the model metadata only selected path parameter.

Example Usage

resource "oci_ai_document_model" "test_model" {
	#Required
	compartment_id = var.compartment_id
	model_id = var.model_model_id
	model_type = var.model_model_type
	project_id = oci_ai_document_project.test_project.id

	#Optional
	alias_name = var.model_alias_name
	component_models {

		#Optional
		model_id = oci_ai_document_model.test_model.id
	}
	defined_tags = var.model_defined_tags
	description = var.model_description
	display_name = var.model_display_name
	freeform_tags = var.model_freeform_tags
	inference_units = var.model_inference_units
	is_quick_mode = var.model_is_quick_mode
	language = var.model_language
	max_training_time_in_hours = var.model_max_training_time_in_hours
	model_sub_type {
		#Required
		model_sub_type = var.model_model_sub_type_model_sub_type
		model_type = var.model_model_sub_type_model_type
	}
	model_version = var.model_model_version
	operations {

		#Optional
		operation = var.model_operations_operation
		path = var.model_operations_path
		value = var.model_operations_value
	}
	testing_dataset {
		#Required
		dataset_type = var.model_testing_dataset_dataset_type

		#Optional
		bucket = var.model_testing_dataset_bucket
		dataset_id = oci_data_labeling_service_dataset.test_dataset.id
		namespace = var.model_testing_dataset_namespace
		object = var.model_testing_dataset_object
	}
	training_dataset {
		#Required
		dataset_type = var.model_training_dataset_dataset_type

		#Optional
		bucket = var.model_training_dataset_bucket
		dataset_id = oci_data_labeling_service_dataset.test_dataset.id
		namespace = var.model_training_dataset_namespace
		object = var.model_training_dataset_object
	}
	validation_dataset {
		#Required
		dataset_type = var.model_validation_dataset_dataset_type

		#Optional
		bucket = var.model_validation_dataset_bucket
		dataset_id = oci_data_labeling_service_dataset.test_dataset.id
		namespace = var.model_validation_dataset_namespace
		object = var.model_validation_dataset_object
	}
}

Argument Reference

The following arguments are supported:

• alias_name - (Optional) the alias name of the model.
• compartment_id - (Required) (Updatable) The compartment identifier.
• component_models - (Optional) The OCID list of active custom Key Value models that need to be composed.
  • model_id - (Optional) The OCID of active custom Key Value model that need to be composed.
• defined_tags - (Optional) (Updatable) Defined tags for this resource. Each key is predefined and scoped to a namespace. For example: {"foo-namespace": {"bar-key": "value"}}
• description - (Optional) (Updatable) An optional description of the model.
• display_name - (Optional) (Updatable) A human-friendly name for the model, which can be changed.
• freeform_tags - (Optional) (Updatable) A simple key-value pair that is applied without any predefined name, type, or scope. It exists for cross-compatibility only. For example: {"bar-key": "value"}
• inference_units - (Optional) (Updatable) Number of replicas required for this model.
• is_quick_mode - (Optional) Set to true when experimenting with a new model type or dataset, so the model training is quick, with a predefined low number of passes through the training data.
• language - (Optional) The document language for model training, abbreviated according to the BCP 47 syntax.
• max_training_time_in_hours - (Optional) The maximum model training time in hours, expressed as a decimal fraction.
• model_id - (Required)
• model_sub_type - (Optional) Applicable to only PRE_TRAINED_KEY_VALUE_EXTRACTION, PRE_TRAINED_DOCUMENT_ELEMENTS_EXTRACTION.
  • model_sub_type - (Required) The model sub type for PRE_TRAINED_KEY_VALUE_EXTRACTION The allowed values are:
    • RECEIPT
    • INVOICE
    • PASSPORT
    • DRIVER_LICENSE
    • HEALTH_INSURANCE_ID
  • model_type - (Required) Sub type model based on the model type. The allowed values are:
    • PRE_TRAINED_KEY_VALUE_EXTRACTION
    • PRE_TRAINED_DOCUMENT_ELEMENTS_EXTRACTION
• model_type - (Required) The type of the Document model.
• model_version - (Optional) The model version
• operations - (Optional) (Updatable)
  • operation - (Optional) (Updatable)
  • path - (Optional) (Updatable)
  • value - (Optional) (Updatable)
• project_id - (Required) The OCID of the project that contains the model.
• testing_dataset - (Optional) The base entity which is the input for creating and training a model.
  • bucket - (Required when dataset_type=OBJECT_STORAGE) The name of the Object Storage bucket that contains the input data file.
  • dataset_id - (Required when dataset_type=DATA_SCIENCE_LABELING) OCID of the Data Labeling dataset.
  • dataset_type - (Required) The dataset type, based on where it is stored.
  • namespace - (Required when dataset_type=OBJECT_STORAGE) The namespace name of the Object Storage bucket that contains the input data file.
  • object - (Required when dataset_type=OBJECT_STORAGE) The object name of the input data file.
• training_dataset - (Optional) The base entity which is the input for creating and training a model.
  • bucket - (Required when dataset_type=OBJECT_STORAGE) The name of the Object Storage bucket that contains the input data file.
  • dataset_id - (Required when dataset_type=DATA_SCIENCE_LABELING) OCID of the Data Labeling dataset.
  • dataset_type - (Required) The dataset type, based on where it is stored.
  • namespace - (Required when dataset_type=OBJECT_STORAGE) The namespace name of the Object Storage bucket that contains the input data file.
  • object - (Required when dataset_type=OBJECT_STORAGE) The object name of the input data file.
• validation_dataset - (Optional) The base entity which is the input for creating and training a model.
  • bucket - (Required when dataset_type=OBJECT_STORAGE) The name of the Object Storage bucket that contains the input data file.
  • dataset_id - (Required when dataset_type=DATA_SCIENCE_LABELING) OCID of the Data Labeling dataset.
  • dataset_type - (Required) The dataset type, based on where it is stored.
  • namespace - (Required when dataset_type=OBJECT_STORAGE) The namespace name of the Object Storage bucket that contains the input data file.
  • object - (Required when dataset_type=OBJECT_STORAGE) The object name of the input data file.

** IMPORTANT ** Any change to a property that does not support update will force the destruction and recreation of the resource with the new property values

Attributes Reference

The following attributes are exported:

• alias_name - the alias name of the model.
• compartment_id - The compartment identifier.
• component_models - The OCID collection of active custom Key Value models that need to be composed.
  • model_id - The OCID of active custom Key Value model that need to be composed.
• defined_tags - Defined tags for this resource. Each key is predefined and scoped to a namespace. For example: {"foo-namespace": {"bar-key": "value"}}
• description - An optional description of the model.
• display_name - A human-friendly name for the model, which can be changed.
• freeform_tags - A simple key-value pair that is applied without any predefined name, type, or scope. It exists for cross-compatibility only. For example: {"bar-key": "value"}
• id - A unique identifier that is immutable after creation.
• inference_units - Number of replicas required for this model.
• is_composed_model - Set to true when the model is created by using multiple key value extraction models.
• is_quick_mode - Set to true when experimenting with a new model type or dataset, so model training is quick, with a predefined low number of passes through the training data.
• labels - The collection of labels used to train the custom model.
• language - The document language for model training, abbreviated according to the BCP 47 syntax.
• lifecycle_details - A message describing the current state in more detail, that can provide actionable information if training failed.
• locks - Locks associated with this resource.
  • compartment_id - The lock compartment ID.
  • message - A message added by the lock creator. The message typically gives an indication of why the resource is locked.
  • related_resource_id - The resource ID that is locking this resource. Indicates that deleting this resource removes the lock.
  • time_created - Indicates when the lock was created, in the format defined by RFC 3339.
  • type - Lock type.
• max_training_time_in_hours - The maximum model training time in hours, expressed as a decimal fraction.
• metrics - Trained Model Metrics.
  • dataset_summary - Summary of count of samples used during model training.
    • test_sample_count - Number of samples used for testing the model.
    • training_sample_count - Number of samples used for training the model.
    • validation_sample_count - Number of samples used for validating the model.
  • label_metrics_report - List of metrics entries per label.
    • confidence_entries - List of document classification confidence report.
      • accuracy - accuracy under the threshold
      • f1score - f1Score under the threshold
      • precision - Precision under the threshold
      • recall - Recall under the threshold
      • threshold - Threshold used to calculate precision and recall.
    • document_count - Total test documents in the label.
    • label - Label name
    • mean_average_precision - Mean average precision under different thresholds
  • model_type - The type of custom model trained.
  • overall_metrics_report - Overall Metrics report for Document Classification Model.
    • confidence_entries - List of document classification confidence report.
      • accuracy - accuracy under the threshold
      • f1score - f1Score under the threshold
      • precision - Precision under the threshold
      • recall - Recall under the threshold
      • threshold - Threshold used to calculate precision and recall.
    • document_count - Total test documents in the label.
    • mean_average_precision - Mean average precision under different thresholds
• model_sub_type - Applicable to only PRE_TRAINED_KEY_VALUE_EXTRACTION, PRE_TRAINED_DOCUMENT_ELEMENTS_EXTRACTION.
  • model_sub_type - The model sub type for PRE_TRAINED_KEY_VALUE_EXTRACTION The allowed values are:
    • RECEIPT
    • INVOICE
    • PASSPORT
    • DRIVER_LICENSE
    • HEALTH_INSURANCE_ID
  • model_type - Sub type model based on the model type. The allowed values are:
    • PRE_TRAINED_KEY_VALUE_EXTRACTION
    • PRE_TRAINED_DOCUMENT_ELEMENTS_EXTRACTION
• model_type - The type of the Document model.
• model_version - The version of the model.
• project_id - The OCID of the project that contains the model.
• state - The current state of the model.
• system_tags - Usage of system tag keys. These predefined keys are scoped to namespaces. For example: {"orcl-cloud": {"free-tier-retained": "true"}}
• tenancy_id - The tenancy id of the model.
• testing_dataset - The base entity which is the input for creating and training a model.
  • bucket - The name of the Object Storage bucket that contains the input data file.
  • dataset_id - OCID of the Data Labeling dataset.
  • dataset_type - The dataset type, based on where it is stored.
  • namespace - The namespace name of the Object Storage bucket that contains the input data file.
  • object - The object name of the input data file.
• time_created - When the model was created, as an RFC3339 datetime string.
• time_updated - When the model was updated, as an RFC3339 datetime string.
• trained_time_in_hours - The total hours actually used for model training.
• training_dataset - The base entity which is the input for creating and training a model.
  • bucket - The name of the Object Storage bucket that contains the input data file.
  • dataset_id - OCID of the Data Labeling dataset.
  • dataset_type - The dataset type, based on where it is stored.
  • namespace - The namespace name of the Object Storage bucket that contains the input data file.
  • object - The object name of the input data file.
• validation_dataset - The base entity which is the input for creating and training a model.
  • bucket - The name of the Object Storage bucket that contains the input data file.
  • dataset_id - OCID of the Data Labeling dataset.
  • dataset_type - The dataset type, based on where it is stored.
  • namespace - The namespace name of the Object Storage bucket that contains the input data file.
  • object - The object name of the input data file.

Timeouts

The timeouts block allows you to specify timeouts for certain operations: * create - (Defaults to 20 minutes), when creating the Model * update - (Defaults to 20 minutes), when updating the Model * delete - (Defaults to 20 minutes), when destroying the Model

Import

Models can be imported using the id, e.g.

$ terraform import oci_ai_document_model.test_model "id"