trained_model

Creates, updates, deletes, gets or lists a trained_model resource.

Overview

Name	trained_model
Type	Resource
Id	azure.ai_text_analytics_authoring.trained_model

Fields

The following fields are returned by SELECT queries:

list_model_evaluation_results

Name	Datatype	Description
language	string	Represents the document language. This is BCP-47 representation of a language. For example, use "en" for English, "en-gb" for English (UK), "es" for Spanish etc. Required.
location	string	Represents the document path. Required.
projectKind	string	Represents the project kind. Required. Known values are: "CustomSingleLabelClassification", "CustomMultiLabelClassification", "CustomEntityRecognition", "CustomAbstractiveSummarization", "CustomHealthcare", and "CustomTextSentiment".

get_trained_model

Name	Datatype	Description
hasSnapshot	boolean	The flag to indicate if the trained model has a snapshot ready. Required.
label	string	The trained model label. Required.
lastTrainedDateTime	string (date-time)	The last trained date time of the model. Required.
lastTrainingDurationInSeconds	integer	The duration of the model's last training request in seconds. Required.
modelExpirationDate	string (date)	The model expiration date. Required.
modelId	string	The model ID. Required.
modelTrainingConfigVersion	string	The model training config version. Required.

Methods

The following methods are available for this resource:

Name	Accessible by	Required Params	Optional Params	Description
list_model_evaluation_results	select	trained_model_label, project_name, stringIndexType, endpoint	top, skip, maxpagesize	Gets the detailed results of the evaluation for a trained model. This includes the raw inference results for the data included in the evaluation process.
get_trained_model	select	trained_model_label, project_name, endpoint		Gets the details of a trained model.
delete_trained_model	delete	trained_model_label, project_name, endpoint		Deletes an existing trained model.
get_model_evaluation_summary	exec	trained_model_label, project_name, endpoint		Gets the evaluation summary of a trained model. The summary includes high level performance measurements of the model e.g., F1, Precision, Recall, etc.
evaluate_model	exec	trained_model_label, project_name, endpoint		Triggers evaluation operation on a trained model.
load_snapshot	exec	trained_model_label, project_name, endpoint		Long-running operation.

Parameters

Parameters can be passed in the WHERE clause of a query. Check the Methods section to see which parameters are required or optional for each operation.

Name	Datatype	Description
endpoint	string	The service endpoint host (no scheme), e.g. myaccount.table.cosmos.azure.com:443 - value of the client Endpoint parameter. (default: )
project_name	string
stringIndexType	string	Specifies the method used to interpret string offsets. For additional information see https://aka.ms/text-analytics-offsets _. "Utf16CodeUnit" Required.
trained_model_label	string	The trained model label. Required.
maxpagesize	integer
skip	integer	The number of result items to skip. Default value is None.
top	integer	The number of result items to return. Default value is None.

SELECT examples

list_model_evaluation_results

Gets the detailed results of the evaluation for a trained model. This includes the raw inference results for the data included in the evaluation process.

SELECT
language,
location,
projectKind
FROM azure.ai_text_analytics_authoring.trained_model
WHERE trained_model_label = '{{ trained_model_label }}' -- required
AND project_name = '{{ project_name }}' -- required
AND stringIndexType = '{{ stringIndexType }}' -- required
AND endpoint = '{{ endpoint }}' -- required
AND top = '{{ top }}'
AND skip = '{{ skip }}'
AND maxpagesize = '{{ maxpagesize }}'
;

get_trained_model

Gets the details of a trained model.

SELECT
hasSnapshot,
label,
lastTrainedDateTime,
lastTrainingDurationInSeconds,
modelExpirationDate,
modelId,
modelTrainingConfigVersion
FROM azure.ai_text_analytics_authoring.trained_model
WHERE trained_model_label = '{{ trained_model_label }}' -- required
AND project_name = '{{ project_name }}' -- required
AND endpoint = '{{ endpoint }}' -- required
;

DELETE examples

delete_trained_model

Deletes an existing trained model.

DELETE FROM azure.ai_text_analytics_authoring.trained_model
WHERE trained_model_label = '{{ trained_model_label }}' --required
AND project_name = '{{ project_name }}' --required
AND endpoint = '{{ endpoint }}' --required
;

Lifecycle Methods

get_model_evaluation_summary

Gets the evaluation summary of a trained model. The summary includes high level performance measurements of the model e.g., F1, Precision, Recall, etc.

EXEC azure.ai_text_analytics_authoring.trained_model.get_model_evaluation_summary 
@trained_model_label='{{ trained_model_label }}' --required, 
@project_name='{{ project_name }}' --required, 
@endpoint='{{ endpoint }}' --required
;

evaluate_model

Triggers evaluation operation on a trained model.

EXEC azure.ai_text_analytics_authoring.trained_model.evaluate_model 
@trained_model_label='{{ trained_model_label }}' --required, 
@project_name='{{ project_name }}' --required, 
@endpoint='{{ endpoint }}' --required 
@@json=
'{
"kind": "{{ kind }}", 
"trainingSplitPercentage": {{ trainingSplitPercentage }}, 
"testingSplitPercentage": {{ testingSplitPercentage }}
}'
;

load_snapshot

Long-running operation.

EXEC azure.ai_text_analytics_authoring.trained_model.load_snapshot 
@trained_model_label='{{ trained_model_label }}' --required, 
@project_name='{{ project_name }}' --required, 
@endpoint='{{ endpoint }}' --required
;