Interaction with Azure ML
- class seeq.addons.azureml.backend._aml_response_models.AmlModel(name, idd)[source]
Bases:
objectGets the serialized response from the Azure ML model management endpoint and deserializes it
- name
Name of the Azure ML model
- Type
str
- id
ID of the Azure ML model
- Type
str
- framework
Framework used for the Azure ML model
- Type
str
- frameworkVersion
Version of the framework used
- Type
str
- version
Azure ML model version
- Type
str
- input_ids
IDs of the Seeq signals that the Azure ML model takes as inputs. If provided, asset_path_ids and asset_input_names are disregarded.
- Type
list
- asset_path_ids
IDs of the Seeq asset tree that the Azure ML model takes as inputs. This is used in conjunction with asset_input_names to specify the signals within the asset tree that are to be used as inputs for the ML model
- Type
list
- asset_input_names
Names of the Seeq signals that are to be used as inputs for the ML model. These signals must be children of each of the asset trees specified in asset_path_ids
- Type
list
- sample_rate
The sampling rate required by the Azure ML model for the input signals.
- Type
str
- class seeq.addons.azureml.backend._aml_response_models.OnlineDeployment(name, idd, computeType)[source]
Bases:
objectGets the serialized response from the Azure ML deployments endpoint and deserializes it
- id
ID of the Azure ML deployment
- Type
str
- name
Name of the Azure ML deployment
- Type
str
- modelId
ID of the model associated with the Azure ML deployment
- Type
str
- traffic
Allowed traffic to this deployment
- Type
str
- model
Name of the model associated with the Azure ML deployment
- Type
str
- location
Location of the Azure ML deployment
- Type
str
- deserialize_aml_deployment_response(json)[source]
Deserializes the Azure ML response from the deployment endpoint
- static deserialize_aml_deployment_response(json, computeType)[source]
Gets a serialized response from a given online endpoint in Azure ML and deserializes each deployment associated with the endpoint.
- Parameters
json (dict) – Serialized response of the Azure ML deployment endpoint
computeType (str) – The type of compute: Managed, ACI, K8S
- Returns
ods – List of deployments associated with the endpoint
- Return type
list
- class seeq.addons.azureml.backend._aml_response_models.OnlineEndpoint(name, idd)[source]
Bases:
objectGets the serialized response from an Azure ML endpoint and deserializes it
- id
ID of the Azure ML endpoint
- Type
str
- name
Name of the Azure ML endpoint
- Type
str
- type
Type of the Azure ML endpoint
- Type
str
- description
Description of the Azure ML endpoint
- Type
str
- scoringUri
The identifier to access the model associated with the online endpoint
- Type
str
- authMode
Type of authentication used by the endpoint
- Type
str
- provisioningState
State of the endpoint
- Type
str
- createdBy
Name of the user that created the endpoint
- Type
str
- lastModifiedAt
Date of the last time the endpoint was modified
- Type
str
- tags
Tags associated with the endpoint. The only valid tag for integration with Seeq is {Seeq: true}. If this tag does not exist, the endpoint will be disregarded as a Seeq compatible endpoint.
- Type
str
- kind
The endpoint deployment target type (Managed, Container, AKS).
- Type
str
- location
The Azure region that this endpoint is hosted in.
- Type
str
- traffic
Percentage of traffic to each deployment contained within the endpoint.
- Type
str
- deployment
Deployments associated with the endpoint
- Type
str
- primaryKey
Primary access key for the endpoint
- Type
str
- secondaryKey
Secondary access key for the endpoint
- Type
str
- deserialize_unmanaged_endpoint_response(json)[source]
Deserializes the Azure ML response from the unmanaged online endpoints endpoint
- deserialize_managed_endpoint_response()[source]
Deserializes the Azure ML response from the managed online endpoints endpoint
- add_deployment(seeq.addons.azureml.backend.OnlineDeployment)[source]
Adds an OnlineDeployment to the endpoint only if the traffic split requirement is met.
- add_deployment(deployment)[source]
- Parameters
deployment (seeq.addons.azureml.backend.OnlineDeployment) – A deserialized OnlineDeployment object
- Returns
-
- Return type
None
- class seeq.addons.azureml.backend._aml_online_endpoint_service.AmlOnlineEndpointService(tenant_id, app_id, app_secret, subscription_id, resource_group, workspace_name)[source]
Bases:
objectProvides a service to connect to Azure ML Studio and get endpoints that are tagged with {Seeq: true} and their associated deployments and models.
- list_online_endpoints()[source]
Returns a list containing online endpoints tagged with {Seeq: true} in Azure ML Studio
- list_online_endpoints()[source]
Public method to get a list of endpoints tagged with {Seeq: true} in Azure ML Studio and attach the associated deployments and models in the endpoint to each OnlineEndpoint object.
- Returns
oes – List of OnlineEndpoint objects with deployments and models attached to each object
- Return type
list
- exception seeq.addons.azureml.utils._exceptions.AzureMLException(code=501, reason='Unknown', message='Internal Error')[source]
Bases:
ExceptionException raised for errors connecting to the Azure ML service.
- code
HTTP response status code
- Type
str, int
- reason
reason for the code
- Type
str
- message
explanation of the error
- Type
str