External Monitoring module

Module with all classes and methods to manage the Monitored External Machine Learning (ML) models.

MLOpsExternalMonitoring

class mlops_codex.external_monitoring.MLOpsExternalMonitoring(group: str, ex_monitoring_hash: str, status: MonitoringStatus | None = MonitoringStatus.Unvalidated, login: str | None = None, password: str | None = None, url: str | None = None)

Bases: BaseMLOps

Class that handles an external monitoring object

host(wait: bool | None = False)

Host the new external monitoring.

Parameters:

wait (Optional[bool], optional) – If true, wait until the host is validated or invalidate.

Raises:

• AuthenticationError – Raised if there is an authentication issue.

• GroupError – Raised if there is an error related to the group.

• ServerError – Raised if the server encounters an issue.

• ExternalMonitoringError – Raised if there is an error specific to external monitoring.

Returns:

The external monitoring hash if the new external monitoring is successfully hosted.

Return type:

str

logs(start: str, end: str)

Get the logs of an external monitoring.

Parameters:

• start (str) – Start date to look for the records. The format must be dd-MM-yyyy.

• end (str) – End date to look for the records. The format must be dd-MM-yyyy.

upload_file(*, model_file: str | None = None, requirements_file: str | None = None, preprocess_file: str | None = None, preprocess_reference: str | None = None, shap_reference: str | None = None, python_version: str | None = '3.10')

Validate inputs before sending files.

Parameters:

• model_file (Optional[str], optional) – Path to your model.pkl file. Defaults to None.

• requirements_file (Optional[str], optional) – Path to your requirements.txt file. Defaults to None.

• preprocess_file (Optional[str], optional) – Path to your preprocessing file. Defaults to None.

• preprocess_reference (Optional[str], optional) – Preprocessing function entrypoint. Defaults to None.

• shap_reference (Optional[str], optional) – Shap function entrypoint. Defaults to None.

• python_version (Optional[str], optional) – Python version. Can be “3.8”, “3.9”, or “3.10”. Defaults to “3.10”.

Raises:

• InputError – Raised if there is an error with the input model_file.

• InputError – Raised if there is an error with the input requirements_file.

wait_ready()

Check the status of the external monitoring.

Returns:

The status of the external monitoring.

Return type:

str

MLOpsExternalMonitoringClient

class mlops_codex.external_monitoring.MLOpsExternalMonitoringClient(*, login: str | None = None, password: str | None = None, url: str | None = None)

Bases: BaseMLOpsClient

Class that handles MLOps External Monitoring Client

Parameters:

• login (str) – Login for authenticating with the client. You can also use the env variable MLOPS_USER to set this

• password (str) – Password for authenticating with the client. You can also use the env variable MLOPS_PASSWORD to set this

• url (str) – URL to MLOps Server. Default value is https://neomaril.datarisk.net/, use it to test your deployment first before changing to production. You can also use the env variable MLOPS_URL to set this

Raises:

• ServerError – Database produced an unexpected error.

• AuthenticationError – If user is not in the master group.

• CredentialError – If the Cloud Credential is Invalid

get_external_monitoring(external_monitoring_hash: str) → MLOpsExternalMonitoring

Return an external monitoring.

Parameters:

external_monitoring_hash (str) – External Monitoring Hash.

Raises:

ExternalMonitoringError – Raised if there is an error fetching the external monitoring.

Returns:

The requested MLOps external monitoring instance.

Return type:

MLOpsExternalMonitoring

list_hosted_external_monitorings() → None

List all hosted external monitoring

register_monitoring(**kwargs) → MLOpsExternalMonitoring | None

Register a MLOps External Monitoring.

Parameters:

• name (str) – External Monitoring name.

• group (str) – External Monitoring group. The group is the same used for the external training and datasource.

• training_execution_id (int) – Valid MLOps training execution id.

• period (str) – The frequency the monitoring will run. It can be one of the following: “Day”, “Week”, “Quarter”, “Month”, “Year”.

• input_cols (list) – List with input column names.

• output_cols (list) – List with output column names.

• datasource_name (str) – Valid MLOps datasource name.

• extraction_type (str) – Type of extraction. It can be one of the following: “Incremental”, “Full”.

• datasource_uri (str) – Valid datasource URI.

• column_name (Optional[str], optional) – Column name of the data column.

• reference_date (Optional[str], optional) – Reference extraction date.

• python_version (Optional[str], optional) – Python version used to run preprocessing scripts. It can be one of the following: “3.8”, “3.9”, “3.10”. Defaults to “3.10”.

Raises:

InputError – Raised if there is an error with the provided input.

Returns:

The newly registered MLOps external monitoring instance.

Return type:

MLOpsExternalMonitoring