MlflowHandle#

class scikitplot.mlflow.MlflowHandle(_mlflow_module, _tracking_uri, _registry_uri, _ui_url, _client, _artifacts, _models, server=None, version=None, experiment_name=None, default_run_name=None, default_run_tags=None)[source]#

A handle that proxies the upstream mlflow module while adding session context.

Attributes:

mlflow_modulemodule: Imported mlflow module for this session.
tracking_uristr: Resolved tracking URI used for this session.
registry_uristr or None: Optional registry URI used for this session.
ui_urlstr: Human-facing URL for opening the MLflow UI.
clientmlflow.tracking.MlflowClient: MLflow client bound to the session.
artifactsArtifactsFacade: Artifact helper facade.
modelsModelsFacade: Model helper facade.
serverSpawnedServer or None: Spawned server handle if start_server=True.
versionMlflowVersion or None: Parsed installed MLflow version, if available.
experiment_namestr or None: Active experiment name configured for this session (if any).
default_run_namestr or None: Default run name used by start_run wrapper.
default_run_tagsMapping[str, str] or None: Default tags applied by start_run wrapper.

Parameters:

_mlflow_module (Any)
_tracking_uri (str)
_registry_uri (str | None)
_ui_url (str)
_client (Any)
_artifacts (ArtifactsFacade)
_models (ModelsFacade)
server (SpawnedServer | None)
version (Any | None)
experiment_name (str | None)
default_run_name (str | None)
default_run_tags (Mapping[str, str] | None)

Notes

Attribute access falls back to the underlying mlflow module. This allows users to omit import mlflow while keeping identical API usage.

property artifacts: ArtifactsFacade#

Artifact helper facade.

Returns:

ArtifactsFacade: Facade providing artifact helpers (log/list/download, etc.).

property client: Any#

MLflow client bound to the session.

Returns:

Any: Instance compatible with mlflow.tracking.MlflowClient.

default_run_name: str | None = None#

default_run_tags: Mapping[str, str] | None = None#

experiment_name: str | None = None#

property mlflow_module: Any#: Imported mlflow module for this session.

property models: ModelsFacade#

Model helper facade.

Returns:

ModelsFacade: Facade providing model registry helpers.

property registry_uri: str | None#: Optional registry URI used for this session.

server: SpawnedServer | None = None#

start_run(*args, **kwargs)[source]#

Start an MLflow run and apply session defaults.

Parameters:

*argsAny: Positional args forwarded to mlflow.start_run.
**kwargsAny: Keyword args forwarded to mlflow.start_run.

Returns:

Iterator[Any]: Context manager yielding the active run object.

Raises:

Exception: Propagates underlying MLflow errors.

Parameters:

args (Any)
kwargs (Any)

Return type:

Iterator[Any]

Notes

Strict behavior: - If run_name is not provided and default_run_name is set, we pass it. - After run starts, if default_run_tags is set, apply them to the active run.

This wrapper does not modify MLflow global state beyond the active run tags.

property tracking_uri: str#: Resolved tracking URI used for this session.

property ui_url: str#: Human-facing URL for opening the MLflow UI.

version: Any | None = None#

MlflowHandle#

This Page