fairops.mlops package

fairops.mlops.autolog module

fairops.mlops.models module

class fairops.mlops.models.LoggedMetric(key: str, value: float, step: int | None = None, timestamp: int | None = None, run_id: str | None = None, experiment_id: str | None = None)

Bases: object

to_dict()

Convert to dictionary for easy logging/exporting.

class fairops.mlops.models.LoggedMetrics

Bases: object

add_metric(metric: LoggedMetric)

Store a new metric.

aggregate(experiment_id: str, run_id: str, key: str, step: int, method='mean')

Aggregate multiple values at the same step for a specific experiment and run.

get_metrics(experiment_id: str | None = None, run_id: str | None = None, key: str | None = None, step: int | None = None)

Retrieve stored metrics for a given experiment_id, run_id, key, and/or step.

to_dataframe()

Convert logged metrics into a Pandas DataFrame for analysis.

to_dict()

Export logged metrics to a JSON file while preserving hierarchy.

class fairops.mlops.models.LoggedParam(key: str, value: float, run_id: str | None = None, experiment_id: str | None = None)

Bases: object

to_dict()

Convert to dictionary for easy logging/exporting.

class fairops.mlops.models.LoggedParams

Bases: object

add_param(param: LoggedParam)

Store a new parameter.

get_params(experiment_id: str | None = None, run_id: str | None = None, key: str | None = None)

Retrieve stored parameters for a given experiment_id, run_id, and/or key.

to_dataframe()

Convert logged parameters into a Pandas DataFrame for analysis.

to_dict()

Export logged parameters to a JSON file while preserving hierarchy.