ForecastingWorkflowConfig#

class openstef_foundation_models.presets.ForecastingWorkflowConfig(**data: Any) → None[source]#

Bases: BaseConfig

Declarative configuration for a foundation-model forecasting workflow.

Selects a model family and the checkpoint that backs it, the requested quantiles and horizons, the target column, and the columns to keep before forecasting. Every kept non-target column is treated as a known covariate, so weather forecasts condition the prediction. Compute settings (execution providers, session options) live on the nested backend config.

Parameters:: data (Any)

model: Literal['chronos2']#: Foundation model family to use.

checkpoint: Annotated[LocalCheckpoint | HubCheckpoint, FieldInfo(annotation=NoneType, required=True, discriminator='kind')]#: Checkpoint (weights + metadata) to load and run. Defaults to the published OpenSTEF Chronos-2 dynamic ONNX checkpoint on the HuggingFace Hub. Pick a size and variant with `Chronos2.<SIZE>.checkpoint(…)`, or pass a LocalCheckpoint to run a file already on disk.

quantiles: list[Quantile]#: Quantile levels to predict.

horizons: list[LeadTime]#: Forecast horizons to predict.

target_column: str#: Name of the target column to forecast.

selected_features: FeatureSelection#: Columns to keep before forecasting. Defaults to all columns. Every kept non-target column is forwarded to the model as a known covariate. Must include the target column.

backend: OnnxBackendConfig#: Inference backend (compute) configuration: execution providers and session options.

chronos2_hyperparams: Chronos2HyperParams#: Hyperparameters for the Chronos-2 forecaster.

model_id: ModelIdentifier#: Unique identifier for the workflow model.

run_name: str | None#: Optional name for this workflow run.

experiment_tags: dict[str, str]#: Optional metadata tags for experiment tracking.

model_config: ClassVar[ConfigDict] = {'arbitrary_types_allowed': False, 'extra': 'ignore', 'protected_namespaces': ()}#: Configuration for the model, should be a dictionary conforming to [ConfigDict][pydantic.config.ConfigDict].