create_liander2024_benchmark_runner#

openstef_beam.benchmarking.benchmarks.create_liander2024_benchmark_runner(data_dir: Path | None = None, storage: BenchmarkStorage | None = None, callbacks: list[BenchmarkCallback] | None = None, target_provider: Liander2024TargetProvider | None = None, revision: str = 'dce7fe9bbae0d62288986fa97fa1ee7e9d3b7044') → BenchmarkPipeline[BenchmarkTarget, list[Liander2024Category]][source]#

Create benchmark pipeline for Liander2024 dataset.

Parameters:

  • data_dir (Path | None) – Dataset directory. Downloads from HuggingFace if None.

  • storage (BenchmarkStorage | None) – Storage backend for results.

  • callbacks (list[BenchmarkCallback] | None) – Callbacks to use during benchmarking.

  • target_provider (Liander2024TargetProvider | None) – Custom target provider. Creates default if None.

  • revision (str) – Specific revision of the dataset to use.

Returns:

Configured benchmark pipeline.

Return type:

BenchmarkPipeline[BenchmarkTarget, list[Liander2024Category]]

Example

>>> from pathlib import Path
>>> runner = create_liander2024_benchmark_runner(
...     data_dir=Path("./liander2024_dataset")
... )
Parameters:
Return type:

BenchmarkPipeline[BenchmarkTarget, list[Literal['mv_feeder', 'station_installation', 'transformer', 'solar_park', 'wind_park']]]