dockyard_rl.data.datasets.response_datasets.aime24

Module Contents

Classes

AIME2024Dataset

Simple wrapper around the AIME2024 dataset with train split.

API

class dockyard_rl.data.datasets.response_datasets.aime24.AIME2024Dataset(repeat: int = 16, **kwargs)

Bases: dockyard_rl.data.datasets.raw_dataset.RawDataset

Simple wrapper around the AIME2024 dataset with train split.

Args: repeat: Number of times to repeat the dataset, default is 16

Initialization

format_data(data: dict[str, Any]) dict[str, Any]