str
): Benchmark dataset name, e.g.: nq
str
): Path to the dataset fileDict[str, str]
): Field mapping rules, where key
represents the name used in the returned result, and value
represents the field name in the original datasetbool
): Whether to shuffle the dataset after loadingint
): Random seed to ensure reproducible shuffle
resultsint
): Number of dataset entries to load, default -1
loads allget_data
Dict[str, Any]
): Configuration parameter dictionaryDict[str, List[Any]]
): Data dictionary containing the mapped content