funding-extraction-lora-predictions-funders-award-ids

该数据集包含多个配置版本，每个版本均包含相同的字段结构，适用于资金信息提取或预测任务。主要字段包括：文献DOI（doi）、资金声明（funding_statement）、真实资助方（ground_truth_funders）、预测资助方（predicted_funders）以及原始输出（raw_output）。所有配置均包含测试集（test），样本量固定为322条，数据大小因配置不同而有所差异，范围从1,283,797字节到6,830,206字节不等。数据集可能用于评估不同模型在资金信息识别任务上的性能。

This dataset contains multiple configuration versions, each with an identical field structure, designed for funding information extraction or prediction tasks. The primary fields include: Document DOI (doi), funding_statement, ground_truth_funders, predicted_funders, and raw_output. All configurations include a test set, with a fixed sample size of 322 records. The data size varies across different configurations, ranging from 1,283,797 bytes to 6,830,206 bytes. This dataset can be utilized to evaluate the performance of various models on funding information recognition tasks.