sartifyllc/belebele-sw-gemma-evaluation

该数据集包含多个字段，包括问题编号、段落、问题、模型回答、标准化回答、正确答案、精确匹配和F1得分。数据集主要用于评估模型在回答问题时的性能，通过比较模型回答与正确答案的匹配程度（精确匹配和F1得分）来衡量模型的表现。数据集分为训练集，包含900个样本，总大小为635788字节。

This dataset includes multiple fields such as question number, passage, question, model answer, normalized response, correct answer, exact match, and F1 score. The dataset is primarily used to evaluate the performance of models in answering questions by comparing the models answers with the correct answers (exact match and F1 score) to measure the models performance. The dataset is divided into a training set, containing 900 samples, with a total size of 635788 bytes.