yuhuanstudio/gsm8k_zhtw

GSM8K是一个包含8500个高品质小学数学文字题的数据集，旨在评估模型在多步骤数学推理任务中的表现。该数据集原始版本为英文，已被完整翻译为繁体中文。数据集包含问题和答案，每个问题需要2到8个步骤来解决，主要涉及基本的算术运算。该数据集可用于评估语言模型在繁体中文环境下的数学推理能力，作为繁体中文自然语言处理研究的基准数据集，以及开发和教育应用程式。

GSM8K is a dataset containing 8,500 high-quality elementary school math word problems, designed to evaluate the performance of models in multi-step mathematical reasoning tasks. The original dataset is in English and has been fully translated into Traditional Chinese. The dataset includes questions and answers, with each question requiring 2 to 8 steps to solve, mainly involving basic arithmetic operations. The dataset can be used to assess the mathematical reasoning ability of language models in a Traditional Chinese environment, as a benchmark dataset for Traditional Chinese natural language processing research, and for developing educational applications.