shoumenchougou/RWKV-7-Arithmetic

RWKV-7-Arithmetic-0.1B数据集用于训练和测试加减法运算模型，包含33000147条单轮加减法QA数据，支持中英文数字、全半角格式以及大小写字符的多种表示形式。数据集包含多种数据类型，如全角、中文数字、大写中文数字、英文数字等随机替换的加减法数据，以及强制使用标准化的英文连接符的加减法数据。此外，还包括解方程格式的数据。数据集还包含了噪声数据和乱码数据，用于测试模型在处理异常输入时的鲁棒性。

The RWKV-7-Arithmetic-0.1B dataset is used for training and testing arithmetic operation models, containing 33,000,147 single-round addition and subtraction QA data, supporting various representations of Chinese and English numerals, full-width and half-width formats, as well as uppercase and lowercase characters. The dataset includes multiple data types, such as randomly replaced addition and subtraction data with full-width, Chinese numerals, uppercase Chinese numerals, and English numerals, as well as addition and subtraction data with standardized English conjunctions. Additionally, it also includes data in the form of equations. The dataset also contains noise data and gibberish data to test the robustness of the model in handling abnormal inputs.