TAUR-dev/evals__LORA_just_attempts__samples

该数据集包含文档ID、文档内容（包括答案、难度级别、问题、解决方案、学科、唯一标识符）、目标、参数（包括生成参数和采样参数）、响应和过滤后的响应、文档哈希、提示哈希、目标哈希、精确匹配和提取的答案等字段。数据集分为训练集，共有500个示例，总大小为1,807,677字节。

The dataset includes fields such as document ID, document content (including answer, difficulty level, problem, solution, subject, unique identifier), target, arguments (including generation parameters and sampling parameters), responses and filtered responses, document hash, prompt hash, target hash, exact match, and extracted answers. The dataset is split into a training set with a total of 500 examples and a size of 1,807,677 bytes.