TAUR-dev/evals__lora_lmfd__rewritten_verification_blocks_v3_to_4omini__train__3to8dig__samples

该数据集包含了文档ID、文档内容（答案、问题、问题提问和解决方案）、目标、参数、响应、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信息的结构化数据。数据集有训练集分割，共100个示例。

The dataset consists of structured data including document ID, document content (answer, problem, question, and solution), target, arguments, responses, filtered responses, document hash, prompt hash, target hash, exact match, and extracted answers. The dataset has a training split with a total of 100 examples.