mlfoundations-dev/r1_annotated_finqa

该数据集包含了问答对、推理过程、解决方案以及对话信息。具体字段包括答案(answer)、问题(problem)、索引(__index_level_0__)、推理过程(reasoning)、解决方案(deepseek_solution)、原始行索引(__original_row_idx)、推理追踪(final_reasoning_trace)和对活信息(conversations)。数据集被分割为训练集，包含5000个示例，总文件大小为97750233字节。

The dataset includes question-answer pairs, reasoning processes, solutions, and conversation information. Specific fields include answer, problem, index (__index_level_0__), reasoning process, solution (deepseek_solution), original row index (__original_row_idx__), reasoning trace (final_reasoning_trace), and conversation information (conversations). The dataset is split into a training set containing 5000 examples, with a total file size of 97750233 bytes.