train-qwen2.5math1.5binstruct-texts

该数据集包含3000个训练样本，总大小为118389476字节。每个样本包含多个字段：dataset_index（整型）、question（字符串）、gold_answer（字符串）、prompt（字符串）、prompt_token_ids（整型列表）、completion（字符串）、completion_token_ids（整型列表）、steps（字符串列表）、step_token_spans（整型列表的列表）和step_continuations（字符串列表的列表）。数据集仅包含训练集，下载大小为113920095字节。

This dataset contains 3,000 training samples, with a total size of 118,389,476 bytes. Each sample includes multiple fields: dataset_index (integer type), question (string type), gold_answer (string type), prompt (string type), prompt_token_ids (list of integers), completion (string type), completion_token_ids (list of integers), steps (list of strings), step_token_spans (list of lists of integers), and step_continuations (list of lists of strings). This dataset only includes a training split, with a download size of 113,920,095 bytes.