dapo-math-17k-qwen3
收藏资源简介:
DAPO-Math-17k English Qwen3 Answers 数据集是从 'open-r1/DAPO-Math-17k-Processed' 数据集的英文子集派生而来。该数据集新增了由不同规模的 Qwen3 模型(1.7B、4B、8B、14B 和 32B 参数)生成的答案列。数据集包含以下字段:'id'(唯一标识符)、'problem'(数学问题)、'prompt_text'(提示文本)、'expected_answer'(期望答案)、'answer_style'(答案风格)、'dataset_config'(数据集配置),以及由 Qwen3 模型生成的答案列(如 'qwen3_1b7_answer' 等)。数据集仅包含训练集,共有 14,116 个样本,总大小为 1,750,337,233 字节,下载大小为 700,286,785 字节。该数据集适用于数学问题解答、模型生成答案评估等任务。
The DAPO-Math-17k English Qwen3 Answers dataset is derived from the English subset of the "open-r1/DAPO-Math-17k-Processed" dataset. This dataset adds answer columns generated by Qwen3 models with different parameter scales (1.7B, 4B, 8B, 14B, and 32B parameters). It contains the following fields: "id" (unique identifier), "problem" (mathematical problem), "prompt_text" (prompt text), "expected_answer" (expected answer), "answer_style" (answer style), "dataset_config" (dataset configuration), as well as answer columns generated by Qwen3 models such as "qwen3_1b7_answer". The dataset only includes a training set, with a total of 14,116 samples, a total size of 1,750,337,233 bytes, and a download size of 700,286,785 bytes. This dataset is suitable for tasks including mathematical problem solving and evaluation of model-generated answers.




