Cornell-AGI/gsm8k_size_qwen2.5_3b_eval

该数据集包含了数据来源、提示、能力、奖励模型以及额外信息等字段，并且提供了多个响应和评估结果。提示字段中包含了内容和角色信息，奖励模型字段中包含了真实情况和风格信息，额外信息字段中包含了答案、索引、问题和分割信息。数据集仅包含训练集部分，共有7473个示例，数据大小为193816575字节。

The dataset includes fields such as data source, prompt, ability, reward model, and extra information, as well as multiple responses and evaluation results. The prompt field contains content and role information, the reward model field contains ground truth and style information, and the extra information field contains answer, index, question, and split information. The dataset only contains the training set with a total of 7473 examples and a data size of 193816575 bytes.