selfrew/llama3_8b_it_math_test_bon_with_rewards
收藏Hugging Face2024-12-09 更新2024-12-14 收录
下载链接：
https://hf-mirror.com/datasets/selfrew/llama3_8b_it_math_test_bon_with_rewards
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含多个特征，如索引、问题、真实推理链、真实答案、难度等级、类型、解决方案、自定义解决方案序列、代码序列、预测序列、报告序列（值为null）和奖励序列（布尔类型）。数据集分为一个训练集，包含1,780,000个样本，总大小为5,746,479,675字节，下载大小为2,647,749,779字节。
The dataset contains multiple features such as idx (index), question, gt_cot (ground truth reasoning chain), gt (ground truth answer), level (difficulty level), type, solution, my_solu (custom solution sequence), code (code sequence), pred (prediction sequence), report (report sequence, null), and rewards (reward sequence, boolean). The dataset is divided into one training set (train) with 1,780,000 samples, a total size of 5,746,479,675 bytes, and a download size of 2,647,749,779 bytes.
提供机构：
selfrew
