mytestdpo/llama3_gsm8k1_first_corr_processed_old_format
收藏Hugging Face2024-12-29 更新2025-04-26 收录
下载链接：
https://hf-mirror.com/datasets/mytestdpo/llama3_gsm8k1_first_corr_processed_old_format
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含多个字段，如索引(idx)、提示(prompt)、是否为第一轮(first_round)、真实标签(gt)、奖励(rewards)、我的解决方案(my_solu)、标志(flag)、轮次(turn)和对话(conversations)。对话字段又包括内容(content)和角色(role)。数据集分为训练集(train)，共有75477个示例，大小为445342399字节。数据集下载大小为133105176字节。
The dataset includes multiple fields such as index (idx), prompt, whether it is the first round (first_round), ground truth label (gt), rewards, my solution (my_solu), flag, turn, and conversation (conversations). The conversation field includes content and role. The dataset is split into a training set (train) with a total of 75477 examples and a size of 445342399 bytes. The download size of the dataset is 133105176 bytes.
提供机构：
mytestdpo
