mytestdpo/llama3_8b_it_gsm8k1_gold_tmp10_with_orm_rewards
收藏Hugging Face2025-01-06 更新2025-04-26 收录
下载链接：
https://hf-mirror.com/datasets/mytestdpo/llama3_8b_it_gsm8k1_gold_tmp10_with_orm_rewards
下载链接
链接失效反馈官方服务：
资源简介：
这是一个包含多个字段的数据集，其中包括索引、真实标签、提示文本、答案文本、用户解决方案、预测结果、消息列表（包括内容和角色）、奖励值、自定义提示和代理奖励。数据集分为训练集，其大小为14,281,002字节，共有2,638个示例。
This dataset contains multiple fields including index, ground truth labels, prompt text, answer text, user solutions, prediction results, a list of messages (including content and role), reward values, custom prompts, and proxy rewards. The dataset is split into a training set, which is 14,281,002 bytes in size and contains 2,638 examples.
提供机构：
mytestdpo
