1231czx/qwen_qwq_3ep_iter1_gen
收藏Hugging Face2025-02-10 更新2025-02-15 收录
下载链接：
https://hf-mirror.com/datasets/1231czx/qwen_qwq_3ep_iter1_gen
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含了四个字段：奖励（rewards，浮点数序列），提示（prompt，字符串），真实标签（gt，字符串），以及响应（responses，字符串序列）。数据集的目的是为了训练模型，能够处理提示并根据真实标签来生成响应，同时评估模型的响应质量（奖励）。提供了训练集，包含20000个示例。
The dataset includes four fields: rewards (float64 sequence), prompt (string), ground truth (gt, string), and responses (string sequence). The dataset is intended for training models to handle prompts, generate responses based on the ground truth, and evaluate the quality of the models responses (rewards). A training set with 20,000 examples is provided.
提供机构：
1231czx
