selfcorrexp2/llama3_sft_first_wrong_prompt1_generation0
收藏Hugging Face2024-12-28 更新2025-04-26 收录
下载链接：
https://hf-mirror.com/datasets/selfcorrexp2/llama3_sft_first_wrong_prompt1_generation0
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含索引、提示文本、答案序列、是否为第一轮以及地面真实值等字段。它似乎用于某种问答或对话系统，其中prompt字段提供问题或对话的起始文本，answers字段包含一系列可能的回答，first_round字段表示是否为对话的第一轮，gt字段则是正确的答案或标注。训练集大小为137,886,100字节，包含10,000个示例。
The dataset includes fields such as index, prompt text, answer sequence, whether it is the first round, and ground truth. It seems to be used for some question-answering or dialogue system, where the prompt field provides the starting text of the question or conversation, the answers field contains a series of possible responses, the first_round field indicates whether it is the first round of the conversation, and the gt field is the correct answer or label. The training set is 137,886,100 bytes in size and contains 10,000 examples.
提供机构：
selfcorrexp2
