TAUR-dev/OTQ7BI_sft_openthoughts_100000_end200000_chunk_4

该数据集包含了会话信息、问题、来源、ID、提示信息、答案以及模型响应的标注信息。会话信息中包含发送者和消息内容，提示信息包含内容和角色。此外，还包含了标注的元数据，如API URL、后端信息、是否应用了聊天模板、生成参数（如最大令牌数、重复惩罚、温度等）、模型名称等。数据集分为训练集，共有80000个示例。

The dataset includes conversation information, questions, sources, IDs, prompt information, answers, and annotations for model responses. Conversation information contains the sender and message content, while prompt information includes content and role. Additionally, it includes metadata for annotations, such as API URL, backend information, whether a chat template was applied, generation parameters (such as max tokens, repetition penalty, temperature, etc.), model name, etc. The dataset is split into a training set with a total of 80,000 examples.