stallone/glaive-code-assistant-v3

该数据集包含三个不同的配置：completion、conversation和source。completion配置用于文本补全任务，包含输入、输出和文档ID三个特征。conversation配置用于对话生成任务，包含文档ID和消息列表（消息内容和角色）两个特征。source配置用于问答系统任务，包含问题、答案和文档ID三个特征。每个配置的训练集大小和示例数量也被详细列出。

The dataset includes three different configurations: completion, conversation, and source. The completion configuration is used for text completion tasks and includes features such as input, output, and document ID. The conversation configuration is used for dialogue generation tasks and includes features such as document ID and a list of messages (message content and role). The source configuration is used for question-answering systems and includes features such as question, answer, and document ID. The size of the training set and the number of examples for each configuration are also detailed.