mc-ai/roleplay-feedback-dataset-v2

这个数据集包含了会话信息（对话内容和角色）、人类场景、AI场景、AI个性、框架、LLAMA反馈、时间戳、反馈以及模型信息。LLAMA反馈和反馈都详细记录了评估者的评价，包括做得好的地方、可以改进的地方、应该提出的问题以及总结。数据集分为训练集，并提供了相关的配置信息。The dataset includes conversation information (dialogue content and role), human scenarios, AI scenarios, AI personality, framework, LLAMA feedback, timestamp, feedback, and model information. LLAMA feedback and feedback both record the evaluators assessments in detail, including what was done well, what could be improved, questions that should be asked, and summaries. The dataset is split into a training set and provides related configuration information.