shisa-ai/wip-shisa-v1-all-scores-full

该数据集是一个包含多个自然语言处理模型评分的训练集，每个示例都包含了id、不同模型的评分、提示文本、模型来源、随机化顺序、响应内容和提示名称等信息。评分包括帮助性、诚实度、指令遵循性和真实性四个方面。

This dataset is a training set containing ratings from multiple natural language processing models, with each example including id, ratings from different models, prompt text, model sources, randomized order, response content, and prompt name. The ratings cover four aspects: helpfulness, honesty, instruction following, and truthfulness.