QED

该数据集名为QED，用于评估分子属性预测和条件生成。它不仅能够生成符合连续关注性质的全新化学物质，而且在这些物质中显示出高创新性评分。该数据集适用于分子语言建模中的并发属性预测和条件序列生成任务。

This dataset, named QED, is designed for evaluating molecular property prediction and conditional generation tasks. It can not only generate novel chemical substances that possess continuous desirable properties, but also demonstrate high novelty scores among these generated substances. This dataset is applicable to concurrent property prediction and conditional sequence generation tasks in the field of molecular language modeling.