dvilasuero/midicaps_benchmark

这是一个包含音乐作品属性的数据集，其中包括位置、标题、类型、类型概率、情绪、情绪概率、关键字、时间签名、节奏、节奏描述、时长、时长描述、和弦摘要、和弦摘要出现次数、乐器摘要、乐器数量排序、所有和弦、和弦时间戳、测试集标记和紧凑序列等多个字段。数据集分为测试集和小型测试集两部分。

This is a dataset containing attributes of music pieces, including location, caption, genre, genre probability, mood, mood probability, key, time signature, tempo, tempo description, duration, duration description, chord summary, chord summary occurrence, instrument summary, instrument numbers sorted, all chords, all chords timestamps, test set flag, and condensed sequence. The dataset is split into a test set and a small test set.