---
title: music_genre
canonical_url: "https://www.modelscope.cn/datasets/ccmusic-database/music_genre"
md_url: "https://www.modelscope.cn/datasets/ccmusic-database/music_genre.md"
repository: ccmusic-database/music_genre
chinese_name: "音乐流派数据集"
last_updated: 2026-08-05
license: CC-BY-NC-ND
storage_size: "17 GB"
domain:
  - audio
tasks:
  - audio-classification
size_scale:
  - 10k-1m
downloads: 13496
stars: 19
---

# music_genre

> music_genre - ccmusic-database 在 ModelScope 开源的数据集。原始数据集包含至少1700首不同流派音乐的音频（.mp3格式），每段音频时长约270~300秒。数据库共分为17个流派，每个流派对应一个标注文件，标注信息为流派分类标签，用于流派分类任务。主要的流派标签：古典（symphony, opera, solo, chamber）、非古典（pop, dance&house, indie, soul/r&b,…

ccmusic-database/music_genre 是 ModelScope 魔搭社区上的audio-classification数据集，涉及 audio 领域，存储大小 17 GB，采用 CC-BY-NC-ND 许可。

- **Repository**: ccmusic-database/music_genre
- **License**: CC-BY-NC-ND
- **Tasks**: audio-classification
- **Domain**: audio
- **Storage size**: 17 GB
- **Downloads**: 13496
- **Stars**: 19
- **Last updated**: 2026-08-05

Source: https://www.modelscope.cn/datasets/ccmusic-database/music_genre

---

# 简介 Dataset Card for Music Genre
原始数据集包含大约 1,700 首音乐作品，格式为 .mp3, 来源是网易云音乐。这些作品的长度从 270 到 300 秒不等。所有作品的采样率均为 22,050 赫兹。由于提供音频音乐的网站为下载的音乐提供了风格标签，因此没有涉及特定的标注人员。验证过程与下载过程同时进行。这些音乐作品共分为 16 种类型。

The raw dataset comprises approximately 1,700 musical pieces in .mp3 format, sourced from the NetEase music. The lengths of these pieces range from 270 to 300 seconds. All are sampled at the rate of 22,050 Hz. As the website providing the audio music includes style labels for the downloaded music, there are no specific annotators involved. Validation is achieved concurrently with the downloading process. They are categorized into a total of 16 genres.

## 数据结构 Dataset Structure
### 原始子集 Raw Subset Structure
<table>
    <tr>
        <th>audio (22050Hz)</th>
        <th>mel (spectrogram)</th>
        <th>fst_level_label (2-class)</th>
        <th>sec_level_label (9-class)</th>
        <th>thr_level_label (16-class)</th>
    </tr>
    <tr>
        <td>.mp3</td>
        <td>.jpg, 22050Hz</td>
        <td>1_Classic / 2_Non_classic</td>
        <td>3_Symphony / 4_Opera / 5_Solo / 6_Chamber / 7_Pop / 8_Dance_and_house / 9_Indie / 10_Soul_or_r_and_b / 11_Rock</td>
        <td>3_Symphony / 4_Opera / 5_Solo / 6_Chamber / 12_Pop_vocal_ballad / 13_Adult_contemporary / 14_Teen_pop / 15_Contemporary_dance_pop / 16_Dance_pop / 17_Classic_indie_pop / 18_Chamber_cabaret_and_art_pop / 10_Soul_or_r_and_b / 19_Adult_alternative_rock / 20_Uplifting_anthemic_rock / 21_Soft_rock / 22_Acoustic_pop</td>
    </tr>
</table>

### 评估(默认)子集 Eval Subset (default)
<table>
    <tr>
        <th>mel (spectrogram)</th>
        <th>cqt (spectrogram)</th>
        <th>chroma (spectrogram)</th>
        <th>fst_level_label (2-class)</th>
        <th>sec_level_label (9-class)</th>
        <th>thr_level_label (16-class)</th>
    </tr>
    <tr>
        <td>.jpg, 11.4s, 48000Hz</td>
        <td>.jpg, 11.4s, 48000Hz</td>
        <td>.jpg, 11.4s, 48000Hz</td>
        <td>1_Classic / 2_Non_classic</td>
        <td>3_Symphony / 4_Opera / 5_Solo / 6_Chamber / 7_Pop / 8_Dance_and_house / 9_Indie / 10_Soul_or_r_and_b / 11_Rock</td>
        <td>3_Symphony / 4_Opera / 5_Solo / 6_Chamber / 12_Pop_vocal_ballad / 13_Adult_contemporary / 14_Teen_pop / 15_Contemporary_dance_pop / 16_Dance_pop / 17_Classic_indie_pop / 18_Chamber_cabaret_and_art_pop / 10_Soul_or_r_and_b / 19_Adult_alternative_rock / 20_Uplifting_anthemic_rock / 21_Soft_rock / 22_Acoustic_pop</td>
    </tr>
</table>

![](https://www.modelscope.cn/datasets/ccmusic-database/music_genre/resolve/master/data/genre.png)

### 数据实例 Data Instances
.zip(.jpg, .wav)

### 数据域 Data Fields
```txt
1_Classic
    3_Symphony
    4_Opera
    5_Solo
    6_Chamber

2_Non_classic
    7_Pop
        12_Pop_vocal_ballad
        13_Adult_contemporary
        14_Teen_pop

    8_Dance_and_house
        15_Contemporary_dance_pop
        16_Dance_pop

    9_Indie
        17_Classic_indie_pop
        18_Chamber_cabaret_and_art_pop

    10_Soul_or_RnB

    11_Rock
        19_Adult_alternative_rock
        20_Uplifting_anthemic_rock
        21_Soft_rock
        22_Acoustic_pop
```

### 数据分割 Data Splits
|      Split       |   Raw |   Eval |
| :--------------: | ----: | -----: |
|      total       | 1,713 | 36,375 |
|   train (80%)    | 1,370 | 29,100 |
| validation (10%) |   171 |  3,637 |
|    test (10%)    |   172 |  3,638 |

## 快速使用 Usage
:modelscope-code[]{type="sdk"}

## 维护 Maintenance
```bash
GIT_LFS_SKIP_SMUDGE=1 
```

:modelscope-code[]{type="git"}

### 环境(仅用于数据处理代码) Requirements(For data processing codes)
```bash
cd data
echo y | conda create -n data python=3.x
conda activate data
pip install -r requirements.txt
```

### 数据处理脚本 Data processor
1. Open project with `VSCode`
2. Select file `./data/data.py`
3. Press `F5`

## 数据集描述 Dataset Description
### 数据集总结 Dataset Summary
For the pre-processed version, audio is cut into an 11.4-second segment, resulting in 36,375 files, which are then transformed into Mel, CQT and Chroma. In the end, the data entry has six columns: the first three columns represent the Mel, CQT, and Chroma spectrogram slices in .jpg format, respectively, while the last three columns contain the labels for the three levels. The first level comprises two categories, the second level consists of nine categories, and the third level encompasses 16 categories. The entire dataset is shuffled and split into training, validation, and test sets in a ratio of 8:1:1. This dataset can be used for genre classification.

### 支持任务 Supported Tasks and Leaderboards
Audio classification, image classification

### 语言 Languages
Multilingual

## 数据集创建 Dataset Creation
### 创建理由 Curation Rationale
Promoting the development of AI in the music industry

### 数据源 Source Data
#### 初始数据收集及正规化 Initial Data Collection and Normalization
Zhaorui Liu, Monan Zhou

#### 语言支持 Who are the source language producers?
Composers of the songs in the dataset

### 标注 Annotations
#### 标注步骤 Annotation process
Students collected about 1700 musical pieces (.mp3 format) with lengths of 270-300s divided into 17 genres in total.

#### 标注者 Who are the annotators?
Students from CCMUSIC

### 个人及敏感信息 Personal and Sensitive Information
Due to copyright issues with the original music, only spectrograms are provided in the dataset for fair use.

## 数据使用考量 Considerations for Using the Data
### 社会影响 Social Impact of Dataset
Promoting the development of AI in the music industry

### 偏好 Discussion of Biases
Most are English songs

### 其它限制 Other Known Limitations
Samples are not balanced enough

## 附加信息 Additional Information
### 策划者 Dataset Curators
Zijin Li

### 镜像 Mirror
<https://huggingface.co/datasets/ccmusic-database/music_genre>

### 评估 Evaluation
<https://www.modelscope.cn/models/ccmusic-database/music_genre>

### 引用 Cite
```bibtex
@dataset{zhaorui_liu_2021_5676893,
  author    = {Zhaorui Liu and Zijin Li},
  title     = {Music Data Sharing Platform for Computational Musicology Research (CCMUSIC DATASET)},
  month     = {nov},
  year      = {2021},
  publisher = {Zenodo},
  version   = {1.1},
  doi       = {10.5281/zenodo.5676893},
  url       = {https://doi.org/10.5281/zenodo.5676893}
}
```

### 贡献 Contributions
Provide a dataset for music genre classification

<div style="display:none">