edwixx/phonk-dataset

Phonk Dataset

A small curated dataset of phonk music audio samples with metadata, suitable for music information retrieval, genre classification, and audio analysis tasks.

Dataset Description

• Genre: Phonk music
• Number of samples: < 1K (11 audio files)
• Format: AudioFolder (WAV + metadata CSV)
• Size: ~575 MB
• Access: Public

Contents

The dataset contains:

• audio_001.wav through audio_011.wav — Individual phonk music audio clips in WAV format
• metadata.csv — Metadata for each audio sample (e.g., filename, duration, annotations)

Usage

Load with 🤗 Datasets:

from datasets import load_dataset

dataset = load_dataset("edwixx/phonk-dataset")
print(dataset)

Or download the raw files:

from huggingface_hub import snapshot_download
snapshot_download("edwixx/phonk-dataset")

Intended Uses

• Music genre classification (phonk vs. other genres)
• Music information retrieval (MIR) research
• Audio feature extraction and analysis
• Training / fine-tuning music tagging models

License

This dataset is released under the MIT License.