Dataset Viewer
Auto-converted to Parquet Duplicate
audio
audioduration (s)
30
30
name
stringlengths
16
16
000/000002_0.ogg
000/000003_0.ogg
000/000005_0.ogg
000/000010_0.ogg
000/000134_0.ogg
000/000136_0.ogg
000/000139_0.ogg
000/000140_0.ogg
000/000141_0.ogg
000/000148_0.ogg
000/000181_0.ogg
000/000182_0.ogg
000/000190_0.ogg
000/000193_0.ogg
000/000194_0.ogg
000/000197_0.ogg
000/000198_0.ogg
000/000200_0.ogg
000/000203_0.ogg
000/000204_0.ogg
000/000207_0.ogg
000/000208_0.ogg
000/000209_0.ogg
000/000210_0.ogg
000/000211_0.ogg
000/000212_0.ogg
000/000213_0.ogg
000/000237_0.ogg
000/000247_0.ogg
000/000249_0.ogg
000/000255_0.ogg
000/000256_0.ogg
000/000257_0.ogg
000/000258_0.ogg
000/000341_0.ogg
000/000343_0.ogg
000/000352_0.ogg
000/000368_0.ogg
000/000369_0.ogg
000/000397_0.ogg
000/000398_0.ogg
000/000399_0.ogg
000/000400_0.ogg
000/000405_0.ogg
000/000414_0.ogg
000/000424_0.ogg
000/000425_0.ogg
000/000458_0.ogg
000/000459_0.ogg
000/000512_0.ogg
000/000534_0.ogg
000/000535_0.ogg
000/000536_0.ogg
000/000537_0.ogg
000/000540_0.ogg
000/000546_0.ogg
000/000550_0.ogg
000/000564_0.ogg
000/000573_0.ogg
000/000574_0.ogg
000/000583_0.ogg
000/000584_0.ogg
000/000585_0.ogg
000/000590_0.ogg
000/000591_0.ogg
000/000592_0.ogg
000/000593_0.ogg
000/000602_0.ogg
000/000603_0.ogg
000/000605_0.ogg
000/000608_0.ogg
000/000615_0.ogg
000/000619_0.ogg
000/000620_0.ogg
000/000621_0.ogg
000/000625_0.ogg
000/000629_0.ogg
000/000631_0.ogg
000/000633_0.ogg
000/000635_0.ogg
000/000664_0.ogg
000/000665_0.ogg
000/000666_0.ogg
000/000667_0.ogg
000/000668_0.ogg
000/000669_0.ogg
000/000672_0.ogg
000/000673_0.ogg
000/000674_0.ogg
000/000676_0.ogg
000/000677_0.ogg
000/000679_0.ogg
000/000680_0.ogg
000/000681_0.ogg
000/000684_0.ogg
000/000686_0.ogg
000/000690_0.ogg
000/000694_0.ogg
000/000695_0.ogg
000/000696_0.ogg
End of preview. Expand in Data Studio

Free Music Archive (FMA-medium)

This is a mirror of FMA-medium.

  • Sampling rate: 24 and 48 kHz
  • Channels: 1 and 2
  • Format: Opus
  • Duration: 208 hours, 24908 tracks
  • License:
    • Each track is distributed under the license chosen by the artist. See tracks.csv for details.
    • The metadata is distributed under CC BY 4.0.
  • Source: https://github.com/mdeff/fma
  • Paper: FMA: A Dataset For Music Analysis

Usage

import io

import soundfile as sf
from datasets import Features, Value, load_dataset

for item in load_dataset(
    "philgzl/fma-medium",
    streaming=True,
    features=Features({"audio": Value("binary"), "name": Value("string")}),
):
    print(item["name"])
    buffer = io.BytesIO(item["audio"])
    x, fs = sf.read(buffer)
    # do stuff...

Citation

@inproceedings{defferrard2017fma,
  title = {{FMA}: {A} dataset for music analysis},
  author = {Defferrard, Micha{\"e}l and Benzi, Kirell and Vandergheynst, Pierre and Bresson, Xavier},
  booktitle = {Proc. ISMIR},
  pages = {316--323},
  year = {2017},
}
Downloads last month
23

Paper for philgzl/fma-medium