Datasets:

Yougen
/

pfann_testset

license: other
task_categories:
  - audio-classification
pretty_name: PFANN Audio Dataset
configs:
  - config_name: default
    data_files:
      - split: test_20211202_query
        path: data/test_20211202_query/audio/*.tar
      - split: test_20211202_seed
        path: data/test_20211202_seed/audio/*.tar
      - split: test_20211224_query
        path: data/test_20211224_query/audio/*.tar
      - split: test_20211224_seed
        path: data/test_20211224_seed/audio/*.tar
      - split: test_20220316_query
        path: data/test_20220316_query/audio/*.tar
      - split: test_20220316_seed
        path: data/test_20220316_seed/audio/*.tar
tags:
  - audio
  - music
  - audio-fingerprinting
  - pfann
  - webdataset

Yougen/pfann_testset

Pfann-style audio dataset, packed as WebDataset tar shards.

Inside each tar, every sample is a pair sharing a unique key:

<key>.<ext>     # raw audio bytes (ext == wav / mp3 / flac / ... as in source)
<key>.json      # {"audio_id":..., "subset":..., "rel_path":..., "duration":..., "sample_rate":..., "channels":...}

metadata.csv columns: key, shard, audio_id, subset, rel_path, duration, sample_rate, channels

Loading

from datasets import load_dataset

ds = load_dataset("Yougen/pfann_testset")
print(ds)
print(ds["test_20211202_query"][0])
# sample keys: 'wav'/'mp3'/... (decoded audio), 'json' (metadata), '__key__', '__url__'

For streaming (no full download needed):

ds = load_dataset("Yougen/pfann_testset", streaming=True)
for example in ds["test_20211202_query"]:
    print(example["__key__"], example["json"]["audio_id"])
    break

HuggingFace's webdataset builder will automatically pair <key>.<audio_ext> with <key>.json inside every tar and decode the audio.