README.md

metadata

license: cc0-1.0
dataset_info:
  features:
    - name: sentence
      dtype: string
    - name: audio
      dtype:
        audio:
          sampling_rate: 16000
  splits:
    - name: train
      num_bytes: 4831507290.906
      num_examples: 127287
    - name: dev
      num_bytes: 1054053505.127
      num_examples: 26231
    - name: test
      num_bytes: 1065284217.135
      num_examples: 26335
    - name: test_cy
      num_bytes: 506350038
      num_examples: 13167
    - name: test_en
      num_bytes: 558933020.632
      num_examples: 13168
  download_size: 7925124771
  dataset_size: 8016128071.8
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: dev
        path: data/dev-*
      - split: test
        path: data/test-*
      - split: test_cy
        path: data/test_cy-*
      - split: test_en
        path: data/test_en-*

Dataset Card for Welsh and English Common Voice Corpus 16.1

Dataset Details

Dataset Description

This dataset consists of curated selection of 179,854 MP3 recordings with corresponding text files from the Welsh and English language Common Voice 16.1 release.

All recordings tagged as being with an English accent from the UK and Ireland are included. The dataset contains an equal balance 50/50 between Welsh and English recordinds.

Dataset Sources

• https://commonvoice.mozilla.org/cy/datasets

• https://commonvoice.mozilla.org/en/datasets

References

@inproceedings{commonvoice:2020,
  author = {Ardila, R. and Branson, M. and Davis, K. and Henretty, M. and Kohler, M. and Meyer, J. and Morais, R. and Saunders, L. and Tyers, F. M. and Weber, G.},
  title = {Common Voice: A Massively-Multilingual Speech Corpus},
  booktitle = {Proceedings of the 12th Conference on Language Resources and Evaluation (LREC 2020)},
  pages = {4211--4215},
  year = 2020
}

More Information

You agree to not attempt to determine the identity of speakers in the Common Voice dataset

Why request access? On its website, the Mozilla Foundation require an e-mail before downloading, in case they need to contact you in the future about changes to the dataset. An email provides us a point of contact for passing on messages.