---
license: cc0-1.0
dataset_info:
  features:
  - name: sentence
    dtype: string
  - name: audio
    dtype:
      audio:
        sampling_rate: 16000
  splits:
  - name: train
    num_bytes: 306395420
    num_examples: 7898
  - name: dev
    num_bytes: 212436000
    num_examples: 5352
  - name: test
    num_bytes: 210776070
    num_examples: 5354
  - name: train_s99
    num_bytes: 2353999659
    num_examples: 63643
  - name: dev_s99
    num_bytes: 491003647
    num_examples: 13115
  - name: test_s99
    num_bytes: 506350038
    num_examples: 13167
  download_size: 4022278461
  dataset_size: 4080960834
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: dev
    path: data/dev-*
  - split: test
    path: data/test-*
  - split: train_s99
    path: data/train_s99-*
  - split: dev_s99
    path: data/dev_s99-*
  - split: test_s99
    path: data/test_s99-*
task_categories:
- automatic-speech-recognition
language:
- cy
pretty_name: Welsh Common Voice Corpus 16.1
---
# Dataset Card for Welsh Common Voice Corpus 16.1

## Dataset Details

### Dataset Description

This dataset consists of all 114,139 MP3 recordings with corresponding text files from the Welsh language Common Voice 16.1 release. 
It contains a total of 155.12 hours of speech from 1,832 contributors, 122.22 hours of which (or 78.79%) has been verified manually
by volunteers.

[Mozilla's corpus creator](https://github.com/common-voice/CorporaCreator) command line tool was used, with the number of repeats 
allowed for any sentence set to 99 (`-s 99`), to build new splits that utilize all validated recordings.

 
### Dataset Sources

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