README.md

metadata

license: cc-by-4.0
dataset_info:
  config_name: toy_corpus_asr_ca
  features:
    - name: audio_id
      dtype: string
    - name: audio
      dtype:
        audio:
          sampling_rate: 16000
    - name: language
      dtype: string
    - name: normalized_text
      dtype: string
    - name: absolute_path
      dtype: string
  splits:
    - name: train
      num_bytes: 871036
      num_examples: 20
    - name: validation
      num_bytes: 370873
      num_examples: 10
    - name: test
      num_bytes: 416706
      num_examples: 10
  download_size: 1639360
  dataset_size: 1658615
configs:
  - config_name: toy_corpus_asr_ca
    data_files:
      - split: train
        path: toy_corpus_asr_ca/train-*
      - split: validation
        path: toy_corpus_asr_ca/validation-*
      - split: test
        path: toy_corpus_asr_ca/test-*
    default: true
task_categories:
  - automatic-speech-recognition
language:
  - ca
tags:
  - asr
  - catalan
  - speech
pretty_name: Toy Corpus for ASR in Catalan
size_categories:
  - n<1K

This is an example of a repository with a standard data loader. The audio files are compressed in tar format. Since this repository contains very few audio files, it can be used to test certain scripts in local machines.