README.md

metadata

dataset_info:
  features:
    - name: audio
      dtype: audio
    - name: sentence
      dtype: string
  splits:
    - name: train
      num_bytes: 470257191.64
      num_examples: 4060
  download_size: 459865552
  dataset_size: 470257191.64
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
license: cc-by-nc-nd-4.0
language:
  - yue

ASR-SCCantDuSC: A Scripted Chinese Cantonese (Canton) Daily-use Speech Corpus

This open-source dataset consists of 4.06 hours of transcribed Guangzhou Cantonese scripted speech focusing on daily use sentences, where 4,060 utterances contributed by ten speakers were contained. Source: https://magichub.com/datasets/guangzhou-cantonese-scripted-speech-corpus-daily-use-sentence/