edited_common_voice / README.md
lunarlist's picture
Update README.md (#1)
1996896
metadata
dataset_info:
  features:
    - name: audio_filepath
      dtype: audio
    - name: text
      dtype: string
    - name: duration
      dtype: float64
  splits:
    - name: train
      num_bytes: 6731304269.504
      num_examples: 36296
    - name: test
      num_bytes: 340059709.94
      num_examples: 1911
  download_size: 6985650459
  dataset_size: 7071363979.443999
license: mit
task_categories:
  - text-to-speech
language:
  - th

Dataset Card for "edited_common_voice"

More Information needed

This dataset is a Thai TTS dataset that use the voice from Common Voice dataset and modify the voice to not to sound like the original.

Medium: Text-To-Speech ภาษาไทยด้วย Tacotron2