Dataset Viewer
The dataset could not be loaded because the splits use different data file formats, which is not supported. Read more about the splits configuration. Click for more details.
Couldn't infer the same data file format for all splits. Got {'train': ('audiofolder', {}), 'bad_subtitles': ('audiofolder', {}), 'whisper_clean_test': ('parquet', {}), 'whisper_clean': ('parquet', {})}
Error code:   FileFormatMismatchBetweenSplitsError

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

Youtube Azatutyun Speech

Armenian YouTube speech dataset collected from Azatutyun.

Splits

  • train: clean Armenian audio + Armenian captions, intended for supervised ASR fine-tuning.
  • bad_subtitles: audio + Armenian caption files where captions are known to be noisy or incorrect. Do not use this split as clean supervised ASR training data.

Columns

The metadata files include:

  • file_name: audio filename for AudioFolder.
  • video_id: YouTube video ID.
  • vtt_file: corresponding .hy.vtt subtitle file.
  • text: cleaned text extracted from .hy.vtt.
  • subtitle_language: hy.
  • subtitle_quality: clean or bad_subtitles.
  • source: source identifier.
Downloads last month
4,152