audiobook_unchunked / README.md
instinct1912's picture
Update README.md
a41842e verified
metadata
dataset_name: TTS Audio Dataset
tags:
  - audio
  - speech
  - tts
language:
  - uz
license: other
task_categories:
  - text-to-speech
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
dataset_info:
  features:
    - name: transcription
      dtype: string
    - name: start
      dtype: float64
    - name: end
      dtype: float64
    - name: duration
      dtype: float64
    - name: word_count
      dtype: int64
    - name: source_file
      dtype: string
    - name: audio_quality
      struct:
        - name: rms_energy
          dtype: float64
        - name: silence_ratio
          dtype: float64
    - name: vad_refinement
      struct:
        - name: ctc_end
          dtype: float64
        - name: ctc_start
          dtype: float64
        - name: enabled
          dtype: bool
        - name: end_adjustment
          dtype: float64
        - name: search_windows
          sequence: float64
        - name: start_adjustment
          dtype: float64
        - name: vad_end
          dtype: float64
        - name: vad_start
          dtype: float64
    - name: speaker_validation
      struct:
        - name: has_overlap
          dtype: bool
        - name: is_valid
          dtype: bool
        - name: overlap_duration
          dtype: float64
        - name: rejection_reason
          dtype: 'null'
        - name: segments
          list:
            - name: end
              dtype: float64
            - name: speaker
              dtype: string
            - name: start
              dtype: float64
        - name: speaker_count
          dtype: int64
    - name: audio
      dtype: string
  splits:
    - name: train
      num_bytes: 472360010
      num_examples: 1292013
  download_size: 208630665
  dataset_size: 472360010

TTS Audio Dataset

Description

Audio dataset for STT and TTS. Each file is ~5 minutes long and may contain multiple utterances. Includes paired audio and transcriptions.


Data Format

  • Audio: FLAC, 24kHz, 16-bit PCM
  • Text: Transcriptions in plain text (UTF-8)

Source

Collected from publicly available materials.


Processing Notes

The audio is ~5 minutes long and sometimes noisy, so it requires denoising and must be chunked into smaller segments using an aligner before training.


Use Cases

Suitable for training speech-to-text (STT) and text-to-speech (TTS) models.