README.md

metadata

configs:
  - config_name: clean
    version: 1.0.0
    data_files:
      - split: train
        path: data/train.clean-*
      - split: test
        path: data/test.clean-*
      - split: dev
        path: data/dev.clean-*
  - config_name: other
    version: 1.0.0
    data_files:
      - split: train
        path: data/train.other-*
      - split: test
        path: data/test.other-*
      - split: dev
        path: data/dev.other-*
  - config_name: all
    version: 1.0.0
    data_files:
      - split: train.clean
        path: data/train.clean-*
      - split: train.other
        path: data/train.other-*
      - split: dev.clean
        path: data/dev.clean-*
      - split: dev.other
        path: data/dev.other-*
      - split: test.clean
        path: data/test.clean-*
      - split: test.other
        path: data/test.other-*
dataset_info:
  - config_name: clean
    features:
      - name: speaker
        dtype: string
      - name: file
        dtype: string
      - name: duration
        dtype: float32
      - name: text
        dtype: string
      - name: text_no_preprocessing
        dtype: string
      - name: text_normalized
        dtype: string
      - name: audio
        dtype:
          audio:
            sampling_rate: 44100
    splits:
      - name: train
        num_bytes: 17023899243
        num_examples: 125989
      - name: dev
        num_bytes: 24204633
        num_examples: 150
      - name: test
        num_bytes: 52040552
        num_examples: 300
    download_size: 16271001158
    dataset_size: 17104553676
  - config_name: other
    features:
      - name: speaker
        dtype: string
      - name: file
        dtype: string
      - name: duration
        dtype: float32
      - name: text
        dtype: string
      - name: text_no_preprocessing
        dtype: string
      - name: text_normalized
        dtype: string
      - name: audio
        dtype:
          audio:
            sampling_rate: 44100
    splits:
      - name: train
        num_bytes: 26755286687
        num_examples: 196489
      - name: dev
        num_bytes: 65601521
        num_examples: 350
      - name: test
        num_bytes: 129348882
        num_examples: 700
    download_size: 25655017468
    dataset_size: 26957939607
  - config_name: all
    features:
      - name: speaker
        dtype: string
      - name: file
        dtype: string
      - name: duration
        dtype: float32
      - name: text
        dtype: string
      - name: text_no_preprocessing
        dtype: string
      - name: text_normalized
        dtype: string
      - name: audio
        dtype:
          audio:
            sampling_rate: 44100
    splits:
      - name: train.clean
        num_bytes: 17023899243
        num_examples: 125989
      - name: train.other
        num_bytes: 26755286687
        num_examples: 196489
      - name: dev.clean
        num_bytes: 24204633
        num_examples: 150
      - name: dev.other
        num_bytes: 65601521
        num_examples: 350
      - name: test.clean
        num_bytes: 52040552
        num_examples: 300
      - name: test.other
        num_bytes: 129348882
        num_examples: 700
    download_size: 7040649041
    dataset_size: 44050381518
pretty_name: HiFi TTS
description: >-
  Hi-Fi Multi-Speaker English TTS Dataset (Hi-Fi TTS) is based on LibriVox's
  public domain audio books and Gutenberg Project texts.
homepage: http://www.openslr.org/109
language:
  - en
license:
  - cc-by-4.0
citation: |
  @article{bakhturina2021hi,
    title={{Hi-Fi Multi-Speaker English TTS Dataset}},
    author={Bakhturina, Evelina and Lavrukhin, Vitaly and Ginsburg, Boris and Zhang, Yang},
    journal={arXiv preprint arXiv:2104.01497},
    year={2021}
  }
task_categories:
  - text-to-speech
  - text-to-audio

Dataset Card for HiFiTTS

Hi-Fi Multi-Speaker English TTS Dataset (Hi-Fi TTS) is based on LibriVox's public domain audio books and Gutenberg Project texts.