metadata

license:
  - cc-by-4.0
task_categories:
  - text-to-speech
language:
  - en
pretty_name: ame
size_categories:
  - 10M<n<100M
configs:
  - config_name: default
    data_files:
      - split: dev
        path: data/dev-*
      - split: test
        path: data/test-*
      - split: train
        path: data/train-*

Aligned English MLS

We use the Montreal Forced Aligner to align English audio files to transcripts. The dataset contains a total of 44.5K hours of English audio speech.

We only provide the aligned TextGrid files. To download the audio files, you can use the Parler MLS English dataset available at parler-tts/mls_eng.

This dataset can be used to train TTS models by leveraging the ground truth phoneme durations to align phoneme features with acoustic features.

Download the Dataset

To download the dataset, use the following command:

huggingface-cli download ntt123/aligned_mls_eng --repo-type dataset

Read TextGrid Files

To read the TextGrid files, you can use the textgrid library. First, install the library:

pip install textgrid

Then, load and process a TextGrid file with the following example code:

import textgrid

# Load the TextGrid file
tg = textgrid.TextGrid.fromFile('test.TextGrid')