Librispeech

This config can be used to prepare Librispeech dataset in the NeMo format.

It produces manifests for the dev-clean split (for other splits, please configure). The options are:

• "dev-clean"

• "dev-other"

• "test-clean"

• "test-other"

• "train-clean-100"

• "train-clean-360"

• "train-other-500"

• "dev-clean-2"

• "train-clean-5"

This config performs the following data processing.

1. Downloads Librispeech data

2. Converts flac files to wav file

3. Calculates the length of wav files

4. Makes capitalization lowercase

Required arguments.

• workspace_dir: specify the workspace folder where all audio files will be stored.

Note that you can customize any part of this config either directly or from command-line.

Output format.

This config generates output manifest file:

• ${workspace_dir}/manifest.json - dev-clean subset of the data.

Output manifest contains the following fields:

• audio_filepath (str): relative path to the audio files.

• text (str): transcription (lower-case without punctuation).

• duration (float): audio duration in seconds.

Config link: dataset_configs/english/librispeech/config.yaml