metadata
license: bsd
task_categories:
- automatic-speech-recognition
language:
- en
pretty_name: timit phoneme datas
size_categories:
- 1K<n<10K
This dataset contains the phonetic transcriptions of audios as well as English transcripts. Phonetic transcriptions are based on the g2p model. It can be used to train phoneme recognition model using wav2vec2.