Datasets:
Tasks:
Automatic Speech Recognition
Formats:
parquet
Size:
10K - 100K
ArXiv:
Tags:
phonetic transcription
License: