formospeech/

hakkaradio_news_clean

	TEST




	

Subset lang_group hours n_utts n_chars secs/utt chars/sec

Hakka_Hailu 客語_海陸 5.99 3,327 82,469 6.48 3.82

Hakka_Sixian 客語_四縣 5.56 1,510 88,013 13.26 4.40

11.55 4,837 170,482 8.60 4.10

	TRAIN




	

Subset lang_group hours n_utts n_chars secs/utt chars/sec

Hakka_Dapu 客語_大埔 19.06 10,860 262,715 6.32 3.83

Hakka_Hailu 客語_海陸 76.17 33,059 1,133,0568.29 4.13

Hakka_Raoping 客語_饒平 5.05 2,507 66,698 7.25 3.67

Hakka_Sixian 客語_四縣 114.72 41… See the full description on the dataset page: https://huggingface.co/datasets/formospeech/hakkaradio_news_clean.