AI-DataTang Corpus#

Source: AI-DataTang Corpus
Language: Mandarin Chinese
Dialects: Standard Mandarin Chinese, Beijing Mandarin
Number of hours: 200.38
Number of utterances: 237,265
Number of speakers: 600
- Female speakers: 0
- Male speakers: 0
- Unknown speakers: 600
License: CC BY-NC-ND 4.0
Citation:

@misc{aidatatang_200zh,
	title = {A free {C}hinese {M}andarin speech corpus by {B}eijing {D}ata{T}ang {T}echnology {C}o., {L}td (www.datatang.com)},
	author = {Beijing DataTang Technology Co., Ltd},
	url = {www.datatang.com}}

Please, note that no corpora are hosted by MFA, please see the link above for accessing the data.
If you have comments or questions about using this corpus for MFA, you can check previous MFA model discussion posts or create a new one.

Pronunciation dictionaries

Acoustic models