Speech data should have a wide coverage, with each speaker recording 1050 sentences of audio

Requirements of recording personnel: native speakers of the corresponding recording language; The age range of recording personnel is 18-60 years old; Gender balance: male 50%±10%, female 50%±10%; Pronunciation naturalness: Record the speech in a natural and smooth manner imitating normal dialogue. The recorded speech should be controlled within 2 to 8 seconds