This dataset was recorded in a quiet office environment, with the participation of 2,250 speakers, including 1,014 males and 1,236 females. All speakers involved in the recording were professionally selected to ensure standardized pronunciation and clear enunciation. The recorded texts cover daily spoken language, education, medical health, sports and entertainment, and social economy.