Different speaker's contents are combined by '\n' into one speaker content

[zshnb]

I test with a 30 minutes audio, found a content below

 {
    "startMs": 53420,
    "endMs": 92090,
    "text": "马友友先生您好，我们非常高兴能邀请到您。您今天早上过得怎么样？您喝咖啡了吗？还是您更喜欢喝茶？\n我什么都喝，无论是咖啡还是茶，这些我都是喝的。早上呢，我习惯性地会来杯咖啡，这可能就是一种习惯了。今天早上，我确实是起得特别早。早晨，嗯，确实是一个非常好的思考时机，因为你的思绪还没有被一天的琐事所占据，头脑会特别清醒。我不知道你喜欢怎么开始早晨，但对我来说，这确实是反思或深度思考的好时机。"
  },

the content before \n is spearker A say, after \n is speaker B say, but in result they are combined by \n, and all contents are put in speaker A content. how to solve it?