Over maximum length,and sentence not complete

I train my data(about 400 records). The outcome consistently exceeds the Maximum length I set, and in the last sentence always not complete.
This problem confuses me for a long time, if someone can solve it, that will be great!!
I think maybe a set “stop” in my model can solve this problem, but I don’t know how to do it.

Hit “SUBMIT” again.ㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤㅤ

Are you using <|endoftext|> in your training data? If so, you should be able to use <|endoftext|> to stop the output correctly…