Hello,
It seems that the auto-splitting logic might not be fully compatible with Chinese-specific punctuation, which could lead to the duplication you’re experiencing.
You are welcome to create a specific parsing rule for Chinese to change the way the system recognizes the file: