请问大模型输出后，怎么tts呢？

[zhaojiangbing]

可以不累计返回吗？

[qbc2016]

You can use pre_print_hook to call TTS. As for accumulation, you can use a prefix to truncate and get incremental content. Please refer to the implementation of the print function for details.

[DavdGao]

@zhaojiangbing The pre_print hook function takes a bool argument last as input, indicating if the current message is the last accumulated chunk in a streaming message. You can use tts model when last is True, and ignore the other messages.