Enqueue multiple talk stream requests

Hi,

in a scenario of long talks (up to 100 words), to reduce latency, I'm trying to submit different talks stream where each one represent a single phrase of about 20/30 words.

I notice that, although i request new stream talks successfully, only the first one in generated.

It's possibile to request a new talk request before the streaming termination of the previous one?

Thanks in advance