Hi, I'm running into an issue with TTS.
Flow:
1. User asks a query
2. LLM sends its response
3. Part 1 of the stream arrives: "Acknowledged, processing your request."
4. 15-second break - thinking and calling tools
5. Part 2 of the stream arrives: full response
The problem: TTS only plays(Agent Speaks) Part 1 after the 15-second delay