Voice generation not working for custom LLM option
# supportp
pascalophon
03/30/2024, 8:03 AMHi, when creating a pipeline with a custom LLM voice generation didn't work. To confirm this, I ran the server side Vercel example and created an example assistant from the custom_llm.json - same result.
When I start a conversation with the new assistant from the dashboard, there is no response generated.
n
nikhil
03/31/2024, 10:59 PMYou need to send response back as SSE
nikhil
03/31/2024, 10:59 PMYou can create a script that uses the OpenAI SDK to call your custom LLM with stream mode enabled
nikhil
03/31/2024, 10:59 PMIf that works, it'll work with us