Input streaming with context

[xumx]

https://docs.cartesia.ai/api-reference/endpoints/stream-speech-websocket#input-streaming-with-contexts

This API can provider better control of the sentences when responses are longer duration. But it is slightly harder to implement for this project.

I'm trying to implement it Cartesia's SDK with https://github.com/cartesia-ai/cartesia-js?tab=readme-ov-file#tts-over-websocket

Adding a reference to cartesia ticket. https://github.com/cartesia-ai/cartesia-js/issues/3