We open sourced the code we use for our grammar assistant app backend. It is a small and hopefully easy to navigate server app.
It could serve as an inspiration for OpenAI token streaming and FastAPI.
The code isn't perfect since we're still validating the idea, and we welcome any feedback. But we also believe, that the minimal viable example has been reached.
Show HN: Token Streaming with OpenAI SDK and FastAPI | Heykuki News