FastAPI deployment about uptrain

[HasnainKhanNiazi]

@ashish-1600 Thanks, that was the issue indeed.

[ashish-1600]

Hi @HasnainKhanNiazi ,
Thanks for raising this issue.

I have figured out a way to get this working in your use-case.
I am using uvicorn as my ASGI web server framework. In uvicorn, we have to set the event loop as 'asyncio'.

I am attaching the code for your reference.

@app.get("/")
async def read_root():
    return {"message": "Welcome to the LLM Evaluation API"}

@app.post("/evaluate/hallucination", response_model=EvaluationResponse)
async def hallucination_detect():
    try:
        data = [
            {
                "question": "Hello",
                "context": "Hello",
                "response": "Hello",
            }
        ]
        result = eval_llm.evaluate(
            data=data,
            checks=[Evals.FACTUAL_ACCURACY],
        )
        print(result)
        return EvaluationResponse(result=result)
    except Exception as e:
        return EvaluationResponse(result={}, error=str(e))
    

if __name__ == "__main__":
    uvicorn.run("app:app", host="0.0.0.0", port=4300, loop='asyncio')

You don't have to make any changes in the EvalLLM. Let me know if it works for you. :)