import os from fastapi import FastAPI, HTTPException from pydantic import BaseModel from langchain_ollama import ChatOllama from langchain.schema import StrOutputParser from langchain.prompts import ChatPromptTemplate from fastapi.responses import StreamingResponse import logging from functools import lru_cache # Set up logging logging.basicConfig(level=logging.INFO) logger = logging.getLogger(__name__) app = FastAPI() MODEL_NAME = 'phi3:mini' @lru_cache() def get_llm(): return ChatOllama(model=MODEL_NAME) @lru_cache() def get_chain(): llm = get_llm() prompt = ChatPromptTemplate.from_template("Question: {question}\n\nAnswer:") return prompt | llm | StrOutputParser() class Question(BaseModel): text: str @app.get("/") def read_root(): return {"Hello": f"Welcome to {MODEL_NAME} FastAPI"} @app.post("/ask") async def ask_question(question: Question): try: logger.info(f"Received question: {question.text}") chain = get_chain() response = chain.invoke({"question": question.text}) logger.info("Response generated successfully") return {"answer": response} except Exception as e: logger.error(f"Error in /ask endpoint: {str(e)}") raise HTTPException(status_code=500, detail=str(e)) @app.post("/ask_streaming") async def ask_question_stream(question: Question): try: logger.info(f"Received question for streaming: {question.text}") chain = get_chain() async def generate(): async for chunk in chain.astream({"question": question.text}): yield chunk + "\n" return StreamingResponse(generate(), media_type="text/plain") except Exception as e: logger.error(f"Error in /ask_stream endpoint: {str(e)}") raise HTTPException(status_code=500, detail=str(e)) @app.on_event("startup") async def startup_event(): logger.info(f"Starting up with model: {MODEL_NAME}") # Warm up the cache get_chain() @app.on_event("shutdown") async def shutdown_event(): logger.info("Shutting down")