main.py

from typing import Union
from fastapi import FastAPI
from pydantic import BaseModel
from llama_cpp import Llama


class InferenceRequest(BaseModel):
    input: Union[str, None] = None
    max_tokens: Union[int, None] = 0


app = FastAPI()

llm = Llama(model_path="./models/vicuna-7b-v1.5.Q4_K_M.gguf",
            verbose=False, n_ctx=4096)


@app.get("/")
async def root():
    return {"message": "Hello World"}


@app.post('/inference')
async def inference(request: InferenceRequest):
    input_text = request.input
    max_tokens = 256
    try:
        max_tokens = int(request.max_tokens)
    except:
        pass

    # process request
    try:
        result = llm(input_text, temperature=0.2,
                     top_k=5, max_tokens=max_tokens)
        return result
    except:
        pass

    # create response
    return {}