main.py

import requests

from . import app


API_URL = "https://api-inference.huggingface.co/models/google/gemma-1.1-7b-it"


def query(payload):
	response = requests.post(API_URL, json=payload)
	return response.json()

@app.get("/", tags=["Home"])
def api_home():
    return {'detail': 'Welcome to FastAPI TextGen Tutorial!'}

@app.get("/api/generate", summary="Generate text from prompt", tags=["Generate"])
def inference(input: str):
    return query({"inputs": input})