inference_server.py

from sklearn.linear_model import SGDClassifier
import uvicorn
from fastapi import FastAPI

app = FastAPI()

def predict(input_text: str):
    data = [[ord(c) for c in input_text]] # Convert the string to a list of ASCII values
    model = train(data)
    # Make a prediction
    prediction = model.predict([[ord(c) for c in 'abc']]) # Convert the input string to a list of ASCII values
    return {"prediction": prediction}

def train(X):
    model = SGDClassifier()
    model.fit(X, X) # In this case, we are using the input data as the labels
    return model

# Here you can do things such as load your models

@app.get("/")
def read_root(input_text):
    return {f"Hello {input_text}!"}