app.py

import numpy as np
from PIL import Image
import torch
from fastapi import FastAPI, UploadFile, File
from fastapi.responses import JSONResponse
from pydantic import BaseModel

app = FastAPI()

# Load the pre-trained model from within the container
model_uri = "model.pth"
model = torch.load(model_uri, weights_only=False)

# Define input schema for JSON requests
class ImageInput(BaseModel):
    image_path: str


# Preprocess the image
def preprocess_image(image):
    image = image.convert('L')  # Convert to grayscale
    image = image.resize((28, 28))
    image = np.array(image) / 255.0  # Normalize to [0, 1]
    image = (image - 0.1307) / 0.3081  # Standardize
    image = torch.tensor(image).unsqueeze(0).float()  # Convert to tensor with batch dimension
    return image


# Root endpoint
@app.get("/")
def greet_json():
    return {"Hello": "World!"}


# Predict endpoint for JSON input
@app.post("/predict")
async def predict_image(file: UploadFile = File(...)):
    try:
        # Read and preprocess the uploaded image
        image = Image.open(file.file)
        image = preprocess_image(image)

        # Make prediction
        model.eval()
        with torch.no_grad():
            output = model(image)
            prediction = output.argmax(dim=1).item()

        return JSONResponse(content={"prediction": f"The digit is {prediction}"})
    except Exception as e:
        return JSONResponse(content={"error": str(e)}, status_code=500)