app.py

### 1. Imports and class names setup ###

import gradio as gr
import os
import torch

from model import create_effnetb2_model
from timeit import default_timer as timer
from typing import Tuple, Dict

### 1. Open class names
with open("class_names.txt", "r") as f:
  class_names = [food_name.strip() for food_name in f.readlines()]

### 2. Model and transforms prep ###

effnetb2, effnetb2_transforms = create_effnetb2_model()

# Load saved weights
effnetb2.load_state_dict(
    torch.load(
        f="pretrained_effnetb2_foodvision.pth",
        map_location=torch.device("cpu")
    )
)

### 3. Predict function ###

def predict(img) -> Tuple[Dict, float]:
  # Start a timer
  start_time = timer()

  # Transform the input image for use with EffNetB2
  img = effnetb2_transforms(img).unsqueeze(0)

  # Put the model into eval mode, make prediction
  effnetb2.eval()
  with torch.inference_mode():
    pred_probs = torch.softmax(effnetb2(img), dim=1)

    pred_label = torch.argmax(pred_probs, dim=1)

  # Create a prediction label and prediction probability dictionary
  pred_labels_and_probs = {class_names[i]: float(pred_probs[0][i]) for i in range(len(class_names))}

  # Calculate pred time
  end_time = timer()
  pred_time = end_time - start_time

  # Return pred dict and pred time
  return pred_labels_and_probs, pred_time

### 4. Gradio app ###

example_list = [["examples/" + example] for example in os.listdir("examples")]
title = "Computer Vision for Food Processing"
description = "An [EfficientNetB2](https://pytorch.org/vision/main/models/generated/torchvision.models.efficientnet_b2.html?highlight=efficientnet_b2#torchvision.models.efficientnet_b2) feature extractor computer vision model to classify images of 101 foods (taken from PyTorch's [Food101](https://pytorch.org/vision/main/generated/torchvision.datasets.Food101.html) dataset). View the foods that this model can classify [here](https://github.com/alpapado/food-101/blob/master/data/meta/classes.txt)."

demo = gr.Interface(fn=predict,
                    inputs=gr.Image(type="pil"),
                    outputs=[gr.Label(num_top_classes=5, label="Predictions"),
                             gr.Number(label="Prediction time (s)")],
                    examples=example_list,
                    title=title,
                    description=description)
demo.launch()