app.py

import gradio as gr
import os
import torch
import torchvision
from modeling import EffNetPlantDiseaseClassification
from timeit import default_timer as timer
from typing import Dict, Tuple
# load the model from hugging face repo
model = EffNetPlantDiseaseClassification.from_pretrained("BrandonFors/effnetv2_s_plant_disease")

# load transforms for effnetv2-s from torchvision
effnetv2_s_weights = torchvision.models.EfficientNet_V2_S_Weights.DEFAULT
effnetv2_s_auto_transforms = effnetv2_s_weights.transforms()

# define class names
class_names = ['Apple___Apple_scab',
  'Apple___Black_rot',
  'Apple___Cedar_apple_rust',
  'Apple___healthy',
  'Blueberry___healthy',
  'Cherry_(including_sour)___Powdery_mildew',
  'Cherry_(including_sour)___healthy',
  'Corn_(maize)___Cercospora_leaf_spot Gray_leaf_spot',
  'Corn_(maize)___Common_rust_',
  'Corn_(maize)___Northern_Leaf_Blight',
  'Corn_(maize)___healthy',
  'Grape___Black_rot',
  'Grape___Esca_(Black_Measles)',
  'Grape___Leaf_blight_(Isariopsis_Leaf_Spot)',
  'Grape___healthy',
  'Orange___Haunglongbing_(Citrus_greening)',
  'Peach___Bacterial_spot',
  'Peach___healthy',
  'Pepper,_bell___Bacterial_spot',
  'Pepper,_bell___healthy',
  'Potato___Early_blight',
  'Potato___Late_blight',
  'Potato___healthy',
  'Raspberry___healthy',
  'Soybean___healthy',
  'Squash___Powdery_mildew',
  'Strawberry___Leaf_scorch',
  'Strawberry___healthy',
  'Tomato___Bacterial_spot',
  'Tomato___Early_blight',
  'Tomato___Late_blight',
  'Tomato___Leaf_Mold',
  'Tomato___Septoria_leaf_spot',
  'Tomato___Spider_mites Two-spotted_spider_mite',
  'Tomato___Target_Spot',
  'Tomato___Tomato_Yellow_Leaf_Curl_Virus',
  'Tomato___Tomato_mosaic_virus',
  'Tomato___healthy']
# create predict function
def predict(img):
  # start the timer to time the prediction
  start_time = timer()
  # Transform the input image for use
  img = effnetv2_s_auto_transforms(img).unsqueeze(0)
  # Put the model in eval mode to make prediction
  model.eval()
  with torch.inference_mode():
    pred_logits = model(img)["logits"]
    pred_probs = torch.softmax(pred_logits, dim=1)
  # create a prediction label and prediction probabilities dict
  pred_labels_and_probs = {class_names[i]: float(pred_probs[0][i]) for i in range(len(class_names))}

  # calc pred time
  pred_time = round(timer() - start_time, 4)
  # return the pred dict and time
  return pred_labels_and_probs, pred_time

# Gradio App

# Create title, description, and article
title = "EffNetV2_S_PlantDisease"
description = "EffNetV2-S trained on Plant Village Dataset"
article = "Personal project"

# Create example list
example_list = [["examples/" + example] for example in os.listdir("examples")]

# Create the Gradio Demo
demo = gr.Interface(fn=predict, # maps inputs to outputs
                    inputs=gr.Image(type="pil"),
                    outputs=[gr.Label(num_top_classes=3, label="Predictions"),
                             gr.Number(label="Prediction time (s)")],
                    examples=example_list,
                    title=title,
                    description=description,
                    article=article)
# launch the demo
demo.launch(debug=False)