app.py

import gradio as gr
import timm
import torch
import torch.nn as nn
from torchvision import datasets, transforms
from PIL import Image

from torch.utils.mobile_optimizer import optimize_for_mobile

model = timm.create_model('vit_base_patch16_224', pretrained=True)
model.head = torch.nn.Linear(in_features=model.head.in_features, out_features=5)

path = "opt_model.pt"

model = model.jit.load(path)

model.eval()

def transform_image(img_sample):
    transform = transforms.Compose([
        transforms.Resize((224, 224)),  # Resize to 224x224
        transforms.ToTensor(),  # Convert PIL image to tensor
        transforms.ColorJitter(contrast=0.5),  # Contrast
        transforms.RandomAdjustSharpness(sharpness_factor=0.5),
        transforms.RandomSolarize(threshold=0.75),
        transforms.RandomAutocontrast(p=1),
    ])
    img = Image.open(img_sample)
    transformed_img = transform(img)
    return transformed_img

def predict(Image):
    model.eval()
    tranformed_img = transform_image(Image)
    img = torch.from_numpy(tranformed_img)

    with torch.no_grad():
        grade = torch.softmax(model(img.float()), dim=1)[0]
    category = ["None", "Mild", "Moderate", "Severe", "Proliferative"]
    output_dict = {}
    for cat, value in zip(category, grade):
        output_dict[cat] = value.item()
    return output_dict



image = gr.Image(shape=(224, 224), image_mode="RGB")
label = gr.Label(label="Grade")

demo = gr.Interface(
    fn=predict,
    inputs=image,
    outputs=label,
    examples=["examples/0.png", "examples/1.png", "examples/2.png", "examples/3.png", "examples/4.png"]
    )

if __name__ == "__main__":
    demo.launch(debug=True)