app.py

import torch
import gradio as gr
from torch import nn
from torch.nn import functional as F
import torchvision
from PIL import Image
from torchvision import transforms



transformer = transforms.Compose([
    transforms.ToPILImage(),
    transforms.Resize((224, 224)),
    transforms.RandomHorizontalFlip(),
    transforms.RandomRotation(degrees=10),
    transforms.ToTensor(),
    transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])
])
model_best=torch.jit.load('model_best1510_hugging.pt')
classes=['Drinking', 'Others', 'Smoking', 'Talking on Phone']

def predict(inp):
  inp=transformer(inp).unsqueeze(0)
  #inp = transforms.ToTensor()(inp).unsqueeze(0)
  
  with torch.no_grad():
    prediction =F.softmax(model_best(inp)[0], dim=0)
    confidences = {classes[i]: float(prediction[i]) for i in range(4)}    
  return confidences
  
#gr.Interface(predict,inputs=gr.inputs.Image(label="Input Image", source="webcam"),outputs='label').launch(debug='True')

gr.Interface(predict,inputs=gr.inputs.Image(label="Input Image"),outputs='label').launch(debug='True')