import gradio as gr
import os
import torch
from model import create_effnet_b2_model
from timeit import default_timer as timer
from typing import Dict, Tuple

class_names = ['pizza', 'steak', 'sushi']

effnetb2, effnetb2_transforms = create_effnet_b2_model(
    num_classes=3)

#load weigths
effnetb2.load_state_dict(
    torch.load(
        f='09_pretrained_effnet_b2_feature_extractor_20%.pth',
        map_location=torch.device('cpu')
    )
)


#predict
def predict(img) -> Tuple[Dict, float]:
    #start a timer
    start_time = timer()
    #transform input image
    img = effnetb2_transforms(img).unsqueeze(0)
    #set model to eval mode
    effnetb2.eval()
    with torch.inference_mode():
        pred_probs = torch.softmax(effnetb2(img), dim=1)
    pred_labels_and_probs = {class_names[i] :float(pred_probs[0,i]) for i in \
                             range(len(class_names))}
    end_time = timer()
    pred_time = round(end_time - start_time, 4)

    return pred_labels_and_probs, pred_time


examples_list = [['examples/' + example] for example in os.listdir('examples')]
#examples_list



title = 'foodvision mini'
description = 'effnet feature extractor for image classification'
article = 'course type-along'


demo = gr.Interface(fn=predict,
                    inputs=gr.Image(type='pil'),
                    outputs = [gr.Label(num_top_classes=3,label='predictions'),
                               gr.Number(label='Prediction time(s)')],
                    examples=examples_list,
                    title=title,
                    description=description,
                    article=article)

#launch demo
demo.launch()