import gradio as gr
from utils.predict import predict_action
import os
import glob

##Create Dataset for loading examples
example_list = glob.glob("examples/*")
example_list = list(map(lambda el:[el], example_list))


# def load_example(video):
#     return video[0]

# demo = gr.Blocks()

input_video = gr.Video(label="Input Video", show_label=True)
output_label = gr.Label(label="Model Output", show_label=True)
output_gif = gr.Image(label="Video Gif", show_label=True)
title = "Video Classification with Transformers"
description = "This space demonstrates the use of a hybrid (CNN-Transformer based) model for video classification. \n The model can classify videos belonging to the following action categories: CricketShot, Punch, ShavingBeard, TennisSwing, PlayingCello. \n Upload a video and try out 🤗 "

article = '\n Demo created by: <a href=\"https://www.linkedin.com/in/shivalika-singh/\">Shivalika Singh</a> <br> Based on this <a href=\"https://keras.io/examples/vision/video_transformers/\">Keras example</a> by <a href=\"https://twitter.com/RisingSayak\">Sayak Paul</a> <br> Demo Powered by this <a href=\"https://huggingface.co/shivi/video-transformers/\"> Video Classification</a> model'

gr.Interface(predict_action, input_video, [output_label, output_gif], examples=example_list, allow_flagging=False, analytics_enabled=False,
  title=title, description=description, cache_examples=True, article=article).launch(enable_queue=True,share=True)


# with demo:
    
#     gr.Markdown("# **<p align='center'>Video Classification with Transformers</p>**")
#     gr.Markdown("This space demonstrates the use of hybrid Transformer-based models for video classification that operate on CNN feature maps.")
    
#     with gr.Tabs():
                
#         with gr.TabItem("Upload & Predict"):
#             with gr.Box():
                
#                 with gr.Row():
#                     input_video = gr.Video(label="Input Video", show_label=True)
#                     output_label = gr.Label(label="Model Output", show_label=True)
#                     output_gif = gr.Image(label="Video Gif", show_label=True)
            
#             gr.Markdown("**Predict**")
            
#             with gr.Box():
#                 with gr.Row():
#                     submit_button = gr.Button("Submit")
            
#             gr.Markdown("**Examples:**")
#             gr.Markdown("The model is trained to classify videos belonging to the following classes:")
#             gr.Markdown("CricketShot, PlayingCello, Punch, ShavingBeard, TennisSwing")

#             with gr.Column():
#                 # gr.Examples("examples", [input_video], [output_label,output_gif], predict_action, cache_examples=True)
#                 examples = gr.components.Dataset(components=[input_video], samples=example_list, type='values')
#                 examples.click(load_example, examples, input_video)
        
#     submit_button.click(predict_action, inputs=input_video, outputs=[output_label,output_gif])
    
#     gr.Markdown('\n Author: <a href=\"https://www.linkedin.com/in/shivalika-singh/\">Shivalika Singh</a> <br> Based on this <a href=\"https://keras.io/examples/vision/video_transformers/\">Keras example</a> by <a href=\"https://twitter.com/RisingSayak\">Sayak Paul</a> <br> Demo Powered by this <a href=\"https://huggingface.co/shivi/video-transformers/\"> Video Classification</a> model')
    
# demo.launch()