import gradio as gr
# import os, subprocess, torchaudio
# import torch
from PIL import Image
from gtts import gTTS
import tempfile
from pydub import AudioSegment
from pydub.generators import Sine
# from fairseq.checkpoint_utils import load_model_ensemble_and_task_from_hf_hub
# from fairseq.models.text_to_speech.hub_interface import TTSHubInterface
import soundfile

import dlib
import cv2
import imageio
import os
import gradio as gr
import os, subprocess, torchaudio
from PIL import Image
import ffmpeg



block = gr.Blocks()



def generate_ocr(method,image,gender):
    print("efef")
    return "Hello"
    
def run():
  with block:
  
    with gr.Group():
      with gr.Box():
        with gr.Row().style(equal_height=True):
          image_in = gr.Image(show_label=False, type="filepath")
          # audio_in = gr.Audio(show_label=False, type='filepath')
          input_text=gr.Textbox(lines=3, value="Hello How are you?", label="Input Text")
          gender = gr.Radio(["Female","Male"],value="Female",label="Gender")
          video_out = gr.Textbox(label="output")
          # video_out = gr.Video(show_label=False)
        with gr.Row().style(equal_height=True):
          btn = gr.Button("Generate")          

    btn.click(generate_ocr, inputs=[image_in, input_text,gender], outputs=[video_out])
    # block.queue()
    block.launch(server_name="0.0.0.0", server_port=7860)

if __name__ == "__main__":
    run()


# image = gr.Image(show_label=True, type="filepath",label="Input Image")
# input_text=gr.Textbox(lines=3, value="Hello How are you?", label="Input Text")
# gender = gr.Radio(["Female","Male"],value="Female",label="Gender")
# output = gr.Video(show_label=True,label="Output")

# demo = gr.Interface(generate_ocr,[image,input_text,gender],[output],title="One Shot Talking Face from Text",)
# demo.launch(server_name="0.0.0.0", server_port=7860)