app.py

#!/usr/bin/python
# -*- coding: utf-8 -*-

import gradio as gr
from lib.config import (
    AUDIO_DIR,
    TRANSCRIPTS_DIR,
)
from lib.media import (
    extract_audio_from_video,
    save_file,
)

from lib.model import (
    transcribe_audio,
)

# -->> Tunables <<---------------------


# -->> Definitions <<------------------


def interface_creation():
    with gr.Blocks() as interface:
        with gr.Row():
            with gr.Column():
                gr.Markdown("## Extract Audio from Video")
                audio_inp = [
                    gr.Video(label="Upload Video"),
                    gr.Textbox(value=AUDIO_DIR, label="Output Audio Folder Path"),
                ]
                audio_btn = gr.Button("Extract Audio")

            with gr.Column():
                gr.Markdown("### Extracted Audio from Video")
                audio_out = [
                    gr.Audio(label="Extracted Audio"),
                    gr.Textbox(label="Audio Path"),
                ]

        with gr.Row():
            with gr.Column():
                gr.Markdown("## Transcribe Audio")
                transcribe_inp = [
                    gr.Audio(label="Upload Audio"),
                    gr.Textbox(label="Audio Path"),
                    gr.Textbox(
                        value=TRANSCRIPTS_DIR, label="Output Transcript Folder Path"
                    ),
                ]
                transcribe_btn = gr.Button("Transcribe Audio")

            with gr.Column():
                gr.Markdown("### Transcription:")
                transcribe_out = gr.Textbox(label="Transcription")
                download_btn = gr.Button("Download Transcript")
                download_path = gr.Textbox(label="Download File Path")
                download_link = gr.Textbox(label="Download Link")


        audio_btn.click(
            fn=extract_audio_from_video, inputs=audio_inp, outputs=audio_out
        ).then(
            fn=lambda audio_file, audio_path: [audio_file, audio_path],
            inputs=audio_out,
            outputs=[transcribe_inp[0], transcribe_inp[1]],
        )

        transcribe_btn.click(
            transcribe_audio, inputs=transcribe_inp, outputs=transcribe_out
        )

        download_btn.click(
            save_file,
            inputs=[transcribe_inp[1], transcribe_inp[2], transcribe_out],
            outputs=[download_path, download_link]
        )

    return interface


# -->> API <<--------------------------

iface = interface_creation()
iface.launch()

# -->> Execute <<----------------------


# -->> Export <<-----------------------