app.py

from huggingface_hub import login
from diarization import start_diarization
from transcribe import start_transcribe
import ffmpeg
import gradio as gr


def prepare_input(input_file, progress=gr.Progress()):
    output_file = "input.wav"
    progress(0.2, desc="Preparing video")
    ffmpeg.input(input_file).audio.output(
        output_file, format="wav").run()
    progress(0.4, desc="Acquiring diarization")
    start_diarization(output_file)
    progress(0.6, desc="Transcribing audio")
    return start_transcribe(progress)


video_interface = gr.Interface(
    fn=prepare_input,
    inputs=gr.Video(type="file"),
    outputs="files",
    title="Test 2"
)

if __name__ == "__main__":
    video_interface.launch()