app.py

from huggingface_hub import login
from diarization import startDiarization
import ffmpeg
import gradio as gr
import os

hugging_face_token = os.environ["HUGGING_FACE_TOKEN"]
login(token=hugging_face_token)

# Get video & convert it
inputs_interface = gr.inputs.Video(label="Insert video")
in_file = ffmpeg.input(inputs_interface).audio
out_file = ffmpeg.output(in_file, "input.wav", f="wav").run()


def prepareInput():
    return startDiarization(out_file)


gr.Interface(
    prepareInput,
    inputs=inputs_interface,
    outputs="text",
    title="Get Diarization"
).launch()