app.py

from transformers.pipelines.audio_utils import ffmpeg_read
from transformers import WhisperForConditionalGeneration, WhisperProcessor, pipeline
import gradio as gr
import numpy as np

def process_transcribe(file):
    audio_nparray = file[1]
    my_list = audio_nparray.tolist()

    endpoint = runpod.Endpoint("14ggfq6a17uim9")

    run_request = endpoint.run_sync(
        {"audio_list": my_list}
    )
    raw_text = run_request
    
    return raw_text

with gr.Blocks(theme=gr.themes.Default(primary_hue="blue")) as iface:
    with gr.Tab("App"):
        with gr.Row():
            with gr.Column():
                audio_id = gr.Textbox(label="Audio")
                audio_file = gr.Audio(sources=["upload"], type="numpy")
                submit_btn = gr.Button("Submit", variant="primary")
            with gr.Column():
                raw_transcript = gr.Textbox(label="Transcription")
        submit_btn.click(process_transcribe, inputs=[audio_file], outputs=[raw_transcript])
iface.launch(debug=True)