app.py

import gradio as gr
import os
import torch
import whisper

voice_model = whisper.load_model("large-v2")

def transcribe_audio(audio_path):
    result = voice_model.transcribe(whisper.pad_or_trim(whisper.load_audio(audio_path)))["text"]
    os.remove(audio_path)
    return {"text": result}

# Create a Gradio interface
input_audio = gr.Audio(type="filepath")
output_text = gr.outputs.Textbox()

iface = gr.Interface(fn=transcribe_audio, inputs=input_audio, outputs=output_text, title="Audio Transcription")
iface.launch()