app.py

import gradio as gr
from transformers import pipeline
import whisper


model_whisper = whisper.load_model("base")


nlp_model = pipeline("summarization", model="t5-base")


def process_audio(audio_file):

    transcription = model_whisper.transcribe(audio_file)["text"]
    
    summary = nlp_model(transcription)[0]["summary_text"]
    
    return transcription, summary

interface = gr.Interface(
    fn=process_audio,
    inputs=gr.Audio(type="filepath"),
    outputs=["text", "text"],
    title="Transcripción y Análisis de Audio - PNL",
    description="Sube un archivo de audio para transcribirlo y resumir el contenido o  activa tu microfono y habla."
)

interface.launch()