app.py

import gradio as gr
from transformers import pipeline

transcription = pipeline("automatic-speech-recognition", model = "facebook/wav2vec2-large-xlsr-53-spanish")
clasificator = pipeline("text-classification", model = "pysentimiento/robertuito-sentiment-analysis")

def speech_to_text(audio):
  text = transcription(audio)["text"]
  return text

def text_to_sentiment(text):
  return clasificator(text)[0]["label"]


demo = gr.Blocks()

with demo:
  gr.Markdown("Obtener texto de un audio y analizar el sentimiento")
  with gr.Tabs():
    with gr.TabItem("Transcribir audio en español"):
      with gr.Row():
        audio = gr.Audio(type="filepath")
        transcripcion = gr.Textbox()
      
      with gr.Row():
        gr.Markdown("")

      b1 = gr.Button("Generar texto")

    with gr.TabItem("Análisis de sentimiento en español"):
      with gr.Row():
        texto = gr.Textbox()
        label = gr.Label()
      b2 = gr.Button("Analizar sentimiento")

    b1.click(speech_to_text, inputs = audio, outputs = transcripcion)
    b2.click(text_to_sentiment, inputs=texto, outputs=label)

demo.launch(share=True)