app.py

from transformers import pipeline
from openai import OpenAI
import gradio as gr
import os
import dotenv



api_key = os.getenv("OPENROUTER_API_KEY")

client = OpenAI(
    base_url="https://openrouter.ai/api/v1",
    api_key=api_key,
)


asr = pipeline("automatic-speech-recognition", "facebook/wav2vec2-base-960h")


def speech_to_text(speech):
    text = asr(speech)["text"]
    text = text.lower()
    return text

def chatbot(message, history):
    messages = []
    for human, assistant in history:
        messages.append({"role": "user", "content": human})
        messages.append({"role": "assistant", "content": assistant})
    messages.append({"role": "user", "content": message})

    response = client.chat.completions.create(
        model="meta-llama/llama-3.1-405b-instruct:free",
        messages=messages,
        temperature=0.7,
        max_tokens=1000
    )
    return response.choices[0].message.content
    

with gr.Blocks() as demo:
    with gr.Row():
        audio_file = gr.Audio(type="filepath", label="Sube tu archivo de audio")

        text_input = gr.Textbox(label="Texto transcrito (o ingresa tu mensaje)", placeholder="El texto aparecerá aquí...")
        text_output = gr.Textbox(label="Respuesta del Chatbot", placeholder="La respuesta del chatbot aparecerá aquí...", interactive=False)

        b1 = gr.Button("Reconocer Audio y Preguntar al Chatbot")

    def process_audio_and_chat(audio_input, history):
        if audio_input:

            audio_text = speech_to_text(audio_input)

            response = chatbot(audio_text, history)
            print(response)
            return audio_text, response
            
        return "", "Por favor, sube un archivo de audio."


    b1.click(process_audio_and_chat, inputs=[audio_file, text_input], outputs=[text_input, text_output])


demo.launch(debug=True)