app_huggingface_spaces.py

import gradio as gr
from transformers import AutoModelForCausalLM, AutoTokenizer

# Cargar el modelo y el tokenizer
model_name = "epfl-llm/meditron-7b"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name)

# Función para generar respuesta del modelo
def generate_response(prompt):
    inputs = tokenizer(prompt, return_tensors="pt")
    outputs = model.generate(inputs.input_ids, max_length=200, pad_token_id=tokenizer.eos_token_id)
    response = tokenizer.decode(outputs[0], skip_special_tokens=True)
    return response

# Configurar la interfaz de Gradio
def chat(paciente_input):
    prompt = f"Paciente: {paciente_input}\nDoctor:"
    response = generate_response(prompt)
    return response

iface = gr.Interface(fn=chat, inputs="text", outputs="text", title="Consulta Médica con Meditron-7B")

iface.launch()