import gradio as gr
from transformers import pipeline
import torch
import os
from huggingface_hub import login

hf_token = os.environ["HF_TOKEN"]
login(token=hf_token)


device = 0 if torch.cuda.is_available() else -1
pipe = pipeline("text-generation", model="google/gemma-3n-E4B-it-litert-preview", device=device)


def responder(prompt):
    respuesta = pipe(prompt, max_new_tokens=80, do_sample=True, temperature=0.7)[0]['generated_text']
    return respuesta

with gr.Blocks() as demo:
    gr.Markdown("## 🤖 AmInside 1.0\nTu asistente IA gratuito en Hugging Face")
    entrada = gr.Textbox(label="Escribe tu mensaje")
    salida = gr.Textbox(label="Respuesta")
    entrada.submit(fn=responder, inputs=entrada, outputs=salida)

demo.launch()