copywriter / app.py
JeCabrera's picture
Update app.py
c99083d verified
raw
history blame
3.97 kB
TITLE = """<h1 align="center">Gemini Playground ✨</h1>"""
SUBTITLE = """<h2 align="center">Play with Gemini Pro and Gemini Pro Vision</h2>"""
import os
import time
from typing import List, Tuple, Optional, Union
import google.generativeai as genai
import gradio as gr
from dotenv import load_dotenv
# Cargar las variables de entorno desde el archivo .env
load_dotenv()
# Obtener la clave de la API de las variables de entorno
GOOGLE_API_KEY = os.getenv("GOOGLE_API_KEY")
# Verificar que la clave de la API esté configurada
if not GOOGLE_API_KEY:
raise ValueError("GOOGLE_API_KEY is not set in environment variables.")
# Configurar la API
genai.configure(api_key=GOOGLE_API_KEY)
# Constantes
IMAGE_WIDTH = 512
CHAT_HISTORY = List[Tuple[Optional[Union[Tuple[str], str]], Optional[str]]]
def user(text_prompt: str, chatbot: CHAT_HISTORY):
"""
Maneja las entradas del usuario en el chatbot.
"""
if text_prompt:
chatbot.append((text_prompt, None))
return "", chatbot
def bot(
model_choice: str,
system_instruction: Optional[str],
chatbot: CHAT_HISTORY
):
"""
Maneja las respuestas del modelo generativo.
"""
generation_config = genai.types.GenerationConfig(
temperature=0.7,
max_output_tokens=8192,
top_k=10,
top_p=0.9
)
# Usar un valor predeterminado si system_instruction está vacío
if not system_instruction:
system_instruction = "You are a helpful assistant."
# Obtener el prompt más reciente del usuario
text_prompt = [chatbot[-1][0]] if chatbot and chatbot[-1][0] else []
# Crear y configurar el modelo generativo
model = genai.GenerativeModel(
model_name=model_choice,
generation_config=generation_config,
system_instruction=system_instruction,
)
# Generar contenido usando streaming
response = model.generate_content(text_prompt, stream=True)
# Preparar la respuesta para el chatbot
chatbot[-1] = (chatbot[-1][0], "")
for chunk in response:
chatbot[-1] = (chatbot[-1][0], chatbot[-1][1] + chunk.text)
yield chatbot
# Componentes de la interfaz de usuario
system_instruction_component = gr.Textbox(
placeholder="Enter system instruction...",
label="System Instruction",
lines=2
)
chatbot_component = gr.Chatbot(label='Gemini', bubble_full_width=False, height=300)
text_prompt_component = gr.Textbox(placeholder="Message...", show_label=False, autofocus=True)
run_button_component = gr.Button(value="Run", variant="primary")
model_choice_component = gr.Dropdown(
choices=["gemini-1.5-flash", "gemini-2.0-flash-exp", "gemini-1.5-pro"],
value="gemini-1.5-flash",
label="Select Model"
)
user_inputs = [text_prompt_component, chatbot_component]
bot_inputs = [model_choice_component, system_instruction_component, chatbot_component]
# Definir la interfaz de usuario
with gr.Blocks() as demo:
gr.HTML(TITLE)
gr.HTML(SUBTITLE)
with gr.Column():
# Campo de selección de modelo arriba
model_choice_component.render()
chatbot_component.render()
with gr.Row():
text_prompt_component.render()
run_button_component.render()
# Crear el acordeón para la instrucción del sistema al final
with gr.Accordion("System Instruction", open=False):
system_instruction_component.render()
run_button_component.click(
fn=user,
inputs=user_inputs,
outputs=[text_prompt_component, chatbot_component],
queue=False
).then(
fn=bot, inputs=bot_inputs, outputs=[chatbot_component],
)
text_prompt_component.submit(
fn=user,
inputs=user_inputs,
outputs=[text_prompt_component, chatbot_component],
queue=False
).then(
fn=bot, inputs=bot_inputs, outputs=[chatbot_component],
)
# Lanzar la aplicación
demo.queue(max_size=99).launch(debug=False, show_error=True)