app.py

import gradio as gr
from playwright.sync_api import sync_playwright
import torch
from transformers import AutoModelForCausalLM, AutoProcessor
from PIL import Image
import os
import subprocess
import spaces

# Устанавливаем браузер и все системные зависимости
subprocess.run(["playwright", "install", "chromium"], check=True)

model_id = "allenai/Molmo-7B-D-0924"
processor = AutoProcessor.from_pretrained(model_id, trust_remote_code=True)

# Загружаем модель один раз при старте
model = AutoModelForCausalLM.from_pretrained(
    model_id,
    trust_remote_code=True,
    torch_dtype=torch.bfloat16,
    low_cpu_mem_usage=True,
    device_map="auto",
)
model.eval()


@spaces.GPU(duration=120)
def run_agent(url, prompt):
    with sync_playwright() as p:
        browser = p.chromium.launch(
            headless=True,
            args=["--no-sandbox", "--disable-setuid-sandbox", "--disable-dev-shm-usage"]
        )
        context = browser.new_context(viewport={"width": 1280, "height": 720})
        page = context.new_page()

        try:
            target_url = url if url.startswith("http") else f"https://{url}"
            page.goto(target_url, timeout=60000)
            page.wait_for_timeout(3000)

            screenshot_path = "/tmp/snapshot.png"
            page.screenshot(path=screenshot_path)
            image = Image.open(screenshot_path).convert("RGB")

            inputs = processor.process(images=[image], text=prompt)
            inputs = {
                k: v.to(device="cuda") if torch.is_tensor(v) else v
                for k, v in inputs.items()
            }

            with torch.no_grad():
                output = model.generate_from_batch(
                    inputs,
                    max_new_tokens=200,
                    stop_strings=["<|endoftext|>"]
                )

            generated_text = processor.tokenizer.decode(
                output[0], skip_special_tokens=True
            )
            browser.close()
            return image, generated_text

        except Exception as e:
            try:
                browser.close()
            except Exception:
                pass
            return None, f"Ошибка: {str(e)}"


with gr.Blocks() as demo:
    gr.Markdown("# 🚀 Molmo AI Web Agent")
    with gr.Row():
        with gr.Column():
            url_input = gr.Textbox(label="URL", value="google.com")
            prompt_input = gr.Textbox(label="Запрос", value="Point to the search bar")
            btn = gr.Button("Запустить")
    with gr.Row():
        out_img = gr.Image(label="Скриншот", type="pil")
        out_txt = gr.Textbox(label="Ответ")

    btn.click(fn=run_agent, inputs=[url_input, prompt_input], outputs=[out_img, out_txt])

demo.launch()