app.py

import spaces
import os
import time
import torch
import gradio as gr
from threading import Thread
from PIL import Image

# Install required packages
import subprocess
subprocess.run('pip install --upgrade transformers', shell=True)
subprocess.run('pip install accelerate', shell=True)

from transformers import AutoProcessor, AutoModelForVisionText2Text

# Model and processor initialization with trust_remote_code=True
processor = AutoProcessor.from_pretrained(
    "Qwen/QVQ-72B-Preview",
    trust_remote_code=True
)

model = AutoModelForVisionText2Text.from_pretrained(
    "Qwen/QVQ-72B-Preview",
    trust_remote_code=True,
    device_map="auto"
).eval()

# Footer
footer = """
<div style="text-align: center; margin-top: 20px;">
    <p>Powered by QVQ-72B Model</p>
</div>
"""

# Vision model function
@spaces.GPU()
def process_image(image, text_input=None):
    try:
        # Convert image to PIL format
        image = Image.fromarray(image).convert("RGB")
        
        # Prepare inputs
        if text_input:
            inputs = processor(text=text_input, images=image, return_tensors="pt")
        else:
            inputs = processor(images=image, return_tensors="pt")
        
        # Move inputs to the same device as the model
        inputs = {k: v.to(model.device) for k, v in inputs.items()}
        
        # Generate output
        outputs = model.generate(**inputs, max_new_tokens=1000)
        
        # Decode response
        response = processor.batch_decode(outputs, skip_special_tokens=True)[0]
        
        return response
    except Exception as e:
        return f"Error processing image: {str(e)}"

# CSS styling
css = """
footer {
    visibility: hidden;
}
"""

# Gradio interface
with gr.Blocks(theme="Yntec/HaleyCH_Theme_Orange", css=css) as demo:
    with gr.Row():
        input_img = gr.Image(label="Input Image")
    with gr.Row():
        text_input = gr.Textbox(label="Question (Optional)")
    with gr.Row():
        submit_btn = gr.Button(value="Submit")
    with gr.Row():
        output_text = gr.Textbox(label="Response")

    submit_btn.click(process_image, [input_img, text_input], [output_text])
    
    gr.HTML(footer)

# Launch the app
demo.launch(debug=True)