app.py

import numpy as np
import torch
import torchvision.transforms as T
import gradio as gr
from PIL import Image
from torchvision.transforms.functional import InterpolationMode

from transformers import AutoModel, AutoTokenizer

IMAGENET_MEAN = (0.485, 0.456, 0.406)
IMAGENET_STD = (0.229, 0.224, 0.225)

def build_transform(input_size):
    MEAN, STD = IMAGENET_MEAN, IMAGENET_STD
    transform = T.Compose([
        T.Lambda(lambda img: img.convert('RGB') if img.mode != 'RGB' else img),
        T.Resize((input_size, input_size), interpolation=InterpolationMode.BICUBIC),
        T.ToTensor(),
        T.Normalize(mean=MEAN, std=STD)
    ])
    return transform

def find_closest_aspect_ratio(aspect_ratio, target_ratios, width, height, image_size):
    best_ratio_diff = float('inf')
    best_ratio = (1, 1)
    area = width * height
    for ratio in target_ratios:
        target_aspect_ratio = ratio[0] / ratio[1]
        ratio_diff = abs(aspect_ratio - target_aspect_ratio)
        if ratio_diff < best_ratio_diff:
            best_ratio_diff = ratio_diff
            best_ratio = ratio
        elif ratio_diff == best_ratio_diff:
            if area > 0.5 * image_size * image_size * ratio[0] * ratio[1]:
                best_ratio = ratio
    return best_ratio

def dynamic_preprocess(image, min_num=1, max_num=12, image_size=448, use_thumbnail=False):
    orig_width, orig_height = image.size
    aspect_ratio = orig_width / orig_height

    # calculate the existing image aspect ratio
    target_ratios = set(
        (i, j) for n in range(min_num, max_num + 1) for i in range(1, n + 1) for j in range(1, n + 1) if
        i * j <= max_num and i * j >= min_num)
    target_ratios = sorted(target_ratios, key=lambda x: x[0] * x[1])

    # find the closest aspect ratio to the target
    target_aspect_ratio = find_closest_aspect_ratio(
        aspect_ratio, target_ratios, orig_width, orig_height, image_size)

    # calculate the target width and height
    target_width = image_size * target_aspect_ratio[0]
    target_height = image_size * target_aspect_ratio[1]
    blocks = target_aspect_ratio[0] * target_aspect_ratio[1]

    # resize the image
    resized_img = image.resize((target_width, target_height))
    processed_images = []
    for i in range(blocks):
        box = (
            (i % (target_width // image_size)) * image_size,
            (i // (target_width // image_size)) * image_size,
            ((i % (target_width // image_size)) + 1) * image_size,
            ((i // (target_width // image_size)) + 1) * image_size
        )
        # split the image
        split_img = resized_img.crop(box)
        processed_images.append(split_img)
    assert len(processed_images) == blocks
    if use_thumbnail and len(processed_images) != 1:
        thumbnail_img = image.resize((image_size, image_size))
        processed_images.append(thumbnail_img)
    return processed_images

def load_image(image, input_size=448, max_num=12):
    transform = build_transform(input_size=input_size)
    images = dynamic_preprocess(image, image_size=input_size, use_thumbnail=True, max_num=max_num)
    pixel_values = [transform(image) for image in images]
    pixel_values = torch.stack(pixel_values)
    return pixel_values

# Load model and tokenizer globally to avoid reloading for each request
@torch.inference_mode()
def load_models():
    model = AutoModel.from_pretrained(
        "5CD-AI/Vintern-1B-v3_5",
        torch_dtype=torch.bfloat16,
        low_cpu_mem_usage=True,
        trust_remote_code=True,
        use_flash_attn=False,
    ).eval()
    
    # Move model to GPU if available
    device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
    model = model.to(device)
    
    tokenizer = AutoTokenizer.from_pretrained("5CD-AI/Vintern-1B-v3_5", trust_remote_code=True, use_fast=False)
    
    return model, tokenizer, device

model, tokenizer, device = load_models()

def generate_response(image, question):
    if image is None:
        return "Vui lòng tải lên một hình ảnh."
    if not question:
        question = '<image>\nTrích xuất thông tin chính trong ảnh và trả về dạng markdown.'
    
    # Convert image to PIL if it's a file upload
    if isinstance(image, str):
        image = Image.open(image).convert('RGB')
    else:
        image = image.convert("RGB")
    
    # Process image
    pixel_values = load_image(image, max_num=6).to(torch.bfloat16).to(device)
    
    # Generate response
    generation_config = dict(max_new_tokens=1024, do_sample=False, num_beams=3, repetition_penalty=2.5)
    
    response, _ = model.chat(tokenizer, pixel_values, question, generation_config, history=None, return_history=True)
    
    return response


    
# # Create Gradio interface
# with gr.Blocks() as demo:
gr.Markdown("# Vintern-1B-v3.5 Demo")
gr.Markdown("Tải lên hình ảnh và đặt câu hỏi (hoặc để trống để trích xuất thông tin).")

with gr.Row():
    with gr.Column():
        image_input = gr.Image(type="pil", label="Hình ảnh")
        question_input = gr.Textbox(
            placeholder="<image>\nTrích xuất thông tin chính trong ảnh và trả về dạng markdown.", 
            label="Câu hỏi (để trống để trích xuất thông tin)"
        )
        submit_btn = gr.Button("Gửi")
    
    with gr.Column():
        output = gr.Markdown(label="Kết quả")

title = "🇻🇳 Vietnamese Image Captioning"
description = "Tải lên một bức ảnh và mô hình sẽ sinh mô tả tiếng Việt tương ứng."

demo = gr.Interface(
    fn=generate_response,
    inputs=[image_input, question_input],
    outputs="text",
    title=title,
    description=description,
)

if __name__ == "__main__":
    demo.launch()
    
#     submit_btn.click(
#         fn=generate_response,
#         inputs=[image_input, question_input],
#         outputs=output
#     )
    
#     # gr.Examples(
#     #     [
#     #         ["example1.jpg", "<image>\nTrích xuất thông tin chính trong ảnh và trả về dạng markdown."],
#     #         ["example2.jpg", "<image>\nĐây là hình ảnh gì?"],
#     #     ],
#     #     inputs=[image_input, question_input],
#     # )

#     demo.launch()