app.py

from PIL import Image
import spaces

import gradio as gr

MODEL_ID = "davidr99/qwen2-7b-instruct-blackjack"


from transformers import Qwen2VLForConditionalGeneration, AutoTokenizer, AutoProcessor
from qwen_vl_utils import process_vision_info

@spaces.GPU(duration=30)
def blackjack_ai(image):

    model = Qwen2VLForConditionalGeneration.from_pretrained(MODEL_ID, torch_dtype="auto", device="auto")
    processor = AutoProcessor.from_pretrained(MODEL_ID)

    instruction = "extract json from this image."
    
    messages = [
        {"role": "user", "content": [
            {"type": "image", "image": image},
            {"type": "text", "text": instruction}
        ]}
    ]

    print(messages)

    # Preparation for inference
    text = processor.apply_chat_template(
        messages, tokenize=False, add_generation_prompt=True
    )
    image_inputs, video_inputs = process_vision_info(messages)
    inputs = processor(
        text=[text],
        images=image_inputs,
        videos=video_inputs,
        padding=True,
        return_tensors="pt",
    )
    inputs = inputs.to("cuda")

    # Inference: Generation of the output
    generated_ids = model.generate(**inputs, max_new_tokens=128)
    generated_ids_trimmed = [
        out_ids[len(in_ids) :] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
    ]
    output_text = processor.batch_decode(
        generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
    )

    return output_text

with gr.Blocks() as demo:
   
    image = gr.Image(type="filepath")
    submit = gr.Button("Submit")
    output = gr.TextArea()

    submit.click(blackjack_ai, inputs=[image], outputs=[output])
    
demo.launch()