app.py

import gradio as gr
import spaces
from transformers import Idefics3ForConditionalGeneration, AutoProcessor
import torch
from PIL import Image
from datetime import datetime
import numpy as np
import os


DESCRIPTION = """
# SmolVLM-trl-dpo-rlaif-v Demo

This is a demo Space for a fine-tuned version of [SmolVLM](https://huggingface.co/HuggingFaceTB/SmolVLM-Instruct) trained using [rlaif-v dataset](https://huggingface.co/datasets/HuggingFaceH4/rlaif-v_formatted).

The corresponding model is located [here](https://huggingface.co/HuggingFaceTB/SmolVLM-Instruct-DPO).

For a full tutorial of fine-tuning using DPO, check out [this link](https://huggingface.co/learn/cookbook/index).
"""

model_id = "HuggingFaceTB/SmolVLM-Instruct"
model = Idefics3ForConditionalGeneration.from_pretrained(
    model_id,
    device_map="auto",
    torch_dtype=torch.bfloat16,
    #_attn_implementation="flash_attention_2",
)

processor = AutoProcessor.from_pretrained(model_id)
#adapter_path = "sergiopaniego/smolvlm-instruct-trl-dpo-rlaif-v"
adapter_path = "HuggingFaceTB/SmolVLM-Instruct-DPO"
model.load_adapter(adapter_path)

def array_to_image_path(image_array):
    if image_array is None:
        raise ValueError("No image provided. Please upload an image before submitting.")
    # Convert numpy array to PIL Image
    img = Image.fromarray(np.uint8(image_array))
    
    # Generate a unique filename using timestamp
    timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
    filename = f"image_{timestamp}.png"
    
    # Save the image
    img.save(filename)
    
    # Get the full path of the saved image
    full_path = os.path.abspath(filename)
    
    return full_path


@spaces.GPU
def run_example(image, text_input=None):
    image_path = array_to_image_path(image)
    image = Image.fromarray(image).convert("RGB")
    
    messages = [
    {
            "role": "user",
            "content": [
                {
                    "type": "image",
                    "text": None,
                },
                {
                    "text": text_input, 
                    "type": "text"
                },
            ],
        }
    ]
        
    # Preparation for inference
    text = processor.apply_chat_template(
        messages, tokenize=False, add_generation_prompt=True
    )
    image_inputs = []
    if image.mode != 'RGB':
        image = image.convert('RGB')
    image_inputs.append([image])
    
    inputs = processor(
        text=text,
        images=image_inputs,
        padding=True,
        return_tensors="pt",
    )
    inputs = inputs.to("cuda")
    
    # Inference: Generation of the output
    generated_ids = model.generate(**inputs, max_new_tokens=1024)
    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[0]

css = """
  #output {
    height: 500px; 
    overflow: auto; 
    border: 1px solid #ccc; 
  }
"""

with gr.Blocks(css=css) as demo:
    gr.Markdown(DESCRIPTION)
    with gr.Tab(label="SmolVLM-Instruct-DPO Input"):
        with gr.Row():
            with gr.Column():
                input_img = gr.Image(label="Input Picture")
                text_input = gr.Textbox(label="Question")
                submit_btn = gr.Button(value="Submit")
            with gr.Column():
                output_text = gr.Textbox(label="Output Text")

        submit_btn.click(run_example, [input_img, text_input], [output_text])

demo.queue(api_open=False)
demo.launch(debug=True)