handler.py

from typing import Dict, Any
from PIL import Image
import torch
import requests
from io import BytesIO
from transformers import BlipForConditionalGeneration, BlipProcessor

device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')

class EndpointHandler():
    def __init__(self, path=""):
        # Load the processor and model
        self.processor = BlipProcessor.from_pretrained("Salesforce/blip-image-captioning-base")
        self.model = BlipForConditionalGeneration.from_pretrained(
            "Salesforce/blip-image-captioning-base"
        ).to(device)
        self.model.eval()

    def __call__(self, data: Any) -> Dict[str, Any]:
        """
        Args:
            data (:obj:`dict`):
                Includes the input data and the parameters for the inference.
        Return:
            A :obj:`dict`. The object returned contains:
                - "caption": A string corresponding to the generated caption.
        """
        # Extract image URL and parameters
        image_url = data.get("image")
        parameters = data.get("parameters", {})

        if not image_url:
            return {"error": "Missing 'image' field in request body."}

        try:
            # Download the image
            response = requests.get(image_url)
            response.raise_for_status()
            raw_image = Image.open(BytesIO(response.content)).convert("RGB")
        except Exception as e:
            return {"error": f"Failed to fetch image from URL: {str(e)}"}

        # Preprocess the image
        processed_image = self.processor(images=raw_image, return_tensors="pt")
        processed_image["pixel_values"] = processed_image["pixel_values"].to(device)

        # Merge parameters if needed
        processed_image = {**processed_image, **parameters}

        with torch.no_grad():
            out = self.model.generate(**processed_image)

        # Decode the output
        caption = self.processor.decode(out[0], skip_special_tokens=True)
        
        return {"caption": caption}