izer = AutoTokenizer.from_pretrained("ydshieh/vit-gpt2-coco-en")
        >>> model = VisionEncoderDecoderModel.from_pretrained("ydshieh/vit-gpt2-coco-en")

        >>> url = "http://images.cocodataset.org/val2017/000000039769.jpg"
        >>> img = Image.open(requests.get(url, stream=True).raw)
        >>> pixel_values = image_processor(images=img, return_tensors="pt").pixel_values  # Batch size 1

        >>> output_ids = model.generate(
        ...     pixel_values, max_length=16, num_beams=4, return_dict_in_generate=True
        ... ).sequences

        >>> preds = decoder_tokenizer.batch_decode(output_ids, skip_special_tokens=True)
        >>> preds = [pred.strip() for pred in preds]

        >>> assert preds == ["a cat laying on top of a couch next to another cat"]
        ```Ú