manual_prompts.json

{
    "Captioning": {
        "raw": [
            "Can you briefly explain what you see in the image?",
            "Describe what's happening in this image in one short sentence.",
            "Write a short caption that accurately represents the content of this image.",
            "Please generate a descriptive caption for the image provided.",
            "How would you summarize the scene depicted in the picture in short?"
        ]
    },
    "OCR": {
        "raw": [
            "Can you read the text from image and output here?",
            "Extract and document the text from the provided image.",
            "Converting the text embedded in this image into a readable document.",
            "Transcribe all the text you find.",
            "Can you extract all visible text from the image here?"
        ]
    },
    "VQA": {
        "raw": [
            "Given the image, answer the following question with few words.",
            "Answer the following question: ",
            "What is the answer to this question?",
            "Write the answer: ",
            "Please answer this question: "
        ]
    }
}