manual_prompts.json

{
    "COMMENT": "Sources for these prompts include https://huggingface.co/datasets/liuhaotian/LLaVA-Pretrain/viewer and https://huggingface.co/datasets/HuggingFaceM4/M3IT",
    "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?",
            "Describe the image briefly.",
            "Write a succinct description of the image, capturing its main components, the relationships between them, and any notable details.",
            "Create a concise caption that accurately describes the main elements in the image provided.",
            "Write a brief, yet comprehensive, description of the image.",
            "Describe the image in a clear and concise manner.",
            "For the given image, provide a one-sentence summary that captures the most important details.",
            "Generate a short caption for the picture.",
            "Write a short and informative description that highlights the primary subjects and actions occurring in the given image.",
            "Provide a concise and informative caption for the image, focusing on the primary subjects.",
            "Write a clear description of the image, make sure the key features are well covered.",
            "Offer a succinct explanation of the picture presented."
        ]
    },
    "CaptioningPretraining": {
        "raw": [
            "Generate a short caption of the image.",
            "Describe the image concisely.",
            "Provide a brief description of the given image."
        ],
        "llava": [
            "Give a brief description of image.",
            "Give a brief description of the image.",
            "Provide a brief description of the given image.",
            "Provide a one-sentence caption for the provided image.",
            "Write a terse but informative summary of the picture.",
            "Describe the image concisely.",
            "Generate a clear and concise summary of the photo."
        ]
    },
    "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?"
        ]
    }
}