llava/datasets/prompts.py

__all__ = ['cc_sbu_prompt', 'internvid_prompt', 'tt_caption_prompt', 'm3it_infer_prompt']

cc_sbu_prompt = [
    'Describe the image briefly.',
    'Write a relevant description to pair with the image.'
]

internvid_prompt = [
    'Describe the video briefly.',
    'Describe the video shortly.',
    'Provide a brief description of the given video clip.',
    'Give a short and clear explanation of the subsequent video clip.',
    'Summarize the visual content of the following video.'
]

tt_caption_prompt = [
    "Describe the given video in detail.",
    "Elaborate on the video's content.",
    "Provide a detailed explanation of the video.",
    "Explain the video thoroughly.",
    "Give an in-depth description of the video."
]

tt_caption_prompt2 = [
    "Describe the following video in detail.",
]

m3it_infer_prompt = "Assess the image provided and respond to the relevant question."

synthdog_prompt = 'write down the text overlays in the provided image.'

ocr_prompt = ['From the video, there are some text overlays:',
             'The video also contains some OCR info:',
             'The text overlay says:']