yxuansu / PandaGPT

[TLLM'23] PandaGPT: One Model To Instruction-Follow Them All
https://panda-gpt.github.io/
Apache License 2.0
768 stars 60 forks source link

The issue of the input of caption #18

Open liuxuannan opened 1 year ago

liuxuannan commented 1 year ago

I have a question about the position format of the caption in the input data in the command data. For example, the following sentence in the paper, A video of a Super-hero Movie. Is this sentence part of the text prompt, or does it need to be embedded through the imagebind model and then input into LLM?