Use BLIP-2 for Image Captioning

salesforce / LAVIS

LAVIS - A One-stop Library for Language-Vision Intelligence

BSD 3-Clause "New" or "Revised" License

9.22k stars 914 forks source link

Use BLIP-2 for Image Captioning #692

Open ArefAz opened 2 months ago

ArefAz commented 2 months ago

Hi,

Firstly, thank you for maintaining such an awesome repository!

I'm particularly interested in using BLIP-2 for image captioning. Could you please provide some guidance on whether it's feasible to use BLIP-2 into for this task and any steps I should take to do so?

Best, Aref

Thomas2419 commented 2 months ago

Hi Aref,

If you check the repository and you go to projects, then to blip-2 there is some example usage of how you can do just that!

Regards, Thomas