ImageSearchLightningCLIP icon indicating copy to clipboard operation
ImageSearchLightningCLIP copied to clipboard
verified publisher icon ForJadeForest

Model

Open matveymotyvin opened this issue 3 years ago • 6 comments

What model architectures have you used to encode text and images?

matveymotyvin avatar Feb 19 '23 19:02 matveymotyvin

Hi, thanks for your interesting for this work. The distillation model is a small vit and a small transformer. These model architecture is similar to the CLIP original model(eg. vit32-B).

Btw, the diatillation code is in the master branch. And the code about app is in the main branch.

You can checkout the master branch to check the implementation details.

ForJadeForest avatar Feb 20 '23 11:02 ForJadeForest

Thanks!

matveymotyvin avatar Feb 20 '23 11:02 matveymotyvin

Hello, can you share the full version of the code to start the model distillation process on a computer (for experiments), I would be very grateful if you could send it by email [email protected]

matveymotyvin avatar Feb 27 '23 15:02 matveymotyvin

Sure, but it will take me some time to recall the code and I will reorganize a new version of the code in this repository.

ForJadeForest avatar Feb 27 '23 15:02 ForJadeForest

Thank you so much, I will be waiting!!!

matveymotyvin avatar Feb 27 '23 16:02 matveymotyvin

Now you can check the new version code! If you have any further questions, please feel free to consult 👏.

ForJadeForest avatar Feb 28 '23 05:02 ForJadeForest