This is a self contained notebook using openai/CLIP to classify images from dataset of images inside a folder using a token and group them in another folder where name of folder having token as the folder name.

For example, if token = "a bike" then all the images resembling the token, a bike, will be classified and grouped into a new folder in the parent directory named "a bike".

1. NumPy
2. PyTorch

• Parent Directory
  • classify.ipynb
  • images
    • car.png
    • bike.png
    • bike2.png

here token = "a bike"

• Parent Directory
  • classify.ipynb
  • bike
    • bike.png
    • bike2.png
  • images
    • car.png