Image-classifier-model

The program learns from a pre-trained model and then come up with its own model which is is then used to classifier

NOTE: The program can be run in a jupyter notebook by loading the Image_classifier_project.html into the3 notebook.

It can also be run from the command lie via running: + python train.py image_dir (You may refer to -h on the file for help) inorder to train the model and save the checkpoints + python predict.py image_path with optional params as stipulated in the docstring) in order to load a saved checkpoint and then use it it to predict input image

The program requires the flowers folder which is in the following layout:

• flowers

  • train

    • 1
      • specific image files
    • 2
      • specific image files . . .
    • 102
      • specific image files

  • test

    • 1
      • specific image files
    • 2
      • specific image files . . .
    • 102
      • specific image files

  • valid

    • 1
      • specific image files
    • 2
      • specific image files . . .
    • 102
      • specific image files

  #Happy testing and learning of the AI products @Pitson Josiah Mwakabila