Classify Flowers in Pytorch

The final project of Udacity's AI Programming with Python Nanodegree program. A classifier built to recognize different species of flowers, trained on the 102 category flower dataset and uses VGG-16 or AlexNet as the backbone. The project consists of a command line application and a notebook.

Command Line Application

• Clone the repository and open it in any IDE
• Run pip install -r requirements.txt
• Run train.py for training the model and predict.py for making predictions
• You can change different parameters like:
  • Architecture: VGG-16 or AlexNet
  • Hyperparameters: Set hyperparamets like learning rate, epochs, hidden units
  • Device: CPU or GPU
  • Top K Classes: Predict top K classes along with associated probabilities

Output Example