NoumbissiValere / imageclassifier

An AI program to classifier images to the respective categories using pre-trained models and Pytorch

imageclassifier

An AI program to classifier images to their respective categories (102 outputs) using pre-trained models and Pytorch.

This project is divided into two sections,

• A train section
• A predict section

The train section of the program is used to train our model. below are the various commands you could use to with this section

• basic usage python train.py data_directory . The data_directory must contain at least a train and test directory. the train directory is used to train the model while the test directory is used to test the model as we train it. This is to enable us visualize the error loss and model accuracy while we train our model. Giving us an idea on how we could better fine tune our model.
• Setting the checkpoint directory python train.py data_dir --save_dir save_directory
• Choose a particular architecture with which you want your model to be trained with python train.py data_dir --arch "vgg13"
• Setting the hyperparameters for fine turning python train.py data_dir --learning_rate 0.01 --hidden_units 512 --epochs 20
• Using the GPU python train.py data_dir --gpu

The predict section involves feeding our trained model with data it has never seen to see if it can do the right predictions

• Basic usage python predict.py /path/to/image checkpoint
• Return top k most likely categories python predict.py input checkpoint --top_k 3
• Using a mapping of categories to real names python predict.py input checkpoint --category_names cat_to_name.json
• Using a GPU python predict.py input checkpoint --gpu

Enjoy...