CPS 803 - Garbage Classification

• This repo contains all the test code samples, data set configurations and pipeline scripts

Prepare the Dataset

• You can download the datset with the following command in the data folder

kaggle datastes download asdasdasasdas/garbage-classification

• move the contents in the "Garbage Classification/Garbage Classification" up 1 directory

Running Models and Data Preparation Steps

• when running scripts change directory in to the my_scripts folder first

  • This is because we use relative pathing

• To prepare the dataset run python create_train_test_split.py and python structure_train_data_for_nn.py

  • This is assuming you have prepared the dataset in "data/Garbage Classification" folder

• To run an efficient net model run: python efficientnetb[n].py where [n] is either 1,2,3,or 4

• To run the resnet models run: python resnet.py

• To run the keras tuner run: python tune_script.py

  • if you are running it on a server and would like it to run in the background run this: nohup python tune_script.py > run_tuner.txt &
  • to check up on the process you can cat run_tuner.txt or ps ax | grep tune_script.py