Some Classification Intergration

Folders & Files description

• Folders

  • checkpoints
    • here to save training checkpoint, log, etc
  • core
    • training, validating fuctions are in this folder
  • loss
    • some loss fuctions
  • models
    • all models are in this folder, you could import from models
  • my_data
    • train data, test data folder
  • result
    • predict result

• Files

  • mean_std_calculation.ipynb
    • calculate train set mean & std
  • predict_classification.py
    • predict are processing in this script
  • test.py
    • test script, you could run some test in this scrip, if you want to
  • torch_convert.py
    • get all the training images and labels, and split train test set
  • torch_dataset.py
    • pytorch dataset setupt according tor data format
  • train_efficient.py
    • train script, here is efficientnet, so name train_efficient
  • train_efficientnet_album.py
    • train script, efficientnet training using albumentations for data augmentation