The code is based on the "ALFA'' public code.

Requirement:
    Anaconda3
    Python
    Pytorch >= 1.5.0
    numpy
    tensorboard
    scipy
    matplotlib
    pbzip2
    pydrive
    pillow
    tqdm

Datasets
    Datasets can be placed in the dataset folder. The dataset should be divided as three subfolder: train\val\test. All image should be the .jpg format.
    The mini_imagenet can be downloaded from the link provided by the "ALFA" public code.

Usage:
    Training: go to the "experiments_scripts" and execute "bash Meta_AdaM.sh $GPU_ID"
    Evaluation: After training process finish, execute "bash Meta_AdaM.sh $GPU_ID" again
    To adjust hyperparameter, go to the "experiments_config" folder and adjust the hyperparameter in file "Meta_AdaM.json"


