config.py

class config_USC_pmnetV3_V2:
    def __init__(self,):
        # basics
        self.batch_size = 16
        self.exp_name = 'augmented_config_USC_pmnetV3_V2'
        self.num_epochs = 100
        self.val_freq = 1
        self.num_workers = 0

        self.train_ratio = 0.9
        self.validation_ratio = 0.1
        self.test_ratio = 0.1

        self.dataset_settings()
        self.optim_settings()
        return

    def dataset_settings(self,):
        self.dataset = 'USC'
        self.cityMap = 'complete'        # complete, height
        self.sampling = 'exclusive' # random, exclusive



    def optim_settings(self,):
        self.lr = 1e-4
        self.lr_decay = 0.45
        self.step = 10

    def get_train_parameters(self,):
      return {'exp_name':self.exp_name,
        'batch_size':self.batch_size,
        'num_epochs':self.num_epochs,
        'lr':self.lr,
        'lr_decay':self.lr_decay,
        'step':self.step,
        'sampling':self.sampling}