_C.RANDOM_ACTIONS = False
_C.ONE_ACTION = False
_C.SEMANTIC_REWARDS = False
_C.DETECTION_REWARDS = False
_C.REMOVE_BEFORE_REWARD = False
_C.USE_INVERSE_ACTIONS = False
_C.VALUE_LEARNING = False
_C.PREVIOUS_IMAGES = False
_C.GAMMA = 0.9
_C.BOOTSTRAP = False
_C.LINEAR = False
_C.LEARNING_RATE = 1e-3
_C.NUM_STEPS = int(1e5)
_C.TARGET_UPDATE_INTERVAL = int(8e3)
_C.CHECKPOINT_INTERVAL = int(2e3)
_C.ACTION_HIDDEN_LAYERS = 1
_C.GUMBEL_TEMP = 0.1
# use confidence from detector as reward instead of binary reward
_C.CONFIDENCE_REWARD = False
_C.DISTRIBUTIONAL = False
_C.KL_BACKWARDS = False
_C.LOG_SIGMA = False
_C.VISUALIZATION_DATA_ROOT = ""
def get_cfg_defaults():
"""Get a yacs CfgNode object with default values for my_project."""
# Return a clone so that the defaults will not be altered
# This is for the "local variable" use pattern
return _C.clone()
