Line 151 of train.py doesn't do anything. The action is store_true, so if the flag is used it is set to True. But the default value is also True, so it's just always set to True. Basically it only serves to make explicit what you're already doing.
parser.add_argument("--enable_val_pplx", action="store_true",
default=True,
help="Calculate and report smoothed validation pplx\
alongside the Keras objective function loss.\
(default=true)")
Line 151 of
train.pydoesn't do anything. The action isstore_true, so if the flag is used it is set to True. But the default value is also True, so it's just always set to True. Basically it only serves to make explicit what you're already doing.