What should I do if I don't have validation data and test data?

    @pl.data_loader
    def val_dataloader(self):
        # I don't need it.

    @pl.data_loader
    def test_dataloader(self):
        # I don't need it.

    def validation_step(self, batch, batch_nb):
         # I don't need it.

    def validation_end(self, outputs):
         # I don't need it.

[williamFalcon]

@Illcat made optional in this upcoming PR #95. check back in a few hours and run from master

[williamFalcon]

install master branch and try again. val and test are optional

Sometimes I just want to load a trained model to predict unlabeled data.
Is there no corresponding test_step method?

This means that trains are also optional.

[williamFalcon]

@Illcat. there is no test step yet but it’s alloted for this week. if you need it before you can load the model using the instructions in the docs and do your own test loop right now. otherwise, just check back in later in the week when it’s live.

[williamFalcon]

i don’t understand what you mean by trains are optional.