Open rockywind opened 3 years ago
Sorry for the bugs, you can delete the related row (line 11 and 26). Or update the trainer_helper.py from our git.
But the self.val_decorator is call in the following code.
loss, stat_dict = self.val_decorator(self.model, batch_data, self.cfg['decorator'])
We do not call the function eval_one_epoch_1(), this function is my debug code.
When I run
python ../../tools/train_val.py --config config_val.yaml
, I get the error as follow.