Q.Liu

I'm not sure what your point is. If you intended to use test-set servering as the val-set, simply change the val-folder to the test-folder.

Now I see, lets saying you have a training set with 100 images, and a test-set with 50 images, you can split the training set into e.g., train/val 80/20 randomly...

Yes, it's possible, just slightly need to change the code `split_train_val_test_sets` as following ( might be some bugs, you can further modify it). ``` # change DATASET ROOT to your...

test_dict is not used at all during training, you can safely deleted it if you want, and just return train_dict, and val_dict. I leave test_dict here just for furture modification...

the code was not written well and contains some confused names / redundant stuff / bugs without refactoring after completing the agriculture workshop. You need to choose the useful part...

ah, there is a bug, should move the three lines within else： ``` gts_all.append(gts.data.squeeze(0).cpu().numpy()) predictions = outputs.data.max(1)[1].squeeze(1).squeeze(0).cpu().numpy() predictions_all.append(predictions) ``` and the value of `save_rate` is used to control what percent...

yes, you are right. I'd refactor the code a bit when I get some free time. I remember the reason I appended all predictions together to compute metrics, it's because...