Closed rowedenny closed 2 years ago
Thanks for the question. Actually the "preprocess_data" is mean to be a function for preprocessing the input data before creating the input_feed, so all train/valid/test data should be preprocessed with the same "preprocess_data" function to make the models work. Here we use the "preprocess_data" function of train_input_feed to preprocess all data.
In fact, we haven't implemented any "preprocess_data" function yet (as can be seen in base_input_feed.py). In the future, we may refactor the code to create a separate data preprocessing class to avoid confusion.
I see and thanks for the clarification.
Hi,
Thanks for open source this amazing work.
I have a question for the current example in main.py. Specifcially, for the following code,
https://github.com/ULTR-Community/ULTRA_pytorch/blob/ec4fe329e4239b588a940cb4bcdd6a321aade679/main.py#L92
If I understand correctly, is this should be
ultra.utils.find_class(exp_settings['valid_input_feed']).preprocess_data(valid_set, exp_settings['valid_input_hparams'], exp_settings)
since for the valid set, we would like to evaluate based on the human annotation for LTR, thus the key should be
valid_input_feed
instead oftrain_input_feed
, and for the same reason, the key for the second parameter shall be altered as well.In addition, should be the
test_set
need to altered correspondingly.Thanks for your time and consideration.