I had known N in dataset.py is the len(yc_part) in preprocessing_data.py . But I have not known how I can get the min_ts and max_ts in dataset.py if ds == 'yc64' or ds == 'test': self.N = 144527 self.reader = self.yc64 self.min_ts, self.max_ts = 1411604904, 1412017199 elif ds == 'yc4': self.N = 2312432 self.reader = self.yc4 self.min_ts, self.max_ts = 1408507486, 1412017199 elif ds == 'yc': self.N = 9249729 self.reader = self.yc self.min_ts, self.max_ts = 1396292400, 1412017199 I think they are buys.dat's max_ts and min_ts but it's not correlative max, min, dt 1396292731 1412015712 self.min_ts, self.max_ts = 1396292400, 1412017199

I want to figure out it . I need some help

Thanks for your great work and the share of code. in the dataset.py
line 119: if len(click_history) >= 5 and len(share_history) >= 1: Could you please tell me that what this "5" means ? The max_seq_len is set as 3 for yoochoose dataset according to the paper but this constraint of min_click_history haven't been noticed.

tensorflow = 2.2.0 python = 3.7 when I run, prog run till base.save While run tf.compat.v1.train.Saver(), it hint

File "main.py", line 131, in <module> main() File "main.py", line 113, in main T.train() File "/home/yang/Documents/paper/refer/MGNN-SPred-master/Train.py", line 40, in train self.model.save() File "/home/yang/Documents/paper/refer/MGNN-SPred-master/models/base.py", line 116, in save self.saver.save(self.sess, name) File "/home/yang/anaconda3/envs/tsorflw/lib/python3.7/site-packages/tensorflow/python/training/saver.py", line 1180, in save checkpoint_file, build_save=True, build_restore=False) File "/home/yang/anaconda3/envs/tsorflw/lib/python3.7/site-packages/tensorflow/python/training/saver.py", line 852, in _build_eager checkpoint_path, build_save=build_save, build_restore=build_restore) File "/home/yang/anaconda3/envs/tsorflw/lib/python3.7/site-packages/tensorflow/python/training/saver.py", line 886, in _build build_restore=build_restore) File "/home/yang/anaconda3/envs/tsorflw/lib/python3.7/site-packages/tensorflow/python/training/saver.py", line 490, in _build_internal names_to_saveables) File "/home/yang/anaconda3/envs/tsorflw/lib/python3.7/site-packages/tensorflow/python/training/saving/saveable_object_util.py", line 360, in validate_and_slice_inputs for converted_saveable_object in saveable_objects_for_op(op, name): File "/home/yang/anaconda3/envs/tsorflw/lib/python3.7/site-packages/tensorflow/python/training/saving/saveable_object_util.py", line 209, in saveable_objects_for_op "executing eagerly, got type: %s." % type(op)) ValueError: Can only save/restore ResourceVariables when executing eagerly, got type: <class 'tensorflow.python.framework.ops.Tensor'>.

I change some tf1 code to tf2. Is it wrong? Where? But I can't figure out where the problem is. Plz help me

Thanks for your great work and inspiration. Regarding the implementation of the baseline in your paper, I would like to ask how can I get the results consistent with your paper. Take STAMP as an example, we keep the dim, random seed and batch size as the same as your implementation of your method, and use gating function to fuse two sequence representations. However, my result is much worse than yours.
I think that’s because of the implementation differences. I have sent an email to you but maybe it went to the dump unfortunately. Hopefully I can receive your kindly reply.