The issue of the results of RQ2

[ichont]

When I try to run this comman：
python train.py --task 'tt' --model 'PolarGate' --device 0 --batch_size 256 --eval_step 1 --split_file 0.05-0.05-0.9 --layer_num 9 --feature_type 'one-hot' --in_dim 3
I have repeated the experiment multiple times but still cannot resolve the issue.
The results of Polargate in the experiment 'Experimental Results on TTDP Task' consistently fail to match the results presented in your paper, with the MAE being 0.1 to 0.2 higher than your reported results. Could there be any issues with the code and data you provided? Thank you for your guidance and response!