need python-seed for replication

[hojjatkarami]

Hello,

could you please tell me the --python-seed argument that you have used?

I would like to replicate discriminative and predictive scores for stock and energy datasets. I tried the following command but the results are different:

python main_2012.py --dataset ./data/physio_data/full2012.pkl --task-name test

Thanks

[acphile]

Hi, I do not save the seed and set the seed of Pytorch. The command you use is not for stock and energy datasets. You should use the following command for stock dataset

python main_stock.py --dataset ./data/stock.pkl --task-name stock

For energy dataset, you should additionally add --hidden-dim 112 --embed-dim 448 --noise-dim 448 --layers 3