pre-processing code

[bingo619]

Hi can you please provide the code you used to pre-process the raw data?

and also the train.sh?

[haoliplus]

Sorry for such a long time before the reply. I have uploaded the preprocess programs. They should be no problem. But they may be not exactly the same with the ones we used before, since it has been a long time...

So if you have any question about the code, please leave me a message.

[bingo619]

Thanks.

I have managed to get the data ready for music dataset.

But the training losses for plstm and timelstm1 go to nan after a few epochs.

[bingo619]

Thanks.

I have managed to get the data ready for music dataset.

But the training losses for plstm and timelstm1 go to nan after a few epochs. Timelstm3’s loss won’t decrease.

Can you test?

[zyody]

@bingo619
The preprocess code for lastFM and readme.m file have already been improved. Now you can successfully run the code following the readme.m.

In addition, the new download link for the CiteULike dataset is also provided.