Hi Malcolm,
Thank you for your great work on RNBERT — it’s a very impressive and inspiring project!
I noticed that the dataset provided in dataset.zip only contains the salami-sliced version of the data.
I was wondering if it would be possible for you to provide (or point me to how I could obtain) a version of the dataset that is not salami-sliced, while keeping all the other features unchanged?
It would be very helpful for some experiments I’m currently working on.
Thanks in advance!
Best regards,
Hewei