How to divide train, val, and test under the new train.txt?

[Chengjian-Li]

Thanks to the authors for publishing the dataset and for code maintenance. When reproducing the author's published code, it was found that the new train.txt contained 7772 samples. In this case train.txt there are validation sets and test sets, where does the author code reflect the data division?

[Chengjian-Li]

@tr3e @Godheritage @0beihai0 Thanks to the authors for publishing the dataset and code maintenance. When reproducing the author's published code, it was found that the new train.txt contained 7772 samples. In this case train.txt there are validation sets and test sets, where does the author code reflect the data division?