Unable to reproduce results

[HarrisonWilde]

Hi, I am going through the repository's codebase and so far cannot reproduce any of the results in the paper. All of the AUC evaluations are lower for me by something in the range of 0.01 - 0.05. I am not sure what more information to provide as I have changed nothing in the repo, but wondered if you had any idea as to why this might be the case? I am using the specified optimal parameters in pretrained_models.md, etc etc.