Hi,
It seems like the "top_10_similar_genes_sim" file is missing from the Harvard Dataverse data files, which is throwing an error in training since it is used in line 143 of train.py. Could you please upload this data file and the source code used to generate it?
Cheers,
Emily