Cannot reproduce evaluation results correctly.

[Jiaqi-zhang]

Thank you very much for your work and the new dataset!
I am trying to reproduce your results with your given model checkpoints/intergen.ckpt and for some reason don't get the same results.

========== Evaluating MultiModality ==========
!!! DONE !!!
========== MM Distance Summary ==========
---> [ground truth] Mean: 3.7852 CInterval: 0.0007
========== R_precision Summary ==========
---> [ground truth](top 1) Mean: 0.4224 CInt: 0.0041;(top 2) Mean: 0.6060 CInt: 0.0041;(top 3) Mean: 0.7043 CInt: 0.0035;
========== FID Summary ==========
---> [ground truth] Mean: 0.2930 CInterval: 0.0057
========== Diversity Summary ==========
---> [ground truth] Mean: 7.7490 CInterval: 0.0275
========== MultiModality Summary ==========

Compared with the R-precision (0.452, 0.610, 0.701) mentioned in the paper, my result is (0.4224, 0.606, 0.7043).

Thank you!