question about retain_perturbed.json in datasets locuslab/TOFU

[wtma1999]

Thank you for your work on machine unlearning! I noticed that retain_perturbed.json in datasets only contains 400 pieces of data, about 20 authors. Then when calculating the Truth Ratio for Retain Set, can we only calculate these 400 pieces of data? In my opinion, it requires calculating the Truth Ratio for each of the 20 authors in the Retain Set. Looking forward to your reply！