Why is the result so much better than the original?
Opened this issue · 2 comments
lmq135831 commented
In the case of the music dataset AUC, the original was close to 80%, but this version gets to 85%,
hh-hua commented
+1
Amethystium-321E1 commented
+1