search

Roche / BalancedLossNLP

Other
125 stars 24 forks source link

It seems that the generalization of DBLoss is enhanced, but the robustness is reduced? #9

Closed yongzhuo closed 1 year ago

yongzhuo commented 1 year ago

It seems that the generalization of DBLoss is enhanced, but the robustness is reduced?

blessu commented 1 year ago

I did not test, but it is possible that the convergence of the model is slower.