Open wildwolff opened 1 year ago
I tried to replicate the experiment following the default settings, and the results I got was similar with those reported in the paper. However, the eval loss increased as the training goes. I am confused why this happens.
I tried to replicate the experiment following the default settings, and the results I got was similar with those reported in the paper. However, the eval loss increased as the training goes. I am confused why this happens.