multilingual-modeling icon indicating copy to clipboard operation
multilingual-modeling copied to clipboard

Published 20 hours ago • verified publisher icon bigscience-workshop

Inconsistent Evaluation Results

Open yongzx opened this issue 3 years ago • 0 comments

I am getting different results by running training/eval together and separately. Rerunning evaluation after training (by removing --do_train) gives me a better result than running training+eval together.

yongzx avatar Jun 22 '22 12:06 yongzx