RLFN icon indicating copy to clipboard operation
RLFN copied to clipboard

Published 20 hours ago • verified publisher icon bytedance

How many iterations of the model can you get the best performance in the benchmark？

Open DyCheer opened this issue 2 years ago • 1 comments

Benchmark like Set5, Set14 and so on. According to the paper, the initial learning rate is 5e-4, and is halved every 2e5 iterations, but the total number of iterations is not specified.

DyCheer avatar Jun 10 '22 07:06 DyCheer

We train 1000 epochs at each stage and halve learning rate every 200 epoch.

fyan111 avatar Jun 13 '22 02:06 fyan111