search

jmiller656 / EDSR-Tensorflow

Tensorflow implementation of Enhanced Deep Residual Networks for Single Image Super-Resolution
MIT License
330 stars 107 forks source link

Is weight decay required? #27

Closed ZhenyF closed 5 years ago

ZhenyF commented 6 years ago

For L2 loss we need a weight decay to prevent over fitting.

Is it required to set a weight decay for L1 loss?

jmiller656 commented 5 years ago

I didn't use any normalization techniques in this implementation because they weren't mentioned in the paper