I run the sine code and print out the outerloop updated weight without multiply the outerstepsize and the nomal SGD weight, and they are same, even I set the innerepochs bigger than 1.
and form the reptile algorithm, we can see that
,so the only difference is the out loop multiply an epsilon, it is only a learning rate defferent with SGD.
I run the sine code and print out the outerloop updated weight without multiply the![CodeCogsEqn](https://user-images.githubusercontent.com/11084708/56708395-1db52980-674f-11e9-87ad-bb771c76ab34.gif)
outerstepsize
and the nomal SGD weight, and they are same, even I set theinnerepochs
bigger than 1. and form the reptile algorithm, we can see that,so the only difference is the out loop multiply an
epsilon
, it is only a learning rate defferent with SGD.