eric-haibin-lin commented on issue #10388: [MXNET-265] Update optimizer doc to
clarify wd behaviors
URL: https://github.com/apache/incubator-mxnet/pull/10388#issuecomment-379615395
I'm still hesitated to change all these behaviors since they're will break
existing training scripts..
eric-haibin-lin commented on issue #10388: [MXNET-265] Update optimizer doc to
clarify wd behaviors
URL: https://github.com/apache/incubator-mxnet/pull/10388#issuecomment-379401680
Ah, good point. I'll change that, too.
eric-haibin-lin commented on issue #10388: [MXNET-265] Update optimizer doc to
clarify wd behaviors
URL: https://github.com/apache/incubator-mxnet/pull/10388#issuecomment-379295348
Merged wd term before clipping grad for AdaGrad because otherwise it's very
likely optimization with Adagrad
eric-haibin-lin commented on issue #10388: [MXNET-265] Update optimizer doc to
clarify wd behaviors
URL: https://github.com/apache/incubator-mxnet/pull/10388#issuecomment-379295348
Merged wd term before clipping grad for AdaGrad.
eric-haibin-lin commented on issue #10388: [MXNET-265] Update optimizer doc to
clarify wd behaviors
URL: https://github.com/apache/incubator-mxnet/pull/10388#issuecomment-378700149
@ZiyueHuang @szhengac @sxjscience could you help review?
