eric-haibin-lin opened a new issue #9363: Incorrect weight_decay implementation in AdaGrad URL: https://github.com/apache/incubator-mxnet/issues/9363 In Adagrad, weight_decay should be applied to grad before clipping the gradient/updating the state/weight. https://github.com/apache/incubator-mxnet/blob/master/python/mxnet/optimizer.py#L856-L858
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services
