how to use customized loss function and modify trainning procedure in caffe

[ZMaple]

May i know if it is possible to use customized loss function in caffe? Also, if I want to change the formula of weight update in the backpropagation, is this allowed in caffe? Could you please give some hints about this? Thanks a lot!

[shelhamer]

Please ask questions on the caffe-users mailing list. As of the latest release we prefer to keep issues reserved for Caffe development. Thanks!

For how to develop a new layer, including a new loss layer, see https://github.com/BVLC/caffe/wiki/Development-Hints#developing-new-layers