I think shifting by `logsumexp` is reasonable. When the inputs are already from log_softmax, `logsumexp(x)` will be 0 so the behavior is the same as the current version. Basically, `logsumexp(log_softmax(x)) == 0`
-- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/apache/incubator-mxnet/issues/19722#issuecomment-751855108
