sxjscience edited a comment on pull request #18531: URL: https://github.com/apache/incubator-mxnet/pull/18531#issuecomment-643361273
@eric-haibin-lin @fhieber I’ve written a new loss layer in the numpy API: https://github.com/dmlc/gluon-nlp/blob/b714eaccc67619d7bdcbd1574d30be87d9c73f0c/src/gluonnlp/loss.py#L4 The speed + memory cost is better than calculating the cross-entropy between softmax and smoothed label (the GluonNLP's original implementation). ---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: [email protected]
