wkcn commented on issue #16057: add operator `cast_like`. URL: https://github.com/apache/incubator-mxnet/pull/16057#issuecomment-526870996 I think ‘cast’ and ‘cast_like’ should pass the gradient, although it will loss the precision. For example, I have an operator which need a higher precision, like float64. However, the input/output dtype of most operators is float32. I need the cast operator, and pass the gradient.
---------------------------------------------------------------- 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] With regards, Apache Git Services
