tqchen edited a comment on issue #17097: [mxnet 2.0][item 10.1] MXNet Imperative Op Invocation Overhead URL: https://github.com/apache/incubator-mxnet/issues/17097#issuecomment-568046278 @sxjscience here are some quick thoughts (of course passing pyobject kind of "solves" the problem, so I am discussing the wrapping that can be done through the tvm ffi). - concatenate seems can be achieved through the python side wrapping ```concat-> concat_internal(*args)``` - return list value can be addressed by introducing a Tuple object to tvm runtime, we recently did that. - numpy native data structure: one way to deal with it is to convert numpy native structure to something related to DLTensor (e.g. dlpack). It may not be as fast as directly treat it as a scalar though The py_slice is the most tricky case, my guess is that it could be accelerated through a cython layer that translate the slice into flattened representation, of course it is not too ideal and pybind maybe better for this case if we only want to handle it through c++.
---------------------------------------------------------------- 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
