I’m wondering how to use mx.nd operations in the forward pass of Gluon to implement cross product style architectures (e.g. bilinear models) and that need to look at pairs of vectors I was trying to implement a Relational Network: https://arxiv.org/abs/1706.01427 Relevant implementation in PyTorch…

Cross product style architectures with Gluon

non_linearity November 14, 2017, 10:06pm 3

Sure - will do - @piiswrong, are you saying that the syntax in the first version is correct?

Topic		Replies	Views
Hybridized RNN State Initialization Error Gluon	1	409	June 3, 2021
mx.nd.contrib.BilinearResize2D (Inside HybridBlock) Gluon	3	1632	July 24, 2018
Problem when hybridizing with sparse dot Gluon	3	793	September 6, 2018
Gluon models with branches	5	1097	June 28, 2018
Converting Gluon Neural Style to C++ Feedforward Discussion	6	1690	March 7, 2018