Can I pass a sparse tensor as weighes to Funcional conv2d?

1

Can I pass a sparse tensor as weighes to Funcional conv2d?

2

We don’t support that yet. Feel free to open a feature request at the github repo (please check to see if there’s an existing one already!)