Vladimir

Yes! We hope to merge it soon into the main branch. The latest development is in https://github.com/dgburnette/hls4ml

You would need to implement a `depthwise_product(...)` for resource strategy in `nnet_sepconv_stream.h` that currently is hardcoded to fail an assertion. This product is not the same as matrix-vector multiplication that...

You really need to provide more information if you want assistance. Please open a [bug report](https://github.com/fastmachinelearning/hls4ml/issues/new?assignees=&labels=bug&projects=&template=bug_report.md&title=) and follow the template.

Hi Lorenzo, This is really interesting, we would definitely welcome an implementation of instance normalization based on the work you already started. From your description, the only thing you're missing...

One design goal that we mostly abandoned but can come back to is to completely modularize hls4ml such that only the core internals are available by default and everything else...

I'm wondering if we should also have an extension API handle adding support for new hardware or new drivers because now one has to make manual changes to the codebase...

Dense on 3D input is supported. Behind the scenes, it will result in exactly the same code as in your 2nd alternative, i.e., pointwise Conv2D. The reason you don't see...

You can see in the report exactly what is the DSP used for. When using the resource strategy, the weights will be stored in BRAM and there will be accounting...

The statement from the paper refers to that particular model, not in general. When using the resource strategy, the weights will be partitioned with an `ARRAY_RESHAPE` pragma (read about it...

You can use extension API in VivadoAccelerator backend. If you say that your model doesn't work in Vivado but works in VivadoAccelerator you're probably doing something wrong since there should...