Jan-Frederik Schulte

Add RNN support for Pytorch

3

Adds support for RNN layers (GRU, LSTM, RNN) to the pytorch parser. Caveat: We currently lack implementation for `getitem` operations, so we can currently not return the hidden state after...

Add support for parsing simple brevitas layers as part of pytorch models

14

This PR adds support for the parsing of simple brevitas layers (QuantLinear, QuantActivation, QuantConv1D, QuantConv2D) to the pytorch parser. More complex models will still have to go through QONNX, but...

Fixes issues with softmax activation in pytorch parser reported in https://github.com/fastmachinelearning/hls4ml/issues/1083 and adds softmax activation to the pytest. I also added a one-line change to fix https://github.com/fastmachinelearning/hls4ml/issues/1074 while I'm messing...

Adds simple pytests for the brevitas -> QONNX -> hls4ml workflow. Tested are models using both power-of-2 and non-power-of-2 scales. For the latter, the accuracy is generally poor, driven by...

We currently don't install hls4ml with the `qkeras` option, which is needed for part 3 and 4 of the tutorial. Fixed here.