Jovan Mitrevski

fix nondefault project name handling

7

# Description As discussed in #624, convert_from_keras_model with project_name set to nondefault value (e.g. mynewprojname) fails to build, because in build.tcl the variable ${mynewprojname} is not defined. This fixes the...

Add quantized sigmoid, fix quantized tanh for QKeras

14

This is an attempt to fix quantized tanh and add quantized sigmoid activation support for QKeras ingestion. In Quartus the tanh handling probably needs an optimizer to rename it to...

## Prerequisites Please make sure to check off these prerequisites before submitting a bug report. - [x] Test that the bug appears on the current version of the master branch....

QONNX ingestion for Vivado and Quartus

6

# Description This is the main development to ingest QONNX in hls4ml. It's a bit of a large PR, and will probably take a while to review and update as...

BatchNormalization scale and bias array size

2

The `normalize` function call in both `nnet_batchnorm.h` and `nnet_batchnorm_stream.h` has an interface of: ```c++ template void normalize( data_T data[CONFIG_T::n_in], res_T res[CONFIG_T::n_in], typename CONFIG_T::scale_t scale[CONFIG_T::n_in], typename CONFIG_T::bias_t bias[CONFIG_T::n_in] ) ``` However,...

For encoded convolution, add check for when min_width would have been larger than in_width

2

# Description In the encoded convolution, the case when the in_width is smaller than the nominal min_width needs to be handled as a special case. This attempts to fix this...

Update inplace variables to not lose connection when source node is optimized.

3

In the original implementation, inplace variables are created when initializing an object, which can run into problems if the source node is changed due to optimizers. This PR moves the...

fifo_depth_optimization_flow details

6

# Description I am opening this pull request more for discussion. If people agree with this change, the same change would also need to be made for the VivadoAccelator backend....

append applied_flows container before filling instead of after

2

# Description This change simply appends `applied_flows` to `ModelGraph._applied_flows` in the beginning, before it is filled, so that `ModelGraph._applied_flows` always has an accurate description of what has been applied. This...

Beginnings of the oneAPI backend

3

# Description This is the beginning of the oneAPI backend, and is just included here for easier sharing and discussion. It is not ready for merging. Currently only `io_parallel` Dense...