Jovan Mitrevski comments

Results 133 comments of


                                            Jovan Mitrevski

Add quantized sigmoid, fix quantized tanh for QKeras

I was able to manually reproduce it. It is in the vivado_accelerator_writer not liking the pre-commit changes: ```python elif 'void myproject(' in line: newline = f'void {model.config.get_project_name()}_axi(\n' ``` This basically...

Vitis HLS backend

When working with CNNs I get lots of warnings of the type: ``` WARNING: [HLS 214-113] Either use an argument of the function or declare the variable inside the dataflow...

BatchNormalization scale and bias array size

This was fixed in #749 for Quartus, so closing.

Relu merge optimizer pass

Any more news on this?

Relu merge optimizer pass

What is the plan for this in general? Is this where we'll attack with the new code generation framework?

QONNX ingestion for Vivado and Quartus

This is replaced by #832

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

This should now also fix the same issue with Conv2D.

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

@vloncar, are you happy with the changes here as they are?

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

Is there anything left to do here before it's merged?

multi-task learning using HLS4ML

For larger models with CNNs you should use `io_type="io_stream"`. That also supports skip stops. For CNNs, `io_type="io_parallel"` is only for small models without skip connections, or MLPs.