This PR cleans up some of the QKeras2QONNX converter code and inserts BinaryQuant nodes instead of Quant nodes for qkeras.binary instances. This PR should follow PR: #139.
As this changes the nodes created by the converter it could affect downstream users.
Questions:
In onnx.py in qact_handler function there was a channel_order_rewriters._to_channel_first_handler(ctx, quant_act_node). I removed this and the tests still passed, however I am not sure what that was doing there. Does anyone know?
In qonnx/custom_op/general/quant.py in function quant there is a "hack" for bitwidth==1 and signed==True, with the following comment.
# BUG: 1-bit Quant ops currently not exported correctly, manually convert to bipolar values
So, should this PR request delete this workaround? or does anyone still depend on this?
This PR cleans up some of the QKeras2QONNX converter code and inserts BinaryQuant nodes instead of Quant nodes for
qkeras.binaryinstances. This PR should follow PR: #139.As this changes the nodes created by the converter it could affect downstream users.
Questions:
channel_order_rewriters._to_channel_first_handler(ctx, quant_act_node). I removed this and the tests still passed, however I am not sure what that was doing there. Does anyone know?# BUG: 1-bit Quant ops currently not exported correctly, manually convert to bipolar valuesSo, should this PR request delete this workaround? or does anyone still depend on this?