Issues

• Minimum python requirement not enforced, but seems to require python>3.10

  bug
  fastmachinelearning/qonnx#237 · Filter by author nickfraser opened on Mar 2, 2026

• hls4ml fpga_backend fails when using a Resize node with no ROI using #150 fix (Brevitas -> QONNX)

  bug
  fastmachinelearning/qonnx#197 · Filter by author The-Padi opened on Jun 24, 2025

• ExtractBiasFromConv is not working properly with quantized biases

  bug
  fastmachinelearning/qonnx#174 · Filter by author balditommaso opened on Feb 27, 2025
  5

• Overhaul or deprecate data layout inference

  enhancement
  fastmachinelearning/qonnx#173 · Filter by author maltanar opened on Feb 25, 2025
  1

  

• Overhaul shape inference for custom ops

  enhancement
  fastmachinelearning/qonnx#161 · Filter by author maltanar opened on Dec 15, 2024
  3

• Investigate onnx.compose as a potential MergeONNXModels replacement

  enhancement
  fastmachinelearning/qonnx#160 · Filter by author maltanar opened on Dec 15, 2024

• cannot import name 'from_keras' from 'qonnx.converters'

  bug
  fastmachinelearning/qonnx#154 · Filter by author AnouarITI opened on Nov 19, 2024
  1

• Resize node empty RoI input cause issue for shape inference

  bug
  fastmachinelearning/qonnx#150 · Filter by author nghielme opened on Nov 6, 2024
  2

  

• TypeError: __init__() takes from 3 to 5 positional arguments but 6 were given when using convert_from_onnx_model

  bug
  fastmachinelearning/qonnx#147 · Filter by author Sayanti17 opened on Oct 21, 2024

• Unsupported Operation Type Error: BipolarQuant in hls4ml Conversion

  bug
  fastmachinelearning/qonnx#146 · Filter by author Sayanti17 opened on Oct 21, 2024
  1

• Change NodeLocalTransformation Pool to use spawn instead of fork

  bug
  fastmachinelearning/qonnx#117 · Filter by author bwintermann opened on Jun 6, 2024

• channels_last transformation support for branching models

  enhancement
  fastmachinelearning/qonnx#100 · Filter by author nghielme opened on Feb 14, 2024
  1

• Multithreshold operator FINN-ONNX execution error after 1D scalar add operator is absorbed into it.

  bug
  fastmachinelearning/qonnx#98 · Filter by author shashwat1198 opened on Feb 8, 2024

• QKeras qconv2d test failure

  bug
  fastmachinelearning/qonnx#54 · Filter by author maltanar opened on May 8, 2023

• Changed behavior after BatchNormToAffine transformation

  bug
  fastmachinelearning/qonnx#50 · Filter by author auphelia opened on Mar 29, 2023

• Support QConv2DBatchnorm and QDenseBatchnorm layers

  enhancement
  fastmachinelearning/qonnx#10 · Filter by author jmduarte opened on Apr 22, 2022
  1

  

• ConvertToChannelsLastAndClean: New transpose nodes don't have names and intermediate tensors are not kept in place

  bug
  fastmachinelearning/qonnx#9 · Filter by author heborras opened on Nov 18, 2021
  2

  

• Transformations should return whether a model was modified or not

  enhancement
  fastmachinelearning/qonnx#6 · Filter by author heborras opened on Sep 1, 2021
  1