bhuman/CompiledNN

Issues

• Support Conv1D and MaxPooling1D

  #22 opened 7 months ago
  0

• Onnx support

  #20 opened 3 years ago
  2

• Installation fails

  #19 opened 3 years ago
  1

• Fusing UInt8Input+BatchNormalization only works if number of channels is a multiple of 4

  #18 opened 4 years ago
  1

• Quantization

  #17 opened 4 years ago
  0

• DConv2D does not work for channel numbers that are not multiples of 4 or do not fit into SSE registers

  #16 opened 4 years ago
  0

• Fuse BatchNormalization followed by (D)Conv2D

  #15 opened 4 years ago
  0

• Implement Conv1x1

  #14 opened 4 years ago
  0

• Finish Im2Col implementation

  #13 opened 4 years ago
  0

• Segmentation Fault in Tensor.h:225

  #12 opened 4 years ago
  1

• Documentation

  #11 opened 4 years ago
  4

• GlobalPooling2D: support an arbitrarily large channel dimension

  #10 opened 5 years ago
  0

• Softmax: support multi-dimensional tensors and an arbitrary softmax axis

  #9 opened 5 years ago
  0

• Conv2D: biasOffset thing is not working

  #8 opened 5 years ago
  0

• Conv2D: canBeInplace-condition is wrong

  #7 opened 5 years ago
  0

• Testing

  #6 opened 5 years ago
  0

• ASSERT(input.dims(2) <= settings.xmmRegs() * 4) failed

  #5 opened 5 years ago
  2

• build failed

  #3 opened 5 years ago
  6

• 'Platform/BHAssert.h' file not found

  #2 opened 5 years ago
  2

• CompiledNN/Src/Model.cpp:300: ASSERT(result) failed

  #1 opened 5 years ago
  6