1.19.14

• Celu
  • Fixed a bug that caused very few elements to diverge to Nan, resulting in inconsistent output.

    onnx_tensor[0, 3, 1, 26]
    97.723495
    
    tf_transposed_tensor[0, 3, 1, 26]
    nan
    

  • poc.onnx.zip

TensorFlow operator combinations were reviewed.

onnx2tf -i poc.onnx -cotof

• There are discrepancies between the outputs of the ONNX model and converted TensorFlow models. #602

What's Changed

• pip env setup guide: Fix onnx package version & add tensorflow and protobuf by @Boulaouaney in #601
• Bug fix for Celu, onnxruntime==1.17.1 by @PINTO0309 in #603

New Contributors

• @Boulaouaney made their first contribution in #601

Full Changelog: 1.19.13...1.19.14

1.19.13

• Gather
  ONNX Gather changes output shape by removing the singled out dimension. When converting to TF using strided_slice, the extra dimension is kept.

  ONNX: [1,50,768] -> Gather (index=0) -> [1,768]
  TF: [1,50,768] -> StridedSlice (... ) -> [1,1,768]

  

What's Changed

• Gather - keep output dimensions by @svobora in #597

New Contributors

• @svobora made their first contribution in #597

Full Changelog: 1.19.12...1.19.13

1.19.12

What's Changed

• fix for float64 error by @khatami-mehrdad in #593
• Add type specification, v1.19.12 by @PINTO0309 in #594

New Contributors

• @khatami-mehrdad made their first contribution in #593

Full Changelog: 1.19.11...1.19.12

1.19.11

• MaxPoolWithArgmax
  • Support for 1D MaxPoolWithArgmax
  • However, TFLite does not support max_pool_with_argmax as standard.
  • UM_best_model.onnx.zip
    
• ValueError in MaxPool Layer during ONNX to TensorFlow Conversion with 3D Input #579

What's Changed

• Support for 1D MaxPoolWithArgmax by @PINTO0309 in #580

Full Changelog: 1.19.10...1.19.11

1.19.10

• Mul, Div, Sub, or Mod
  • Modified to suppress broadcast when Gemm is immediately followed by a scalar Mul, Div, Sub, or Mod.
  • This is a very specific workaround for the special specifications of NNAPI.
  • Address the issue of NNAPI not allowing multi-dimensional bias.
  • The problem is not reproducible for variable batch sizes.
  • The problem reproduces only when changing to a fixed batch size.
• pre_explicit_broadcast should not expend scalar tensor #573

What's Changed

• Modified to suppress broadcast when Gemm is immediately followed by a scalar Mul, Div, Sub, or Mod. by @PINTO0309 in #577

Full Changelog: 1.19.9...1.19.10

1.19.9

• Improved GatherElements conversion stability.
  • bodypix_mobilenet050_stride8_1x3x128x160.onnx.zip

What's Changed

• Improved GatherElements conversion stability. by @PINTO0309 in #576

Full Changelog: 1.19.8...1.19.9

1.19.8

• Gemm

  • Fixed Abort bug when converting models with --optimization_for_gpu_delegate.

    onnx	before tflite	after tflite

• pre_explicit_broadcast should not expend scalar tensor #573

What's Changed

• Fixed Abort bug when converting models with --optimization_for_gpu_delegate. by @PINTO0309 in #574

Full Changelog: 1.19.7...1.19.8

1.19.7

• Improved the display of auxiliary error messages regarding flatc installation instructions.
• Full int8 quantized model doesn't work when "flatc" is installed on the system #567

What's Changed

• improve flatc warning message by @spacycoder in #570

New Contributors

• @spacycoder made their first contribution in #570

Full Changelog: 1.19.6...1.19.7

1.19.6

• Support for STFT (Partially supported)
  1. add onnx2tf/onnx2tf/ops/STFT.py. This is the main implementation of the STFT op.
  2. provide example STFT onnx model

What's Changed

• Add STFT op onnx2tf support by @guanerdan in #566

New Contributors

• @guanerdan made their first contribution in #566

Full Changelog: 1.19.5...1.19.6

1.19.5

• Resize
  • Implement workaround for cases where the input tensor contains multiple undefined dimensions.
  • Addressed the issue of Resize aborting only when resizing by the scales parameter, not by the sizes parameter.
  • fp32.onnx.zip
    
• Failed to convert Real-ESRGAN x4-general (NCHW) #563

What's Changed

• Implement workaround for cases where the input tensor contains multiple undefined dimensions Resize by @PINTO0309 in #564

Full Changelog: 1.19.4...1.19.5