1.25.12

• Flatten
  • Improved handling when axis attribute is not defined and the batch size of the first dimension is undefined.
  • ONNX
    

    wget https://github.com/PINTO0309/onnx2tf/releases/download/0.0.2/resnet18-v1-7.onnx
    onnx2tf -i resnet18-v1-7.onnx
    
    ls -lh saved_model/
    
    assets
    fingerprint.pb
    resnet18-v1-7_float16.tflite
    resnet18-v1-7_float32.tflite
    saved_model.pb
    variables
    
    TF_CPP_MIN_LOG_LEVEL=3 \
    saved_model_cli show \
    --dir saved_model \
    --signature_def serving_default \
    --tag_set serve
    
    The given SavedModel SignatureDef contains the following input(s):
      inputs['data'] tensor_info:
          dtype: DT_FLOAT
          shape: (-1, 224, 224, 3)
          name: serving_default_data:0
    The given SavedModel SignatureDef contains the following output(s):
      outputs['output_0'] tensor_info:
          dtype: DT_FLOAT
          shape: (-1, 1000)
          name: PartitionedCall:0
    Method name is: tensorflow/serving/predict

  • TFLite
    

What's Changed

• Improved handling when axis attribute is not defined and the batch size of the first dimension is undefined. by @PINTO0309 in #692

Full Changelog: 1.25.11...1.25.12

1.25.11

• BatchNormalization
  • Improved the conversion stability of BatchNormalization.

What's Changed

• Improved the conversion stability of BatchNormalization. by @PINTO0309 in #690

Full Changelog: 1.25.10...1.25.11

1.25.10

• Flatten
  • Addressed the issue of missing conversions when multi-dimensional flattening is performed and the batch size of the first dimension is an undefined dimension.

    ONNX	TFLite

• [EfficientNetV2_m] Output size is constant not variable #688

What's Changed

• Addressed the issue of missing conversions when multi-dimensional flattening is performed and the batch size of the first dimension is an undefined dimension. by @PINTO0309 in #689

Full Changelog: 1.25.9...1.25.10

1.25.9

• Add, Sub
  • https://huggingface.co/onnx-community/metric3d-vit-small/blob/main/onnx/model.onnx
  • metric3d-vit-small.onnx
  • Fixed a bug in the optimization process for arithmetic operations. The error of the final output was less than 1e-4.
    
• A bug in the optimization process for the y = (200 - x) - 200 operation caused an incorrect Sub merge operation to be performed.
• Model gives inaccurate results post conversion to tflite #685

What's Changed

• Fixed a bug in the optimization process for arithmetic operations by @PINTO0309 in #686

Full Changelog: 1.25.8...1.25.9

1.25.8

• Shape
  • Fixed problem of being stuck in an infinite loop.
• An input to an ADD node keeps getting casted to float32 despite being float16 in the onnx file causing issues with the ADD op #681

What's Changed

• Fixed problem of being stuck in an infinite loop by @PINTO0309 in #682

Full Changelog: 1.25.7...1.25.8

1.25.7

• Expand, BatchNormalization, Gather
  • Automatic accuracy compensation.
  • Added the ability to automatically compensate for accuracy degradation due to dimensional transposition errors.
• AveragePool

  • Only very few edge cases are supported.

  • The dynamic tensor AveragePool is difficult to replace exactly with TensorFlow's AveragePooling.

  • I have fixed and released the critical problems except for AveragePool, but AveragePool (with ceil_mode=1) with dynamic tensor as input is extremely difficult to fix due to compatibility issues with TensorFlow.

  • The problem is that the error was not occurring in the AveragePool (with ceil_mode=1) where the conversion error should have occurred, and the latest onnx2tf should now generate a conversion error in the AveragePool (with ceil_mode=1).

    INFO: 39 / 1464
    INFO: onnx_op_type: AveragePool onnx_op_name: wa/xvector/block1/tdnnd1/cam_layer/AveragePool
    INFO:  input_name.1: wa/xvector/block1/tdnnd1/nonlinear2/relu/Relu_output_0 shape: [1, 128, 'unk__71'] dtype: float32
    INFO:  output_name.1: wa/xvector/block1/tdnnd1/cam_layer/AveragePool_output_0 shape: [1, 128, 'unk__77'] dtype: float32
    ERROR: The trace log is below.
    Traceback (most recent call last):
      File "/home/xxxxx/git/onnx2tf/onnx2tf/utils/common_functions.py", line 312, in print_wrapper_func
        result = func(*args, **kwargs)
      File "/home/xxxxx/git/onnx2tf/onnx2tf/utils/common_functions.py", line 385, in inverted_operation_enable_disable_wrapper_func
        result = func(*args, **kwargs)
      File "/home/xxxxx/git/onnx2tf/onnx2tf/utils/common_functions.py", line 55, in get_replacement_parameter_wrapper_func
        func(*args, **kwargs)
      File "/home/xxxxx/git/onnx2tf/onnx2tf/ops/AveragePool.py", line 171, in make_node
        output_spatial_shape = [
      File "/home/xxxxx/git/onnx2tf/onnx2tf/ops/AveragePool.py", line 172, in <listcomp>
        func((i + pb + pe - d * (k - 1) - 1) / s + 1)
    TypeError: unsupported operand type(s) for +: 'NoneType' and 'int'
    
    ERROR: input_onnx_file_path: ../cam++_vin.onnx
    ERROR: onnx_op_name: wa/xvector/block1/tdnnd1/cam_layer/AveragePool
    ERROR: Read this and deal with it. https://github.com/PINTO0309/onnx2tf#parameter-replacement
    ERROR: Alternatively, if the input OP has a dynamic dimension, use the -b or -ois option to rewrite it to a static shape and try again.
    ERROR: If the input OP of ONNX before conversion is NHWC or an irregular channel arrangement other than NCHW, use the -kt or -kat option.
    ERROR: Also, for models that include NonMaxSuppression in the post-processing, try the -onwdt option.
    

• Unable to convert a model with 3d input shape of dynamic length into tflite int8 format #673

What's Changed

• Automatic accuracy compensation Expand, BatchNormalization, Gather by @PINTO0309 in #675

Full Changelog: 1.25.6...1.25.7

1.25.6

• Concat

  • Bug fix for Dynamic Resize optimization pattern.
    https://github.com/yakhyo/face-parsing

    ONNX	TFLite

    

What's Changed

• Bug fix for Resize optimization pattern by @PINTO0309 in #672

Full Changelog: 1.25.5...1.25.6

1.25.5

• Transpose
  • Fixed NHWC flag judgment bug in Transpose of ViT for 3D tensor

What's Changed

• Fixed NHWC flag judgment bug in Transpose of ViT for 3D tensor by @PINTO0309 in #671

Full Changelog: 1.25.4...1.25.5

1.25.4

• Addition of automatic INT8 calibration process for RGBA 4-channel images.
• It should be noted that the MS-COCO image set does not include an alpha channel, so this auto-calibration does not allow for decent quantization.
• There is no correlation between the channel being 4 and the input data being an image. Therefore, if the model to be converted is data other than a 4-channel image, automatic calibration should not be used.
  • The following means and standard deviations are used as fixed values.

    mean = np.asarray([[[[0.485, 0.456, 0.406, 0.000]]]], dtype=np.float32)
    std = np.asarray([[[[0.229, 0.224, 0.225, 1.000]]]], dtype=np.float32)

  • Also, use a fixed value of 0.5 for the alpha channel of the calibration data.

    new_element_array = np.full((*calib_data.shape[:-1], 1), 0.500, dtype=np.float32)

• magic_touch.onnx.zip
  
• [TODO] Add 4 channels of image data to the sample data for quantization #411

What's Changed

• Addition of automatic INT8 calibration process for RGBA 4-channel images by @PINTO0309 in #670

Full Changelog: 1.25.3...1.25.4

1.25.3

• Improvements to ScatterND.

• Improved conversion stability for Transpose -> Softmax -> Transpose combinations.

• Improved error message regarding OP name error when GroupConvolution is included.

• Change -osd option to True by default.

  

  ERROR: Generation of saved_model failed because the OP name does not match the following pattern. ^[A-Za-z0-9.][A-Za-z0-9_.\\/>-]*$
  ERROR: /model.22/cv2.2/cv2.2.2/Conv/kernel
  ERROR: Please convert again with the `-osd` or `--output_signaturedefs` option.
  

What's Changed

• Improvements to ScatterND, Transpose -> Softmax -> Transpose combinations by @PINTO0309 in #669

Full Changelog: 1.25.2...1.25.3