ConvertOnnxModel
2018-07-19..2021-04-25 Dwayne Robinson (FDwR)

• Converts a binary ONNX model file to text (which can be edited in any simple text editor) and vice versa.
• Converts an ONNX tensor protobuf to text/CSV/PNG/NPY and vice versa.
• Generates output tensor values (ones, zeros, iota series, random).
• Exports model tensors to directory of tensor files.

Usage

ConvertOnnxModel.exe [options] inputFilename [outputFilename]

Example usage

• Model to model

  • Model from ONNX binary protobuf format to prototxt format
    ConvertOnnxModel input.onnx output.prototxt
  • Model in prototxt text format to binary protobuf ONNX format
    ConvertOnnxModel input.prototxt output.onnx
  • Model from ONNX binary protobuf back to binary with zeroed tensor values
    ConvertOnnxModel -zeromodelvalues input.onnx output.onnx

• Model to multiple tensors

  • Model from ONNX binary protobuf format to directory of NumPy tensors
    ConvertOnnxModel resnet50.onnx x:\resnet_*.npy
  • Model from ONNX binary protobuf format to directory of raw data files
    ConvertOnnxModel squeezenet.onnx z:\folder\*_weight.dat

• Tensor to tensor

  • Tensor from ONNX binary protobuf to comma separated values
    ConvertOnnxModel input.onnxtensor output.csv
  • Tensor from ONNX binary protobuf (.pb) to image file
    ConvertOnnxModel -tensor input.pb output.png
  • Tensor data as comma separated values to raw binary file
    ConvertOnnxModel -datatype uint8 -dimensions 224,224 Foo.csv Foo.dat
  • Tensor from NumPy array format to protobuf ONNX binary format
    ConvertOnnxModel input.npy output.onnxtensor

• Tensor from generated randomness to ONNX binary protobuf format
  ConvertOnnxModel -dimensions 3,4 -datatype float16 generate(random,1,24) output.onnxtensor

Parameters

• input/output files - graph (onnx/pb/text) or tensor (onnxtensor/npy/pb/text/csv/dat).
• -tensor - specifies the input file is a tensor (only needed if ambiguous file type like .pb)
• -graph - specifies the input file is a model (only needed if ambiguous file type like .pb)
• -dimensions - explicit tensor dimensions for .csv or .dat file. Defaults to 1D element count from source data.
• -datatype - tensor element type (float16,float32,float64,int8,uint8,int16,uint16,int32,uint32,int64,uint64,bool8). This isn't usually needed unless reading from raw data.
• -zeromodelvalues - zero any tensor values in model (clears model initializer weights - useful for sharing confidential models without revealing trained results) except tiny 1D tensors needed for shapes.
• -row - single row or range for .csv.
• -column - single column or range for .csv.

File Types

• Model file types:
  • .onnx - Open Neural Exchange model protobuf
  • .pb - Google Protobuf (with -graph)
  • .txt/.prototxt - Protobuf text
• Tensor file types:
  • .onnxtensor - Open Neural Exchange tensor
  • .pb - Google Protobuf (with -tensor)
  • .csv - Comma Separate Values (no dimensions, just data)
  • .png - Image (Portable Network Graphics)
  • .jpg - Image (Joint Photographic Experts Group)
  • .npy - NumPyArray single tensor
  • .dat/.bin - Raw binary data (no header, just contiguous array elements)
  • generate() - Generator tensor input pseudo filename:
    • generate(ones) - all ones. [1,1,1,1...]
    • generate(zeros) - all zeros [0,0,0,0...]
    • generate(values,3) - specific value [3,3,3,3...]
    • generate(iota,1,2) - increasing sequence [1,3,5...]
    • generate(random,1,100) - random values between min/max [31,56,2,69...]

Building

Load Visual Studio solution (ConvertOnnxModel.sln), and build.

Build Google Protobuf .lib yourself:

The protobuf-3.5.1 directory in this project contains the bare minimum .lib and .h files to build the executable. If you need a different configuration:

Download/clone: https://github.com/protocolbuffers/protobuf/releases/tag/v3.5.1 or git clone -b v3.5.1 https://github.com/google/protobuf.git

md protobuf-3.5.1\cmake\build\solution\
cd protobuf-3.5.1\cmake\build\solution\
cmake -G "Visual Studio 15 2017" -DCMAKE_INSTALL_PREFIX=../../../../install ../.. -Dprotobuf_BUILD_TESTS=OFF
#cmake -G "Visual Studio 15 2017" -DCMAKE_INSTALL_PREFIX=../../../../install ../.. -Dprotobuf_BUILD_TESTS=OFF -Ax64
start protobuf-3.5.1\cmake\build\solution\protobuf.sln
# Build it

Build libprotobuf.vcxproj project. Find the .lib files under protobuf-3.5.1\cmake\build\solution*.

Updating ONNX Proto file.

Copy the newest version of onnx.proto from https://github.com/onnx/onnx/blob/master/onnx/onnx.proto.

Run the ProtoBuf compiler.

protobuf3.5.1\bin\protoc.exe onnx.proto --cpp_out=. --error_format=msvs