ONNX Command Line Toolbox

• Aim to improve your experience of investigating ONNX models.
• Use it like onnx infershape /path/to/input/model.onnx /path/to/output/model.onnx. (See the usage section.)

Installation

Recommand to install via GitHub repo for the latest functionality.

pip install git+https://github.com/zhenhuaw-me/onnxcli.git

Two alternative ways are:

1. Install via pypi package pip install onnxcli
2. Download and add the code tree to your $PYTHONPATH. (For development purpose and the command line is different.）

   git clone https://github.com/zhenhuaw-me/onnxcli.git
   export PYTHONPATH=$(pwd)/onnxcli:${PYTHONPATH}
   python onnxcli/cli/dispatcher.py <sub command> <more args...>
   

Requirements

onnxcli depends on different packages w.r.t. different functionality and may extend. However, we only include only several basic ones (onnx for example) since you may use only a small portion of the functionalities, or you may like have a different version.

Depending on the sub command, error will be raised if the requirements are not met. Follow the error message to install the requirements.

Usage

Once installed, the onnx and onnxcli commands are avaiable. You can play with commands such as onnx infershape /path/to/input/model.onnx /path/to/output/model.onnx.

The general format is onnx <sub command> <dedicated arguments ...>. The sub commands are as sections below.

Check the online help with onnx --help and onnx <subcmd> --help for latest usage.

infershape

onnx infershape performs shape inference of the ONNX model. It's an CLI wrapper of onnx.shape_inference. You will find it useful to generate shape information for the models that are extracted by onnx extract.

extract

onnx extract extracts the sub model that is determined by the names of the input and output tensor of the subgraph from the original model. It's a CLI wrapper of onnx.utils.extract_model (which I authorized in the ONNX repo).

inspect

onnx inspect gives you quick view of the information of the given model. It's inspired by the tf-onnx tool.

When working on deep learning, you may like to take a look at what's inside the model. Netron is powerful but doesn't provide fine-grain view.

With onnx inspect, you no longer need to scroll the Netron window to look for nodes or tensors. Instead, you can dump the node attributes and tensor values with a single command.

$ onnx inspect ./assets/tests/conv.float32.onnx --node --indices 0 --detail
Inpect of model ./assets/tests/conv.float32.onnx
Graph name: 9
Graph inputs: 1
Graph outputs: 1
Nodes in total: 1
ValueInfo in total: 2
Initializers in total: 2
Sparse Initializers in total: 0
Quantization in total: 0
Node information:
Node "output": type "Conv", inputs "['input', 'Variable/read', 'Conv2D_bias']", outputs "['output']"
attributes: [name: "dilations"
ints: 1
ints: 1
type: INTS
, name: "group"
i: 1
type: INT
, name: "kernel_shape"
ints: 3
ints: 3
type: INTS
, name: "pads"
ints: 1
ints: 1
ints: 1
ints: 1
type: INTS
, name: "strides"
ints: 1
ints: 1
type: INTS
]

$ onnx inspect ./assets/tests/conv.float32.onnx --tensor --names Conv2D_bias --detail
Inpect of model ./assets/tests/conv.float32.onnx
Graph name: 9
Graph inputs: 1
Graph outputs: 1
Nodes in total: 1
ValueInfo in total: 2
Initializers in total: 2
Sparse Initializers in total: 0
Quantization in total: 0
Tensor information:
Initializer "Conv2D_bias": type FLOAT, shape [16],
float data: [0.4517577290534973, -0.014192663133144379, 0.2946248948574066, -0.9742919206619263, -1.2975586652755737, 0.7223454117774963, 0.7835700511932373, 1.7674627304077148, 1.7242872714996338, 1.1230682134628296, -0.2902531623840332, 0.2627834975719452, 1.0175092220306396, 0.5643373131752014, -0.8244842290878296, 1.2169424295425415]

draw

onnx draw draws the graph in dot, svg, png formats. It gives you quick view of the type and shape of the tensors that are fed to a specific node. You can view the model topology in image viewer of browser without waiting for the model to load, which I found is really helpful for large models.

If you are viewing svg in browser, you can even quick search for the nodes and tensors. Together with onnx inspect, it will be very efficient to understand the issue you are looking into.

The node are in ellipses and tensors are in rectangles where the rounded ones are initializers. The node type of the node and the data type and shape of the tenors are also rendered. Here is a Convolution node example.

Note: The onnx draw requires dot command (graphviz) to be avaiable on your machine - which can be installed by command as below on Ubuntu/Debian.

sudo apt install -y graphviz

optimize

onnx optimize optimizes the input model with ONNX Optimizer.

Contributing

Welcome to contribute new commands or enhance them. Let's make our life easier together.

The workflow is pretty simple:

1. Starting with GitHub Codespace or clone locally.

2. make setup to config the dependencies (or pip install -r ./requirements.txt if you prefer).

3. Create a new subcommand

   1. Starting by copying and modifying infershape.
   2. Register the command in the dispatcher
   3. Create a new command line test
   4. make test to build and test.
   5. make check and make format to fix any code style issues.

4. Try out, debug, commit, push, and open pull request.

   1. The code has been protected by CI. You need to get a pass before merging.
   2. Ask if any questions.

License

Apache License Version 2.0.