Skip to content
struct2tensor is a library for parsing and manipulating structured data inside of tensorflow.
Python C++ Starlark Shell Smarty
Branch: master
Clone or download


Python PyPI


struct2tensor is a library for parsing structured data inside of tensorflow. In particular, it makes it easy to manipulate structured data, e.g., slicing, flattening, copying substructures, and so on, as part of a TensorFlow model graph. The notebook in 'examples/prensor_playground.ipynb' provides a few examples of struct2tensor in action and an introduction to the main concepts. You can run the notebook in your browser through Google's colab environment, or download the file to run it in your own Jupyter environment.

There are two main use cases of this repo:

  1. To create a PIP package. The PIP package contains plug-ins (OpKernels) to an existing tensorflow installation.
  2. To staticlly link with tensorflow-serving.

As these processes are independent, one can follow either set of directions below.

Use a pre-built Linux PIP package.

From a virtual environment, run:

pip install struct2tensor

Creating a PIP package.

The struct2tensor PIP package is useful for creating models. It works with either tensorflow 2.0 or tensorflow 1.15.0.

In order to unify the process, we recommend compiling struct2tensor inside a docker container.

Downloading the Code

Go to your home directory.

Download the source code.

git clone
cd ~/struct2tensor

Use docker-compose

Install docker-compose.

Use it to build a pip wheel for Python 3.6 with tensorflow version 2:

docker-compose build manylinux2010
docker-compose run -e PYTHON_VERSION=36 -e TF_VERSION=NIGHTLY_TF manylinux2010

Or build a pip wheel for Python 3.7 with tensorflow version 2 (note that if you run one of these docker-compose commands after the other, the second will erase the result from the first):

docker-compose build manylinux2010
docker-compose run -e PYTHON_VERSION=37 -e TF_VERSION=NIGHTLY_TF manylinux2010

This will create a manylinux package in the ~/struct2tensor/dist directory.

Creating a static library.

In order to construct a static library for tensorflow-serving, we run:

bazel build -c opt struct2tensor:prensor_kernels_and_ops

This can also be linked into another library.


struct2tensor tensorflow* 1.15
You can’t perform that action at this time.