NOTE: This repository does not seem to yield the correct output anymore with the latest versions of Keras and PyTorch. Take care to verify the results or use an alternative method for conversion.
This repository contains utilities for converting PyTorch models to Keras and the other way around. More specifically, it allows you to copy the weights from a PyTorch model to an identical model in Keras and vice-versa.
From Keras you can then run it on the TensorFlow, Theano and CNTK backend. You can also convert it to a pure TensorFlow model (see [1] and [2]), which allows you to choose more robust deployment options in the cloud, or even mobile devices. From Keras you can also do inference in browsers with keras-js.
Clone this repository, and simply run
pip install .
You need to have PyTorch and torchvision installed beforehand, see the PyTorch website for how to easily install that.
To run the unit and integration tests:
python setup.py test
# OR, if you have nose2 installed,
nose2
There is also Travis CI which will automatically build every commit, see the button at the top of the readme. You can test the direction of weight transfer individually using the TEST_TRANSFER_DIRECTION environment variable, see .travis.yml.
See example.ipynb for a small tutorial on how to use this library.
- This repository is fully PEP8 compliant, I recommend
flake8. - It works for both Python 2 and 3.