Release 0.9.6
·
280 commits
to master
since this release
Switch over to Tensorflow-optimized graph version of model. This almost doubled the inference speed on my machine!