Open-source to speed up deep learning inference by leveraging multiple optimization techniques (deep learning compilers, quantization, half precision, etc)

Follow the full discussion on Reddit.
Hi everyone, my name is Emile. After trying for a long time many tools to speed up AI inference, with some colleagues I have built an open-source library to bundle the best techniques you could find around and consolidated them into a single interface, this opensource library (today's release features the integration of most of these techniques). Let me know what you think of this OSS! Thank you