Towards the ultimate framework efficiency скачать в хорошем качестве

Towards the ultimate framework efficiency

r. Przemyslaw Tredak, Sr. DL Frameworks Engineer @ NVIDIA The dependency engine inside MXNet has a very elegant design that enables efficient multithreaded execution by bypassing Python's Global Interpreter Lock (GIL). While the design works very well for hybridized execution, imperative execution exposes small inefficiencies in the approach. Furthermore, as GPUs get faster and the number of GPUs used in training jobs gets larger, those inefficiencies become a problem even for the hybridized models. In this talk we will explore the current MXNet dependency engine design and look at examples of the inefficient execution. We will also propose ways to improve this design to achieve the fastest training and inference times for both imperative and hybridized models.