Question about PyTorch 1.8 performance optimization for CPU inference

deployment
2

Just answered this here Inference extremely slow on AWS (CPU)? - #2 by marksaroufim