by Team PyTorch

Reduce inference costs by 71% and drive scale out using PyTorch, TorchServe, and AWS Inferentia.