A tutorial on pin_memory and non_blocking usage #2570

Job	Run time
pytorch_tutorial_build_worker (1, 15, linux.g5.12xlarge.nvidia.gpu)	15m 37s
pytorch_tutorial_build_worker (2, 15, linux.g5.4xlarge.nvidia.gpu)	53m 45s
pytorch_tutorial_build_worker (3, 15, linux.g5.4xlarge.nvidia.gpu)	20m 20s
pytorch_tutorial_build_worker (4, 15, linux.g5.4xlarge.nvidia.gpu)	29m 33s
pytorch_tutorial_build_worker (5, 15, linux.g5.4xlarge.nvidia.gpu)	21m 8s
pytorch_tutorial_build_worker (6, 15, linux.g5.4xlarge.nvidia.gpu)	18m 8s
pytorch_tutorial_build_worker (7, 15, linux.4xlarge.nvidia.gpu)	21m 52s
pytorch_tutorial_build_worker (8, 15, linux.4xlarge.nvidia.gpu)	20m 10s
pytorch_tutorial_build_worker (9, 15, linux.4xlarge.nvidia.gpu)	38m 42s
pytorch_tutorial_build_worker (10, 15, linux.4xlarge.nvidia.gpu)	44m 52s
pytorch_tutorial_build_worker (11, 15, linux.4xlarge.nvidia.gpu)	33m 22s
pytorch_tutorial_build_worker (12, 15, linux.4xlarge.nvidia.gpu)	20m 39s
pytorch_tutorial_build_worker (13, 15, linux.4xlarge.nvidia.gpu)	22m 12s
pytorch_tutorial_build_worker (14, 15, linux.4xlarge.nvidia.gpu)	15m 25s
pytorch_tutorial_build_worker (15, 15, linux.4xlarge.nvidia.gpu)	23m 4s
pytorch_tutorial_build_manager	22m 45s
	7h 1m 34s

Provide feedback