michaelhsieh42 / sagemaker-gpu-performance-io-deeplearning Goto Github PK
View Code? Open in Web Editor NEWThis project forked from aws-samples/sagemaker-gpu-performance-io-deeplearning
GPU IO optimizations for Deep Learning - Illustrated with Caltech 256 dataset on SageMaker. Here, we will be focusing on optimizations for improving I/O for GPU performance tuning. You can typically obtain 10X improvements in overall GPU training performance by just optimizing IO processing routines
License: MIT No Attribution