How to Optimize Your DL Data-Input Pipeline with a Custom PyTorch Operator

PyTorch Model Performance Analysis and Optimization — Part 5

This post is the fifth in a series of posts on the topic of performance analysis and optimization of GPU-based PyTorch workloads and a direct sequel to part four. In part four, we demonstrated how PyTorch Profiler and TensorBoard can be used to identify, analyze, and address performance bottlenecks in the data pre-processing pipeline of a DL training workload. In this post we discuss …

analysis artificial intelligence data deep learning gpu optimization part performance performance analysis pipeline pytorch series training data workloads