Optimizing ML Workload Network Efficiency (Part I): Feature Trimmer (opens in new tab)

Guangtong Bai \| Staff Software Engineer, Product ML Infrastructure\*; Shantam Shorewala \| Software Engineer II, Product ML Infrastructure\*; Chi Zhang \| Staff Software Engineer, AI Platform\*; Neha Upadhyay \| Software Engineer II, AI Platform\*; Haoyang Li \| Director, Product ML Infrastructure \*These authors contributed equally to this article\. Background At Pinterest, our online ML serving systems employ a root-leaf architecture\. On a high level, the architecture looks as follows: Fi...