Multi-Faceted Large Embedding Tables for Pinterest Ads Ranking

TL;DR

This paper presents a novel multi-faceted pretraining scheme for large embedding tables in Pinterest's ads ranking, improving performance and scalability through a hybrid serving infrastructure, leading to better CTR and CPC metrics.

Contribution

Introduces a multi-faceted pretraining approach and a CPU-GPU hybrid infrastructure to enhance large embedding table effectiveness and scalability in recommendation systems.

Findings

Significant improvements in CTR and CVR metrics.

Achieved 1.34% CPC reduction and 2.60% CTR increase.

Enhanced embedding quality through multi-algorithm pretraining.

Abstract

Large embedding tables are indispensable in modern recommendation systems, thanks to their ability to effectively capture and memorize intricate details of interactions among diverse entities. As we explore integrating large embedding tables into Pinterest's ads ranking models, we encountered not only common challenges such as sparsity and scalability, but also several obstacles unique to our context. Notably, our initial attempts to train large embedding tables from scratch resulted in neutral metrics. To tackle this, we introduced a novel multi-faceted pretraining scheme that incorporates multiple pretraining algorithms. This approach greatly enriched the embedding tables and resulted in significant performance improvements. As a result, the multi-faceted large embedding tables bring great performance gain on both the Click-Through Rate (CTR) and Conversion Rate (CVR) domains. Moreover, we designed a CPU-GPU hybrid serving infrastructure to overcome GPU memory limits and elevate the scalability. This framework has been deployed in the Pinterest Ads system and achieved 1.34% online CPC reduction and 2.60% CTR increase with neutral end-to-end latency change.