Clustering the Sketch: Dynamic Compression for Embedding Tables

Published: 21 Sept 2023, Last Modified: 02 Nov 2023NeurIPS 2023 posterEveryoneRevisionsBibTeX
Keywords: Embedding table compression, Clustering and sketching, Memory-efficient training
TL;DR: The CQR algorithm effectively combines clustering and sketching techniques to achieve memory-efficient training and compression rates close to post-training quantization methods in large-scale embedding tables.
Abstract: Embedding tables are used by machine learning systems to work with categorical features. In modern Recommendation Systems, these tables can be very large, necessitating the development of new methods for fitting them in memory, even during training. We suggest Clustered Compositional Embeddings (CCE) which combines clustering-based compression like quantization to codebooks with dynamic methods like The Hashing Trick and Compositional Embeddings [Shi et al., 2020]. Experimentally CCE achieves the best of both worlds: The high compression rate of codebook-based quantization, but \emph{dynamically} like hashing-based methods, so it can be used during training. Theoretically, we prove that CCE is guaranteed to converge to the optimal codebook and give a tight bound for the number of iterations required.
Supplementary Material: pdf
Submission Number: 5402
Loading