English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Embedding tables dominate industrial-scale recommendation model sizes, usingup to terabytes of memory. A popular and the largest publicly available machinelearning MLPerf benchmark on recommendation data is a Deep LearningRecommendation Model (DLRM) trained on a terabyte of click-through data. Itcontains 100GB of embedding memory (25+Billion parameters). DLRMs, due to theirsheer size and the associated volume of data, face difficulty in training,deploying for inference, and memory bottlenecks due to large embedding tables.This paper analyzes and extensively evaluates a generic parameter sharing setup(PSS) for compressing DLRM models. We show theoretical upper bounds on thelearnable memory requirements for achieving $(1 \pm \epsilon)$ approximationsto the embedding table. Our bounds indicate exponentially fewer parameterssuffice for good accuracy. To this end, we demonstrate a PSS DLRM reaching10000$\times$ compression on criteo-tb without losing quality. Such acompression, however, comes with a caveat. It requires 4.5 $\times$ moreiterations to reach the same saturation quality. The paper argues that thistradeoff needs more investigations as it might be significantly favorable.Leveraging the small size of the compressed model, we show a 4.3$\times$improvement in training latency leading to similar overall training times.Thus, in the tradeoff between system advantage of a small DLRM model vs. slowerconvergence, we show that scales are tipped towards having a smaller DLRMmodel, leading to faster inference, easier deployment, and similar trainingtimes.

The trade-offs of model size in large recommendation models : A 10000 $\times$ compressed criteo-tb DLRM model (100 GB parameters to mere 10MB)