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

Machine Learning models are often composed of pipelines of transformations.While this design allows to efficiently execute single model components attraining time, prediction serving has different requirements such as lowlatency, high throughput and graceful performance degradation under heavy load.Current prediction serving systems consider models as black boxes, wherebyprediction-time-specific optimizations are ignored in favor of ease ofdeployment. In this paper, we present PRETZEL, a prediction serving systemintroducing a novel white box architecture enabling both end-to-end andmulti-model optimizations. Using production-like model pipelines, ourexperiments show that PRETZEL is able to introduce performance improvementsover different dimensions; compared to state-of-the-art approaches PRETZEL ison average able to reduce 99th percentile latency by 5.5x while reducing memoryfootprint by 25x, and increasing throughput by 4.7x.

PRETZEL: Opening the Black Box of Machine Learning Prediction Serving Systems