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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

With the increasing popularity of graph-based learning, Graph Neural Networks(GNNs) win lots of attention from the research and industry field because oftheir high accuracy. However, existing GNNs suffer from high memory footprints(e.g., node embedding features). This high memory footprint hurdles thepotential applications towards memory-constrained devices, such as thewidely-deployed IoT devices. To this end, we propose a specialized GNNquantization scheme, SGQuant, to systematically reduce the GNN memoryconsumption. Specifically, we first propose a GNN-tailored quantizationalgorithm design and a GNN quantization fine-tuning scheme to reduce memoryconsumption while maintaining accuracy. Then, we investigate themulti-granularity quantization strategy that operates at different levels(components, graph topology, and layers) of GNN computation. Moreover, we offeran automatic bit-selecting (ABS) to pinpoint the most appropriate quantizationbits for the above multi-granularity quantizations. Intensive experiments showthat SGQuant can effectively reduce the memory footprint from 4.25x to 31.9xcompared with the original full-precision GNNs while limiting the accuracy dropto 0.4% on average.

SGQuant: Squeezing the Last Bit on Graph Neural Networks with Specialized Quantization