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

It is usually infeasible to fit and train an entire large deep neural network(DNN) model using a single edge device due to the limited resources. Tofacilitate intelligent applications across edge devices, researchers haveproposed partitioning a large model into several sub-models, and deploying eachof them to a different edge device to collaboratively train a DNN model.However, the communication overhead caused by the large amount of datatransmitted from one device to another during training, as well as thesub-optimal partition point due to the inaccurate latency prediction ofcomputation at each edge device can significantly slow down training. In thispaper, we propose AccEPT, an acceleration scheme for accelerating the edgecollaborative pipeline-parallel training. In particular, we propose alight-weight adaptive latency predictor to accurately estimate the computationlatency of each layer at different devices, which also adapts to unseen devicesthrough continuous learning. Therefore, the proposed latency predictor leads tobetter model partitioning which balances the computation loads acrossparticipating devices. Moreover, we propose a bit-level computation-efficientdata compression scheme to compress the data to be transmitted between devicesduring training. Our numerical results demonstrate that our proposedacceleration approach is able to significantly speed up edge pipeline paralleltraining up to 3 times faster in the considered experimental settings.

AccEPT: An Acceleration Scheme for Speeding Up Edge Pipeline-parallel Training