MétaCan
Menu
Back to cohort
Record W3171596153 · doi:10.1109/jsac.2021.3087272

LOSP: Overlap Synchronization Parallel With Local Compensation for Fast Distributed Training

2021· article· en· W3171596153 on OpenAlexaff
Haozhao Wang, Zhihao Qu, Song Guo, Ningqi Wang, Ruixuan Li, Weihua Zhuang

Bibliographic record

VenueIEEE Journal on Selected Areas in Communications · 2021
Typearticle
Languageen
FieldComputer Science
TopicStochastic Gradient Optimization Techniques
Canadian institutionsUniversity of Waterloo
FundersNational Key Research and Development Program of ChinaResearch Grants Council, University Grants CommitteeChina Postdoctoral Science FoundationImpact FundScience, Technology and Innovation Commission of Shenzhen MunicipalityNational Natural Science Foundation of China
KeywordsComputer scienceScalabilitySynchronization (alternating current)ComputationSpeedupDistributed computingOverhead (engineering)Convergence (economics)Stochastic gradient descentRate of convergenceCompensation (psychology)Data synchronizationMathematical optimizationParallel computingAlgorithmKey (lock)Computer networkArtificial intelligenceArtificial neural network

Abstract

fetched live from OpenAlex

When running in Parameter Server (PS), the Distributed Stochastic Gradient Descent (D-SGD) incurs significant communication delays and huge communication overhead due to the model synchronization. Moreover, considering the heterogeneity of computational capability among workers, traditional synchronization modes incur under-utilization of computational resources because fast workers have to wait for slow ones finishing the computation. Although our previous work OSP can effectively solve these problems by overlapping the computation and communication procedures and allowing adaptive multiple local updates in distributed training, it causes the staleness problem brought by the overlap, yielding a performance degradation. In this paper, we propose a new method named LOSP by introducing local compensation to our previous synchronization mechanism, which mitigates adverse effects caused by the overlapping synchronization. We theoretically prove that LOSP (1) preserves the same convergence rate as the sequential SGD for non-convex problems, and (2) exhibits good scalability due to the linear speedup property with respect to both the number of workers and the average number of local updates. Evaluations show that LOSP significantly improves performance over the state-of-the-art ones in terms of both convergence accuracy and communication cost.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame machine prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.

metaresearch head score (Codex)0.001
metaresearch head score (Gemma)0.002
Version: metacan-v3-hybrid-931329e0061cValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Simulation or modeling · Consensus signal: Simulation or modeling
GenreCandidate signal: Empirical · Consensus signal: none
Teacher disagreement score0.004
Threshold uncertainty score0.008

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0010.002
Meta-epidemiology (narrow)0.0010.000
Meta-epidemiology (broad)0.0010.000
Bibliometrics0.0000.001
Science and technology studies0.0010.001
Scholarly communication0.0010.001
Open science0.0020.002
Research integrity0.0010.001
Insufficient payload (model declined to judge)0.0020.001

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.039
GPT teacher head0.286
Teacher spread0.247 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.

The models applied no category: nothing in the taxonomy fit this work.
Study designSimulation or modeling
Domainnot available
GenreEmpirical

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations21
Published2021
Admission routes1
Has abstractyes

Explore more

Same venueIEEE Journal on Selected Areas in CommunicationsSame topicStochastic Gradient Optimization TechniquesFrench-language works237,207