Pith. sign in

REVIEW 5 cited by

LocMoE: A Low-Overhead MoE for Large Language Model Training

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2401.13920 v3 pith:Y6NUXQRE submitted 2024-01-25 cs.LG cs.AIcs.CL

classification cs.LGcs.AIcs.CL
keywords modelcommunicationlargeloadroutingtrainingall-to-allcapacity
verification ladder T0 review T1 audit T2 compute T3 formal

Signed reviews

No signed human review yet.

0 comments
read the original abstract

The Mixtures-of-Experts (MoE) model is a widespread distributed and integrated learning method for large language models (LLM), which is favored due to its ability to sparsify and expand models efficiently. However, the performance of MoE is limited by load imbalance and high latency of All-to-All communication, along with relatively redundant computation owing to large expert capacity. Load imbalance may result from existing routing policies that consistently tend to select certain experts. The frequent inter-node communication in the All-to-All procedure also significantly prolongs the training time. To alleviate the above performance problems, we propose a novel routing strategy that combines load balance and locality by converting partial inter-node communication to that of intra-node. Notably, we elucidate that there is a minimum threshold for expert capacity, calculated through the maximal angular deviation between the gating weights of the experts and the assigned tokens. We port these modifications on the PanGu-Sigma model based on the MindSpore framework with multi-level routing and conduct experiments on Ascend clusters. The experiment results demonstrate that the proposed LocMoE reduces training time per epoch by 12.68% to 22.24% compared to classical routers, such as hash router and switch router, without impacting the model accuracy.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 5 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Towards Paradigm-General Suicide Risk Detection via Speech LLM

    eess.AS 2025-09 conditional novelty 5.0 of 10

    A speech LLM with a mixture of DoRA experts (MoDE) improves cross-task suicide risk detection accuracy (0.656 vs 0.635 joint tuning) and calibration on 1,223 Chinese adolescents across ten speech tasks, though the uns...

  2. Theory of Mixture-of-Experts for Mobile Edge Computing

    cs.LG 2024-12 conditional novelty 5.0 of 10

    For continual learning in mobile edge computing, a mixture-of-experts router with enough edge servers can bound overall generalization error to the minimum within-task-type gap, unlike standard offloading.

  3. Software Engineering for Large Language Models: Research Status, Challenges and the Road Ahead

    cs.SE 2025-06 conditional novelty 4.0 of 10

    A literature review organizes LLM development into a six-phase software engineering lifecycle and identifies challenges and research directions for each phase.

  4. MoE$^2$: Optimizing Collaborative Inference for Edge Large Language Models

    cs.NI 2025-01 reject novelty 4.0 of 10

    MoE2 trains a gating network once on all edge LLMs, then selects a resource-feasible subset and a per-prompt top-k of experts to fuse token probabilities under latency and energy budgets.

  5. A Survey on Inference Optimization Techniques for Mixture of Experts Models

    cs.LG 2024-12 conditional novelty 3.0 of 10

    A structured survey of MoE inference optimization that categorizes existing techniques into model, system, and hardware levels and summarizes reported speedups and memory savings.

Pith tools