paper-with-me

홈 › Papers

HMoE: Heterogeneous Mixture of Experts for Language Modeling

2024-08-20 · An Wang, Xingwu Sun, Ruobing Xie, Shuaipeng Li, Jiaqi Zhu, Zhen Yang, Pinxue Zhao, J. N. Han, Zhanhui Kang, Di Wang, Naoaki Okazaki, Cheng-Zhong Xu

Mixture of Experts (MoE) offers remarkable performance and computational efficiency by selectively activating subsets of model parameters. Traditionally, MoE models use homogeneous experts, each with identical capacity. However, varying complexity in input data necessitates experts with diverse capabilities, while homogeneous MoE hinders effective expert specialization and efficient parameter utilization. In this study, we propose a novel Heterogeneous Mixture of Experts (HMoE), where experts differ in size and thus possess diverse capacities. This heterogeneity allows for more specialized experts to handle varying token complexities more effectively. To address the imbalance in expert activation, we propose a novel training objective that encourages the frequent activation of smaller experts, enhancing computational efficiency and parameter utilization. Extensive experiments demonstrate that HMoE achieves lower loss with fewer activated parameters and outperforms conventional homogeneous MoE models on various pre-training evaluation benchmarks. Codes will be released upon acceptance.

📄 PDF Abstract BibTeX arXiv:2408.10681

Code (0)

등록된 구현이 없습니다.

Tasks

Computational EfficiencyLanguage ModelingLanguage ModellingMixture-of-Experts

Methods 이 논문이 사용한 방법론

MoE 설명 없음

Similar Papers 제목 키워드 기반

HMOE: Hypernetwork-based Mixture of Experts for Domain Generalization

2022-11-15 · Jingang Qu, Thibault Faney, Ze Wang, Patrick Gallinari 외

Due to domain shifts, machine learning systems typically struggle to generalize well to new domains that differ from those of training data, which is what domain generalization (DG) aims to address. Although a variety of…

Domain GeneralizationMixture-of-Experts

GRAPHMOE: Amplifying Cognitive Depth of Mixture-of-Experts Network via Introducing Self-Rethinking Mechanism

2025-01-14 · Chen Tang, Bo Lv, Zifan Zheng, Bohao Yang 외

Traditional Mixture-of-Experts (MoE) networks benefit from utilizing multiple smaller expert models as opposed to a single large network. However, these experts typically operate independently, leaving a question open ab…

Mixture-of-Experts

SHMoAReg: Spark Deformable Image Registration via Spatial Heterogeneous Mixture of Experts and Attention Heads

2025-09-24 · Yuxi Zheng, Jianhui Feng, Tianran Li, Marius Staring 외 arxiv

Encoder-Decoder architectures are widely used in deep learning-based Deformable Image Registration (DIR), where the encoder extracts multi-scale features and the decoder predicts deformation fields by recovering spatial …

Image Registration

On Expert Estimation in Hierarchical Mixture of Experts: Beyond Softmax Gating Functions

2024-10-03 · Huy Nguyen, Xing Han, Carl William Harris, Suchi Saria 외

With the growing prominence of the Mixture of Experts (MoE) architecture in developing large-scale foundation models, we investigate the Hierarchical Mixture of Experts (HMoE), a specialized variant of MoE that excels in…

image-classificationImage ClassificationMixture-of-Experts

FlashMoE: Reducing SSD I/O Bottlenecks via ML-Based Cache Replacement for Mixture-of-Experts Inference on Edge Devices

2026-01-22 · Byeongju Kim, Jungwan Lee, Donghyeon Han, Hoi-Jun Yoo 외 arxiv

Recently, Mixture-of-Experts (MoE) models have gained attention for efficiently scaling large language models. Although these models are extremely large, their sparse activation enables inference to be performed by acces…