paper-with-me

Papers

SEER-MoE: Sparse Expert Efficiency through Regularization for Mixture-of-Experts

2024-04-07 · Alexandre Muzio, Alex Sun, Churan He

The advancement of deep learning has led to the emergence of Mixture-of-Experts (MoEs) models, known for their dynamic allocation of computational resources based on input. Despite their promise, MoEs face challenges, particularly in terms of memory requirements. To address this, our work introduces SEER-MoE, a novel two-stage framework for reducing both the memory footprint and compute requirements of pre-trained MoE models. The first stage involves pruning the total number of experts using a heavy-hitters counting guidance, while the second stage employs a regularization-based fine-tuning strategy to recover accuracy loss and reduce the number of activated experts during inference. Our empirical studies demonstrate the effectiveness of our method, resulting in a sparse MoEs model optimized for inference efficiency with minimal accuracy trade-offs.

📄 PDF Abstract BibTeX arXiv:2404.05089

Code (0)

등록된 구현이 없습니다.

Tasks

Mixture-of-Experts

Methods 이 논문이 사용한 방법론

Pruning 설명 없음
MoE 설명 없음

Similar Papers 제목 키워드 기반

SeerAttention-R: Sparse Attention Adaptation for Long Reasoning

2025-06-10 · Yizhao Gao, Shuming Guo, Shijie Cao, Yuqing Xia 외

We introduce SeerAttention-R, a sparse attention framework specifically tailored for the long decoding of reasoning models. Extended from SeerAttention, SeerAttention-R retains the design of learning attention sparsity t…

4kGPU

ATMSeer: Increasing Transparency and Controllability in Automated Machine Learning

2019-02-13 · Qianwen Wang, Yao Ming, Zhihua Jin, Qiaomu Shen 외

To relieve the pain of manually selecting machine learning algorithms and tuning hyperparameters, automated machine learning (AutoML) methods have been developed to automatically search for good models. Due to the huge m…

AutoMLBIG-bench Machine Learning

SeerAttention: Learning Intrinsic Sparse Attention in Your LLMs

2024-10-17 · Yizhao Gao, Zhichen Zeng, Dayou Du, Shijie Cao 외

Attention is the cornerstone of modern Large Language Models (LLMs). Yet its quadratic complexity limits the efficiency and scalability of LLMs, especially for those with a long-context window. A promising approach addre…

Geometric Regularization in Mixture-of-Experts: The Disconnect Between Weights and Activations

2026-01-01 · Hyunjun Kim arxiv

Mixture-of-Experts (MoE) models achieve efficiency through sparse activation, but the role of geometric regularization in expert specialization remains unclear. We apply orthogonality loss to enforce expert diversity and…

Efficient Preference-based Reinforcement Learning via Aligned Experience Estimation

2024-05-29 · Fengshuo Bai, Rui Zhao, Hongming Zhang, Sijia Cui 외

Preference-based reinforcement learning (PbRL) has shown impressive capabilities in training agents without reward engineering. However, a notable limitation of PbRL is its dependency on substantial human feedback. This …

reinforcement-learningReinforcement Learning