SiDA-MoE: Sparsity-Inspired Data-Aware Serving for Efficient and Scalable Large Mixture-of-Experts Models

• URL: http://arxiv.org/abs/2310.18859v2
• Date: Fri, 17 May 2024 23:33:23 GMT
• Title: SiDA-MoE: Sparsity-Inspired Data-Aware Serving for Efficient and Scalable Large Mixture-of-Experts Models
• Authors: Zhixu Du, Shiyu Li, Yuhao Wu, Xiangyu Jiang, Jingwei Sun, Qilin Zheng, Yongkai Wu, Ang Li, Hai "Helen" Li, Yiran Chen,
• Abstract summary: Mixture-of-Experts (MoE) has emerged as a favorable architecture in the era of large models. Yet, the realization of such benefits often results in ineffective GPU memory utilization. We introduce SiDA-MoE, an efficient inference approach tailored for large MoE models.
• Score: 20.16600129902895
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Mixture-of-Experts (MoE) has emerged as a favorable architecture in the era of large models due to its inherent advantage, i.e., enlarging model capacity without incurring notable computational overhead. Yet, the realization of such benefits often results in ineffective GPU memory utilization, as large portions of the model parameters remain dormant during inference. Moreover, the memory demands of large models consistently outpace the memory capacity of contemporary GPUs. Addressing this, we introduce SiDA-MoE ($\textbf{S}$parsity-$\textbf{i}$nspired $\textbf{D}$ata-$\textbf{A}$ware), an efficient inference approach tailored for large MoE models. SiDA-MoE judiciously exploits both the system's main memory, which is now abundant and readily scalable, and GPU memory by capitalizing on the inherent sparsity on expert activation in MoE models. By adopting a data-aware perspective, SiDA-MoE achieves enhanced model efficiency with a neglectable performance drop. Specifically, SiDA-MoE attains a remarkable speedup in MoE inference with up to $3.93\times$ throughput increasing, up to $72\%$ latency reduction, and up to $80\%$ GPU memory saving with down to $1\%$ performance drop. This work paves the way for scalable and efficient deployment of large MoE models, even with constrained resources. Code is available at: https://github.com/timlee0212/SiDA-MoE.

Related papers

• S3D: A Simple and Cost-Effective Self-Speculative Decoding Scheme for Low-Memory GPUs [7.816840847892339]
  Speculative decoding (SD) has attracted a significant amount of research attention due to the substantial speedup it can achieve for LLM inference. We propose Skippy Simultaneous Speculative Decoding (or S3D), a cost-effective self-speculative SD method based on simultaneous multi-token decoding and mid-layer skipping. Our method has achieved one of the top performance-memory ratios while requiring minimal architecture changes and training data.
  arXiv  Detail & Related papers  (2024-05-30T17:54:35Z)
• DiG: Scalable and Efficient Diffusion Models with Gated Linear Attention [82.24166963631949]
  We introduce Diffusion Gated Linear Attention Transformers (DiG), a simple, adoptable solution with minimal parameter overhead, following the Diffusion Transformers (DiT) design. In addition to better performance than DiT, DiG-S/2 exhibits $2.5times$ higher training speed than DiT-S/2 and saves $75.7%$ memory resolution $179times 1792$. With the same model size, DiG-XL/2 is $4.2times$ faster than the recent Mamba-based diffusion model at a $1024$ resolution, and is $1.8times$ faster than DiT with FlashAttention-2
  arXiv  Detail & Related papers  (2024-05-28T17:59:33Z)
• Dense Training, Sparse Inference: Rethinking Training of Mixture-of-Experts Language Models [62.4691912312317]
  Mixture-of-Experts (MoE) language models can reduce computational costs by 2-4$times$ compared to dense models without sacrificing performance. We propose a hybrid dense training and sparse inference framework for MoE models (DS-MoE) which achieves strong computation and parameter efficiency.
  arXiv  Detail & Related papers  (2024-04-08T14:39:49Z)
• MEMORYLLM: Towards Self-Updatable Large Language Models [101.3777486749529]
  Existing Large Language Models (LLMs) usually remain static after deployment. We introduce MEMORYLLM, a model that comprises a transformer and a fixed-size memory pool. MEMORYLLM can self-update with text knowledge and memorize the knowledge injected earlier.
  arXiv  Detail & Related papers  (2024-02-07T07:14:11Z)
• MoE-Infinity: Offloading-Efficient MoE Model Serving [15.826989637041907]
  MoE-Infinity is an offloading-efficient serving system for sparse mixture-of-experts (MoE) models. To optimize offloading, MoE-Infinity achieves novel request-level tracing for expert activation. MoE-Infinity exhibits superior latency performance, providing 2-20X improvements when serving various MoE models.
  arXiv  Detail & Related papers  (2024-01-25T18:07:50Z)
• FFSplit: Split Feed-Forward Network For Optimizing Accuracy-Efficiency Trade-off in Language Model Inference [57.119047493787185]
  This paper shows how to reduce model size by 43.1% and bring $1.25sim1.56times$ wall clock time speedup on different hardware with negligible accuracy drop. In practice, our method can reduce model size by 43.1% and bring $1.25sim1.56times$ wall clock time speedup on different hardware with negligible accuracy drop.
  arXiv  Detail & Related papers  (2024-01-08T17:29:16Z)
• KOALA: Empirical Lessons Toward Memory-Efficient and Fast Diffusion Models for Text-to-Image Synthesis [52.42320594388199]
  We present three key practices in building an efficient text-to-image model. Based on these findings, we build two types of efficient text-to-image models, called KOALA-Turbo &-Lightning. Unlike SDXL, our KOALA models can generate 1024px high-resolution images on consumer-grade GPUs with 8GB of VRAMs (3060Ti)
  arXiv  Detail & Related papers  (2023-12-07T02:46:18Z)
• Pre-gated MoE: An Algorithm-System Co-Design for Fast and Scalable Mixture-of-Expert Inference [23.207326766883405]
  Mixture-of-Experts (MoE) is able to scale its model size without proportionally scaling up its computational requirements. Pre-gated MoE employs our novel pre-gating function which alleviates the dynamic nature of sparse expert activation. We demonstrate that Pre-gated MoE is able to improve performance, reduce GPU memory consumption, while also maintaining the same level of model quality.
  arXiv  Detail & Related papers  (2023-08-23T11:25:37Z)
• Towards MoE Deployment: Mitigating Inefficiencies in Mixture-of-Expert (MoE) Inference [7.743308058511418]
  We provide a characterization of two MoE workloads, namely Language Modeling (LM) and Machine Translation (MT) We propose three optimization techniques to mitigate sources of inefficiencies, namely (1) Dynamic gating, (2) Expert Buffering, and (3) Expert load balancing.
  arXiv  Detail & Related papers  (2023-03-10T19:30:15Z)
• Efficient Large Scale Language Modeling with Mixtures of Experts [61.45159383372181]
  Mixture of Experts layers (MoEs) enable efficient scaling of language models through conditional computation. This paper presents a detailed empirical study of how autoregressive MoE language models scale in comparison with dense models in a wide range of settings.
  arXiv  Detail & Related papers  (2021-12-20T17:05:11Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.