Flash Communication: Reducing Tensor Parallelization Bottleneck for Fast Large Language Model Inference

• URL: http://arxiv.org/abs/2412.04964v2
• Date: Wed, 11 Dec 2024 13:27:00 GMT
• Title: Flash Communication: Reducing Tensor Parallelization Bottleneck for Fast Large Language Model Inference
• Authors: Qingyuan Li, Bo Zhang, Liang Ye, Yifan Zhang, Wei Wu, Yerui Sun, Lin Ma, Yuchen Xie,
• Abstract summary: We introduce Flash Communication, a novel low-bit compression technique designed to alleviate the tensor-parallelism communication bottleneck during inference.<n>Our method substantially boosts intra-node communication speed by more than 3x and reduces the time-to-first-token by 2x, with nearly no sacrifice in model accuracy.
• Score: 14.805702987440512
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: The ever-increasing sizes of large language models necessitate distributed solutions for fast inference that exploit multi-dimensional parallelism, where computational loads are split across various accelerators such as GPU clusters. However, this approach often introduces significant communication overhead, especially on devices with limited bandwidth. In this paper, we introduce Flash Communication, a novel low-bit compression technique designed to alleviate the tensor-parallelism communication bottleneck during inference. Our method substantially boosts intra-node communication speed by more than 3x and reduces the time-to-first-token by 2x, with nearly no sacrifice in model accuracy. Extensive experiments on various up-to-date LLMs demonstrate the effectiveness of our approach.

Related papers

• FlashOverlap: A Lightweight Design for Efficiently Overlapping Communication and Computation [6.284874558004134]
  We propose FlashOverlap, a lightweight design characterized by tile-wise overlapping, interference-free computation, and communication agnosticism. Experiments show that such a lightweight design achieves up to 1.65x speedup, outperforming existing works in most cases.
  arXiv  Detail & Related papers  (2025-04-28T06:37:57Z)
• SPD: Sync-Point Drop for efficient tensor parallelism of Large Language Models [6.065998616707588]
  We introduce Sync-Point Drop (SPD) to reduce communication overheads in tensor parallelism by selectively dropping synchronization on attention outputs. SPD offered about 20% overall inference latency reduction with 1% accuracy regression for LLaMA2-70B inference over 8 GPUs.
  arXiv  Detail & Related papers  (2025-02-28T05:20:48Z)
• DeMo: Decoupled Momentum Optimization [6.169574689318864]
  Training large neural networks typically requires sharing between accelerators through specialized high-speed interconnects.<n>We introduce bfDecoupled textbfMomentum (DeMo), a fused magnitude and data parallel algorithm that reduces inter-accelerator communication requirements.<n> Empirical results show that models trained with DeMo match or exceed the performance of equivalent models trained with AdamW.
  arXiv  Detail & Related papers  (2024-11-29T17:31:47Z)
• Communication Compression for Tensor Parallel LLM Inference [1.199955563466263]
  Large Language Models (LLMs) have pushed the frontier of artificial intelligence but are comprised of hundreds of billions of parameters and operations. For faster inference latency, LLMs are deployed on multiple hardware accelerators through various Model Parallelism strategies. Our paper looks into the details on one such strategy - Parallel - and proposes to reduce latency by compressing inter-accelerator communication.
  arXiv  Detail & Related papers  (2024-11-14T15:19:01Z)
• High-Dimensional Distributed Sparse Classification with Scalable Communication-Efficient Global Updates [50.406127962933915]
  We develop solutions to problems which enable us to learn a communication-efficient distributed logistic regression model. In our experiments we demonstrate a large improvement in accuracy over distributed algorithms with only a few distributed update steps needed.
  arXiv  Detail & Related papers  (2024-07-08T19:34:39Z)
• LoCoDL: Communication-Efficient Distributed Learning with Local Training and Compression [56.01900711954956]
  We introduce LoCoDL, a communication-efficient algorithm that leverages the two popular and effective techniques of Local training, which reduces the communication frequency, and Compression, in which short bitstreams are sent instead of full-dimensional vectors of floats. LoCoDL provably benefits from local training and compression and enjoys a doubly-accelerated communication complexity, with respect to the condition number of the functions and the model dimension, in the general heterogenous regime with strongly convex functions.
  arXiv  Detail & Related papers  (2024-03-07T09:22:50Z)
• Towards a Better Theoretical Understanding of Independent Subnetwork Training [56.24689348875711]
  We take a closer theoretical look at Independent Subnetwork Training (IST) IST is a recently proposed and highly effective technique for solving the aforementioned problems. We identify fundamental differences between IST and alternative approaches, such as distributed methods with compressed communication.
  arXiv  Detail & Related papers  (2023-06-28T18:14:22Z)
• SqueezeLLM: Dense-and-Sparse Quantization [80.32162537942138]
  Main bottleneck for generative inference with LLMs is memory bandwidth, rather than compute, for single batch inference. We introduce SqueezeLLM, a post-training quantization framework that enables lossless compression to ultra-low precisions of up to 3-bit. Our framework incorporates two novel ideas: (i) sensitivity-based non-uniform quantization, which searches for the optimal bit precision assignment based on second-order information; and (ii) the Dense-and-Sparse decomposition that stores outliers and sensitive weight values in an efficient sparse format.
  arXiv  Detail & Related papers  (2023-06-13T08:57:54Z)
• TAMUNA: Doubly Accelerated Distributed Optimization with Local Training, Compression, and Partial Participation [53.84175614198885]
  In distributed optimization and learning, several machines alternate between local computations in parallel and communication with a distant server. We propose TAMUNA, the first algorithm for distributed optimization that leveraged the two strategies of local training and compression jointly and allows for partial participation.
  arXiv  Detail & Related papers  (2023-02-20T08:37:44Z)
• Optimus-CC: Efficient Large NLP Model Training with 3D Parallelism Aware Communication Compression [8.591088380355252]
  We present Optimus-CC, a fast and scalable distributed training framework for large NLP models with aggressive communication compression. We propose techniques to avoid the model quality drop that comes from the compression. We demonstrate our solution on a GPU cluster, and achieve superior speedup from the baseline state-of-the-art solutions for distributed training without sacrificing the model quality.
  arXiv  Detail & Related papers  (2023-01-24T06:07:55Z)
• Provably Doubly Accelerated Federated Learning: The First Theoretically Successful Combination of Local Training and Compressed Communication [7.691755449724637]
  We propose the first algorithm for distributed optimization and federated learning. Our algorithm converges linearly to an exact solution, with a doubly accelerated rate.
  arXiv  Detail & Related papers  (2022-10-24T14:13:54Z)
• Correlating sparse sensing for large-scale traffic speed estimation: A Laplacian-enhanced low-rank tensor kriging approach [76.45949280328838]
  We propose a Laplacian enhanced low-rank tensor (LETC) framework featuring both lowrankness and multi-temporal correlations for large-scale traffic speed kriging. We then design an efficient solution algorithm via several effective numeric techniques to scale up the proposed model to network-wide kriging.
  arXiv  Detail & Related papers  (2022-10-21T07:25:57Z)
• Asynchronous Parallel Incremental Block-Coordinate Descent for Decentralized Machine Learning [55.198301429316125]
  Machine learning (ML) is a key technique for big-data-driven modelling and analysis of massive Internet of Things (IoT) based intelligent and ubiquitous computing. For fast-increasing applications and data amounts, distributed learning is a promising emerging paradigm since it is often impractical or inefficient to share/aggregate data. This paper studies the problem of training an ML model over decentralized systems, where data are distributed over many user devices.
  arXiv  Detail & Related papers  (2022-02-07T15:04:15Z)
• A Linearly Convergent Algorithm for Decentralized Optimization: Sending Less Bits for Free! [72.31332210635524]
  Decentralized optimization methods enable on-device training of machine learning models without a central coordinator. We propose a new randomized first-order method which tackles the communication bottleneck by applying randomized compression operators. We prove that our method can solve the problems without any increase in the number of communications compared to the baseline.
  arXiv  Detail & Related papers  (2020-11-03T13:35:53Z)

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.