paper-with-me

Papers

Pathways: Asynchronous Distributed Dataflow for ML

2022-03-23 · Paul Barham, Aakanksha Chowdhery, Jeff Dean, Sanjay Ghemawat, Steven Hand, Dan Hurt, Michael Isard, Hyeontaek Lim, Ruoming Pang, Sudip Roy, Brennan Saeta, Parker Schuh, Ryan Sepassi, Laurent El Shafey, Chandramohan A. Thekkath, Yonghui Wu

We present the design of a new large scale orchestration layer for accelerators. Our system, Pathways, is explicitly designed to enable exploration of new systems and ML research ideas, while retaining state of the art performance for current models. Pathways uses a sharded dataflow graph of asynchronous operators that consume and produce futures, and efficiently gang-schedules heterogeneous parallel computations on thousands of accelerators while coordinating data transfers over their dedicated interconnects. Pathways makes use of a novel asynchronous distributed dataflow design that lets the control plane execute in parallel despite dependencies in the data plane. This design, with careful engineering, allows Pathways to adopt a single-controller model that makes it easier to express complex new parallelism patterns. We demonstrate that Pathways can achieve performance parity (~100% accelerator utilization) with state-of-the-art systems when running SPMD computations over 2048 TPUs, while also delivering throughput comparable to the SPMD case for Transformer models that are pipelined across 16 stages, or sharded across two islands of accelerators connected over a data center network.

📄 PDF Abstract BibTeX arXiv:2203.12533

Code (0)

등록된 구현이 없습니다.

Tasks

Playing the Game of 2048

Methods 이 논문이 사용한 방법론

Multi-Head Attention 설명 없음
Attention 설명 없음
Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
BPE Byte Pair Encoding, or BPE, is a subword segmentation algorithm that encodes rare and unknown words as sequences of subword units. The intuition is that various word…
Residual Connection 설명 없음
Position-Wise Feed-Forward Layer 설명 없음
Dense Connections Dense Connections, or Fully Connected Connections, are a type of layer in a deep neural network that use a linear operation where every input is connected to every output…
Softmax The Softmax output function transforms a previous layer's output into a vector of probabilities. It is commonly used for multiclass classification. Given an input vector $x$…

Similar Papers 제목 키워드 기반

RLlib Flow: Distributed Reinforcement Learning is a Dataflow Problem

2020-11-25 · NeurIPS 2021 12 · Eric Liang, Zhanghao Wu, Michael Luo, Sven Mika 외

Researchers and practitioners in the field of reinforcement learning (RL) frequently leverage parallel computation, which has led to a plethora of new algorithms and systems in the last few years. In this paper, we re-ex…

reinforcement-learningReinforcement LearningReinforcement Learning (RL)rllib

Current Trends and Future Research Directions for Interactive Music

2018-10-05 · Mauricio Toro

In this review, it is explained and compared different software and formalisms used in music interaction: sequencers, computer-assisted improvisation, meta- instruments, score-following, asynchronous dataflow languages, …

Enel: Context-Aware Dynamic Scaling of Distributed Dataflow Jobs using Graph Propagation

2021-08-27 · Dominik Scheinert, Houkun Zhu, Lauritz Thamsen, Morgan K. Geldenhuys 외

Distributed dataflow systems like Spark and Flink enable the use of clusters for scalable data analytics. While runtime prediction models can be used to initially select appropriate cluster resources given target runtime…

Descriptive

Tawa: Automatic Warp Specialization for Modern GPUs with Asynchronous References

2025-10-16 · Hongzheng Chen, Bin Fan, Alexander Collins, Bastian Hagedorn 외 arxiv

Modern GPUs feature specialized hardware units that enable high-performance, asynchronous dataflow execution. However, the conventional SIMT programming model is fundamentally misaligned with this task-parallel hardware,…

AsyncFlow: An Asynchronous Streaming RL Framework for Efficient LLM Post-Training

2025-07-02 · Zhenyu Han, Ansheng You, Haibo Wang, Kui Luo 외 arxiv

Reinforcement learning (RL) has become a pivotal technology in the post-training phase of large language models (LLMs). Traditional task-colocated RL frameworks suffer from significant scalability bottlenecks, while task…

Reinforcement Learning