paper-with-me

Papers

Approximate Top-$k$ for Increased Parallelism

2024-12-05 · Oscar Key, Luka Ribar, Alberto Cattaneo, Luke Hudlass-Galley, Douglas Orr

We present an evaluation of bucketed approximate top-$k$ algorithms. Computing top-$k$ exactly suffers from limited parallelism, because the $k$ largest values must be aggregated along the vector, thus is not well suited to computation on highly-parallel machine learning accelerators. By relaxing the requirement that the top-$k$ is exact, bucketed algorithms can dramatically increase the parallelism available by independently computing many smaller top-$k$ operations. We explore the design choices of this class of algorithms using both theoretical analysis and empirical evaluation on downstream tasks. Our motivating examples are sparsity algorithms for language models, which often use top-$k$ to select the most important parameters or activations. We also release a fast bucketed top-$k$ implementation for PyTorch.

📄 PDF Abstract BibTeX arXiv:2412.04358

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Topkapi: Parallel and Fast Sketches for Finding Top-K Frequent Elements

2018-12-01 · NeurIPS 2018 12 · Ankush Mandal, He Jiang, Anshumali Shrivastava, Vivek Sarkar

Identifying the top-K frequent items is one of the most common and important operations in large data processing systems. As a result, several solutions have been proposed to solve this problem approximately. In this pap…

A Two-Tier Perspective on Inference-Time Parallelism in Multi-Agent LLM Systems

2026-08-06 · Zihan Xu, Haolin Tian, Hai Jiang arxiv

Large language model (LLM)-driven multi-agent systems typically require multiple model invocations and complex coordination during inference, and their execution strategies directly affect system accuracy, latency, and c…

Research on Model Parallelism and Data Parallelism Optimization Methods in Large Language Model-Based Recommendation Systems

2025-06-21 · Haowei Yang, Yu Tian, Zhongheng Yang, Zhao Wang 외

With the rapid adoption of large language models (LLMs) in recommendation systems, the computational and communication bottlenecks caused by their massive parameter sizes and large data volumes have become increasingly p…

Language ModelingLanguage ModellingLarge Language Modelmodel+2

APB: Accelerating Distributed Long-Context Inference by Passing Compressed Context Blocks across GPUs

2025-02-17 · Yuxiang Huang, Mingye Li, Xu Han, Chaojun Xiao 외

While long-context inference is crucial for advancing large language model (LLM) applications, its prefill speed remains a significant bottleneck. Current approaches, including sequence parallelism strategies and compute…

Language ModelingLanguage ModellingLarge Language Model

DeepSpeed Ulysses: System Optimizations for Enabling Training of Extreme Long Sequence Transformer Models

2023-09-25 · Sam Ade Jacobs, Masahiro Tanaka, Chengming Zhang, Minjia Zhang 외

Computation in a typical Transformer-based large language model (LLM) can be characterized by batch size, hidden dimension, number of layers, and sequence length. Until now, system works for accelerating LLM training hav…

Language ModellingLarge Language Model