https://arxiv.org/abs/2503.01890

Skip to main content
Cornell University
We gratefully acknowledge support from the Simons Foundation, member
institutions, and all contributors. Donate
 
arxiv logo > cs > arXiv:2503.01890
[                    ]

Help | Advanced Search

[All fields        ]
Search
arXiv logo
Cornell University Logo
[                    ] GO
quick links

  * Login
  * Help Pages
  * About

Computer Science > Machine Learning

arXiv:2503.01890 (cs)
[Submitted on 27 Feb 2025]

Title:AutoHete: An Automatic and Efficient Heterogeneous Training
System for LLMs

Authors:Zihao Zeng, Chubo Liu, Xin He, Juan Hu, Yong Jiang, Fei Huang
, Kenli Li, Wei Yang Bryan Lim
View a PDF of the paper titled AutoHete: An Automatic and Efficient
Heterogeneous Training System for LLMs, by Zihao Zeng and 7 other
authors
View PDF HTML (experimental)

    Abstract:Transformer-based large language models (LLMs) have
    demonstrated exceptional capabilities in sequence modeling and
    text generation, with improvements scaling proportionally with
    model size. However, the limitations of GPU memory have
    restricted LLM training accessibility for many researchers.
    Existing heterogeneous training methods significantly expand the
    scale of trainable models but introduce substantial communication
    overheads and CPU workloads. In this work, we propose AutoHete,
    an automatic and efficient heterogeneous training system
    compatible with both single-GPU and multi-GPU environments.
    AutoHete dynamically adjusts activation checkpointing, parameter
    offloading, and optimizer offloading based on the specific
    hardware configuration and LLM training needs. Additionally, we
    design a priority-based scheduling mechanism that maximizes the
    overlap between operations across training iterations, enhancing
    throughput. Compared to state-of-the-art heterogeneous training
    systems, AutoHete delivers a 1.32x~1.91x throughput improvement
    across various model sizes and training configurations.

Subjects: Machine Learning (cs.LG)
Cite as:  arXiv:2503.01890 [cs.LG]
          (or arXiv:2503.01890v1 [cs.LG] for this version)
          https://doi.org/10.48550/arXiv.2503.01890
          Focus to learn more
          arXiv-issued DOI via DataCite

Submission history

From: Zihao Zeng [view email]
[v1] Thu, 27 Feb 2025 14:46:22 UTC (1,423 KB)
Full-text links:

Access Paper:

    View a PDF of the paper titled AutoHete: An Automatic and
    Efficient Heterogeneous Training System for LLMs, by Zihao Zeng
    and 7 other authors
  * View PDF
  * HTML (experimental)
  * TeX Source
  * Other Formats

view license
Current browse context:
cs.LG
< prev   |   next >
new | recent | 2025-03
Change to browse by:
cs

References & Citations

  * NASA ADS
  * Google Scholar
  * Semantic Scholar

a export BibTeX citation Loading...

BibTeX formatted citation

x
[loading...          ]
Data provided by:

Bookmark

BibSonomy logo Reddit logo
(*) Bibliographic Tools

Bibliographic and Citation Tools

[ ] Bibliographic Explorer Toggle
Bibliographic Explorer (What is the Explorer?)
[ ] Connected Papers Toggle
Connected Papers (What is Connected Papers?)
[ ] Litmaps Toggle
Litmaps (What is Litmaps?)
[ ] scite.ai Toggle
scite Smart Citations (What are Smart Citations?)
( ) Code, Data, Media

Code, Data and Media Associated with this Article

[ ] alphaXiv Toggle
alphaXiv (What is alphaXiv?)
[ ] Links to Code Toggle
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
[ ] DagsHub Toggle
DagsHub (What is DagsHub?)
[ ] GotitPub Toggle
Gotit.pub (What is GotitPub?)
[ ] Huggingface Toggle
Hugging Face (What is Huggingface?)
[ ] Links to Code Toggle
Papers with Code (What is Papers with Code?)
[ ] ScienceCast Toggle
ScienceCast (What is ScienceCast?)
( ) Demos

Demos

[ ] Replicate Toggle
Replicate (What is Replicate?)
[ ] Spaces Toggle
Hugging Face Spaces (What is Spaces?)
[ ] Spaces Toggle
TXYZ.AI (What is TXYZ.AI?)
( ) Related Papers

Recommenders and Search Tools

[ ] Link to Influence Flower
Influence Flower (What are Influence Flowers?)
[ ] Core recommender toggle
CORE Recommender (What is CORE?)
[ ] IArxiv recommender toggle
IArxiv Recommender (What is IArxiv?)

  * Author
  * Venue
  * Institution
  * Topic

( ) About arXivLabs

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and
share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have
embraced and accepted our values of openness, community, excellence,
and user data privacy. arXiv is committed to these values and only
works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community?
Learn more about arXivLabs.

Which authors of this paper are endorsers? | Disable MathJax (What is
MathJax?)

  * About
  * Help

  * Click here to contact arXiv Contact
  * Click here to subscribe Subscribe

  * Copyright
  * Privacy Policy

  * Web Accessibility Assistance
  * arXiv Operational Status
    Get status notifications via email or slack