Papers
arxiv:2503.10377

SPPO:Efficient Long-sequence LLM Training via Adaptive Sequence Pipeline Parallel Offloading

Published on Mar 13, 2025
Authors:
,
,
,
,
,

Abstract

Adaptive Sequence Pipeline Parallel Offloading (SPPO) framework enhances training efficiency for Large Language Models with long input sequences by optimizing GPU memory and computational resource usage.

In recent years, Large Language Models (LLMs) have exhibited remarkable capabilities, driving advancements in real-world applications. However, training LLMs on increasingly long input sequences imposes significant challenges due to high GPU memory and computational demands. Existing solutions face two key limitations: (1) memory reduction techniques, such as activation recomputation and CPU offloading, compromise training efficiency; (2) distributed parallelism strategies require excessive GPU resources, limiting the scalability of input sequence length. To address these gaps, we propose Adaptive Sequence Pipeline Parallel Offloading (SPPO), a novel LLM training framework that optimizes memory and computational resource efficiency for long-sequence training. SPPO introduces adaptive offloading, leveraging sequence-aware offloading, and two-level activation management to reduce GPU memory consumption without degrading the training efficiency. Additionally, SPPO develops an adaptive pipeline scheduling approach with a heuristic solver and multiplexed sequence partitioning to improve computational resource efficiency. Experimental results demonstrate that SPPO achieves up to 3.38x throughput improvement over Megatron-LM and DeepSpeed, realizing efficient training of a 7B LLM with sequence lengths of up to 4M tokens on only 128 A100 GPUs.

Community

Sign up or log in to comment

Get this paper in your agent:

hf papers read 2503.10377
Don't have the latest CLI?
curl -LsSf https://hf.co/cli/install.sh | bash

Models citing this paper 0

No model linking this paper

Cite arxiv.org/abs/2503.10377 in a model README.md to link it from this page.

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2503.10377 in a dataset README.md to link it from this page.

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2503.10377 in a Space README.md to link it from this page.

Collections including this paper 0

No Collection including this paper

Add this paper to a collection to link it from this page.