Pith. sign in

REVIEW 9 cited by

Simple and Scalable Strategies to Continually Pre-train Large Language Models

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2403.08763 v4 pith:SX2P6YOO submitted 2024-03-13 cs.LG cs.AIcs.CL

classification cs.LGcs.AIcs.CL
keywords datalearningre-trainingshiftdistributionenglishlanguagelarge
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
abstract

Large language models (LLMs) are routinely pre-trained on billions of tokens, only to start the process over again once new data becomes available. A much more efficient solution is to continually pre-train these models, saving significant compute compared to re-training. However, the distribution shift induced by new data typically results in degraded performance on previous data or poor adaptation to the new data. In this work, we show that a simple and scalable combination of learning rate (LR) re-warming, LR re-decaying, and replay of previous data is sufficient to match the performance of fully re-training from scratch on all available data, as measured by the final loss and the average score on several language model (LM) evaluation benchmarks. Specifically, we show this for a weak but realistic distribution shift between two commonly used LLM pre-training datasets (English$\rightarrow$English) and a stronger distribution shift (English$\rightarrow$German) at the $405$M parameter model scale with large dataset sizes (hundreds of billions of tokens). Selecting the weak but realistic shift for larger-scale experiments, we also find that our continual learning strategies match the re-training baseline for a 10B parameter LLM. Our results demonstrate that LLMs can be successfully updated via simple and scalable continual learning strategies, matching the re-training baseline using only a fraction of the compute. Finally, inspired by previous work, we propose alternatives to the cosine learning rate schedule that help circumvent forgetting induced by LR re-warming and that are not bound to a fixed token budget.

Discussion (0). Sign in to comment.

Forward citations

Cited by 9 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Dynamic Chunking for End-to-End Hierarchical Sequence Modeling

    cs.LG 2025-07 conditional novelty 7.0 of 10

    A learned dynamic chunking hierarchy lets byte-level language models match or beat BPE-tokenized Transformers at matched compute, with larger gains on Chinese, code, and DNA.

  2. WSqD: A Horizon-Free Learning Rate Schedule for Large Model Training

    cs.LG 2026-07 conditional novelty 6.0 of 10

    WSqD uses a horizon-independent shifted inverse-square-root base plus linear cooldown to get optimal last-iterate rates in convex stochastic optimization and match or beat tuned WSD on LLM pretraining with one reused ...

  3. Weight Decay Improves Language Model Plasticity

    cs.LG 2026-02 conditional novelty 6.0 of 10

    Pretrained models trained with larger weight decay fine-tune better on downstream tasks, so the best pretraining checkpoint by loss is not always the best starting point for later training.

  4. BioClinical ModernBERT: A State-of-the-Art Long-Context Encoder for Biomedical and Clinical NLP

    cs.CL 2025-06 conditional novelty 6.0 of 10

    A continued-pretrained ModernBERT encoder for biomedical and clinical text claims SOTA on several clinical NLP tasks, with caveats about data overlap between pretraining and evaluation.

  5. The Common Pile v0.1: An 8TB Dataset of Public Domain and Openly Licensed Text

    cs.CL 2025-06 conditional novelty 6.0 of 10

    A new 8TB openly-licensed text corpus trains 7B LLMs that are competitive with Llama 1/2, showing that performant models need not depend on unlicensed web data.

  6. GEM: Empowering LLM for both Embedding Generation and Language Understanding

    cs.CL 2025-06 conditional novelty 6.0 of 10

    GEM fine-tunes decoder-only LLMs with bottleneck special tokens and a mixed next-token prediction and contrastive objective, giving them text-embedding ability with only a modest MMLU drop.

  7. ZUNA1.1: A more flexible EEG foundation model for Denoising and Super-resolution

    cs.LG 2026-07 conditional novelty 5.0 of 10

    ZUNA1.1, an open-source 380M EEG diffusion autoencoder, reconstructs variable-length, flexibly masked EEG at least as well as its predecessor and far better than spherical spline interpolation.

  8. Dynamic Context-oriented Decomposition for Task-aware Low-rank Adaptation with Less Forgetting and Faster Convergence

    cs.LG 2025-06 conditional novelty 5.0 of 10

    CorDA++ uses data-driven SVD to initialize LoRA adapters, adding per-layer covariance selection and rank allocation that reduce forgetting and speed convergence compared to LoRA, PiSSA, QLoRA, and other baselines.

  9. Software Engineering for Large Language Models: Research Status, Challenges and the Road Ahead

    cs.SE 2025-06 conditional novelty 4.0 of 10

    A literature review organizes LLM development into a six-phase software engineering lifecycle and identifies challenges and research directions for each phase.

Pith tools