arXiv:2506.10911v2 Announce Type: replace Abstract: Training large language models is generally done on clusters containing thousands of accelerators, communicating over a high-bandwidth interconnect...
arXiv:2508. 15706v3 Announce Type: replace Abstract: Communication-efficient distributed training algorithms (e.
arXiv:2606. 11081v1 Announce Type: cross Abstract: Communication-efficient pre-training of LLMs is increasingly important as training draws on compute distributed across clusters, data centers, and lower-bandwidth links.
arXiv:2607. 01678v1 Announce Type: new Abstract: Communication increasingly dominates the cost of Large Language Model (LLM) pre-training, especially under data-parallel and sharded training schemes, where gradient synchronization and parameter reconstruction overhead increase with model size and system scale.
Communication-efficient pre-training of LLMs is increasingly important as training draws on compute distributed across clusters, data centers, and lower-bandwidth links. Many practical methods reduce communication frequency but still rely on synchronous All-Reduce operations that maintain identical model states and tie progress to global collectives.
arXiv:2607. 03011v1 Announce Type: cross Abstract: Model merging techniques, which aggregate independently finetuned models into one to combine their capabilities, have become a topic of significant interest in recent years, with a broad array of methods having been proposed to tackle this problem.
arXiv:2502. 11034v3 Announce Type: replace Abstract: Loss spikes remain a persistent obstacle in large-scale language model pretraining.
arXiv:2602. 21788v2 Announce Type: replace-cross Abstract: Scaling long-context capabilities is crucial for Large Language Models (LLMs).
arXiv:2602. 04396v2 Announce Type: replace-cross Abstract: Distributed training of foundation models via $\texttt{DDP}$ is limited by interconnect bandwidth.
arXiv:2609.36830v1 Announce Type: new Abstract: Fully asynchronous reinforcement learning (RL) improves resource utilization in large language model post-training by overlapping rollout generation wi...
The paper proposes a theoretical framework for scheduling high‑quality data in large language model training by extending functional scaling laws to account for time‑varying data quality. It identifies two regimes—noise‑limited and signal‑limited—where high‑quality data should be used differently, and introduces a Drop‑Stable‑Rampup training schedule that adjusts batch size at the quality transition. Experiments on 15B MoE and 600M dense models show significant accuracy gains over conventional decay schedules across multiple benchmarks.
arXiv:2608. 06025v1 Announce Type: new Abstract: In simulation-in-the-loop decision-making systems, reinforcement learning (RL) inference is often constrained by simulator-side execution overhead, where workloads are highly dynamic and sensitive to runtime thread configurations.