Communication-Efficient LLM Adaptation over Decentralized GPU Meshes
cs.LG
Submitted: 2026-09-13
Updated: 2026-09-13
Comments: 38 pages, 2 figures, 12 tables
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- AsyncMesh: Fully Asynchronous Optimization for Data and Pipeline Parallelism
- DiLoCo: Distributed Low-Communication Training of Language Models
- Streaming DiLoCo with overlapping communication: Towards a Distributed Free Lunch
- The Llama 3 Herd of Models
- Gradient Descent Happens in a Tiny Subspace
- TAH-QUANT: Effective Activation Quantization in Pipeline Parallelism over Slow Network
- OpenDiLoCo: An Open-Source Framework for Globally Distributed Low-Communication Training
- ReLoRA: High-Rank Training Through Low-Rank Updates
- The AdEMAMix Optimizer: Better, Faster, Older
- DeMo: Decoupled Momentum Optimization
- Photon: Federated LLM Pre-Training
- DAPO: An Open-Source LLM Reinforcement Learning System at Scale
- Group Sequence Policy Optimization
Related papers
- Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
- AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
- Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
- Convergence issues in Relational Concept Analysis based on AOC-posets
- Beliefs Beyond Posteriors: Local-Consistency Optimisation for Bayesian Neural Networks
- Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks