SeedFlood: A Step Toward Scalable Decentralized Fine-Tuning of LLMs

arXiv:2602.18181 · cs.LG · Submitted 2026-02-20 · Read on arXiv

cs.LG

Submitted: 2026-02-20

Updated: 2026-10-01

Terminology

• Communication Overhead — 2× in this paper · explained in 2 episodes
• Fine-Tuning — 2× in this paper · explained in 5 episodes
• Generalization — 1× in this paper · explained in 13 episodes

Sources

• AdaGossip: Adaptive Consensus Step-size for Decentralized Deep Learning with Communication Compression
• Petals: Collaborative Inference and Fine-tuning of Large Models
• DeepZero: Scaling up Zeroth-Order Optimization for Deep Model Training
• Enhancing Zeroth-order Fine-tuning for Language Models with Low-rank Structures
• BoolQ: Exploring the Surprising Difficulty of Natural Yes/No Questions
• A Survey on Decentralized Federated Learning
• Variance-reduced Zeroth-Order Methods for Fine-Tuning Language Models
• Decentralized Low-Rank Fine-Tuning of Large Language Models
• Measuring Massive Multitask Language Understanding
• LoRA: Low-Rank Adaptation of Large Language Models
• Tulu 3: Pushing Frontiers in Open Language Model Post-Training
• Achieving Dimension-Free Communication in Federated Learning via Zeroth-Order Optimization
• Federated Full-Parameter Tuning of Billion-Sized Language Models with Communication Cost under 18 Kilobytes
• Ferret: Federated Full-Parameter Tuning at Scale for Large Language Models
• Refining Adaptive Zeroth-Order Optimization at Ease
• $\texttt{DeepSqueeze}$: Decentralization Meets Error-Compensated Compression
• PowerGossip: Practical Low-Rank Communication Compression in Decentralized Deep Learning
• ADF-LoRA: Alternating Low-Rank Aggregation for Decentralized Federated Fine-Tuning
• Decentralized SGD with Controlled Disagreement Finds Flatter Minima
• Qwen3 Technical Report

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: The Predictive Cost of ELBO Binding
• Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks