7 papers
VVS: Accelerating Speculative Decoding for Visual Autoregressive Generation via Partial Verification Skipping
Haotian Dong, Ye Li, Rongwei Lu +3
Visual autoregressive (AR) generation models have demonstrated strong potential for image generation, yet their next-token-prediction paradigm introduces considerable inference lat…
Taming Latency and Bandwidth: A Theoretical Framework and Adaptive Algorithm for Communication-Constrained Training
Rongwei Lu, Jingyan Jiang, Chunyang Li +2
Regional energy caps limit the growth of any single data center used for large-scale model training. This single-center training paradigm works when model size remains manageable,…
Accelerating Parallel Diffusion Model Serving with Residual Compression
Jiajun Luo, Yicheng Xiao, Jianru Xu +5
Diffusion models produce realistic images and videos but require substantial computational resources, necessitating multi-accelerator parallelism for real-time deployment. However,…
Staleness-Centric Optimizations for Parallel Diffusion MoE Inference
Jiajun Luo, Lizhuo Luo, Jianru Xu +4
Mixture-of-Experts-based (MoE-based) diffusion models demonstrate remarkable scalability in high-fidelity image generation, yet their reliance on expert parallelism introduces crit…
Beyond A Single AI Cluster: A Survey of Decentralized LLM Training
Haotian Dong, Jingyan Jiang, Rongwei Lu +5
The emergence of large language models (LLMs) has revolutionized AI development, yet the resource demands beyond a single cluster or even datacenter, limiting accessibility to well…
-FedHT: Stepsize-Aware Hard-Threshold Gradient Compression in Federated Learning
Rongwei Lu, Yutong Jiang, Jinrui Zhang +4
Gradient compression can effectively alleviate communication bottlenecks in Federated Learning (FL). Contemporary state-of-the-art sparse compressors, such as Top-, exhibit high…