6 papers
Complete-muE: Optimal Hyperparameter Transfer and Scaling for MoE Models
Hongwu Peng, Ohiremen Dibua, Yuanjun Xiong +3
We propose Complete-muE, a framework which targets hyperparameter transfer across dense FFN and any Mixture-of-Experts (MoE) setups in transformer blocks. Existing tools such as $Î…
AlignTok: Aligning Visual Foundation Encoders to Tokenizers for Diffusion Models
Bowei Chen, Sai Bi, Hao Tan +6
In this work, we propose aligning pretrained visual encoders to serve as tokenizers for latent diffusion models in image generation. Unlike training a variational autoencoder (VAE)…
Staggered Batch Scheduling: Co-optimizing Time-to-First-Token and Throughput for High-Efficiency LLM Inference
Jian Tian, Shuailong Li, Yang Cao +8
The evolution of Large Language Model (LLM) serving towards complex, distributed architectures--specifically the P/D-separated, large-scale DP+EP paradigm--introduces distinct sche…
ESS: An Offload-Centric Latent-Cache Management Architecture for DeepSeek-V3.2-Exp
Xinhang Chen, Chao Zhang, Jiahuan He +9
DeepSeek-V3.2-Exp introduces a sparse attention mechanism that significantly reduces inference latency in long-context scenarios. Although the overall throughput has improved great…
KnapFormer: An Online Load Balancer for Efficient Diffusion Transformers Training
Kai Zhang, Peng Wang, Sai Bi +2
We present KnapFormer, an efficient and versatile framework to combine workload balancing and sequence parallelism in distributed training of Diffusion Transformers (DiT). KnapForm…
EP-CFG: Energy-Preserving Classifier-Free Guidance
Kai Zhang, Fujun Luan, Sai Bi +1
Classifier-free guidance (CFG) is widely used in diffusion models but often introduces over-contrast and over-saturation artifacts at higher guidance strengths. We present EP-CFG (…