12 papers
Beyond the Target: From Imitation to Collaboration in Speculative Decoding
Jinze Li, Yixing Xu, Guanchen Li +7
Speculative decoding (SPD) accelerates large language model (LLM) inference by letting a smaller draft model propose multiple future tokens that are verified in parallel by a large…
Theory-optimal Quantization Based on Flatness
Xiusheng Huang, Zhe Li, Xuanwu Yin +5
Post-training quantization has emerged as a widely adopted technique for compressing and accelerating the inference of Large Language Models (LLMs). The primary challenges in LLMs…
Learnable Permutation for Structured Sparsity on Transformer Models
Zekai Li, Ji Liu, Guanchen Li +5
Structured sparsity has emerged as a popular model pruning technique, widely adopted in various architectures, including CNNs, Transformer models, and especially large language mod…
Dual LoRA: Enhancing LoRA with Magnitude and Direction Updates
Yixing Xu, Chao Li, Xuanwu Yin +4
Low-rank adaptation (LoRA) is one of the most popular methods among parameter-efficient fine-tuning (PEFT) methods to adapt pre-trained large language models (LLMs) to specific dow…
Training-Free Loosely Speculative Decoding: Accepting Semantically Correct Drafts Beyond Exact Match
Jinze Li, Yixing Xu, Guanchen Li +6
Large language models (LLMs) achieve strong performance across diverse tasks but suffer from high inference latency due to their autoregressive generation. Speculative Decoding (SP…
SpecVLM: Fast Speculative Decoding in Vision-Language Models
Haiduo Huang, Fuwei Yang, Zhenhua Liu +4
Speculative decoding is a powerful way to accelerate autoregressive large language models (LLMs), but directly porting it to vision-language models (VLMs) faces unique systems cons…
SparK: Query-Aware Unstructured Sparsity with Recoverable KV Cache Channel Pruning
Huanxuan Liao, Yixing Xu, Shizhu He +6
Long-context inference in large language models (LLMs) is increasingly constrained by the KV cache bottleneck: memory usage grows linearly with sequence length, while attention com…
Athena: Enhancing Multimodal Reasoning with Data-efficient Process Reward Models
Shuai Wang, Zhenhua Liu, Jiaheng Wei +3
We present Athena-PRM, a multimodal process reward model (PRM) designed to evaluate the reward score for each step in solving complex reasoning problems. Developing high-performanc…
Gumiho: A Hybrid Architecture to Prioritize Early Tokens in Speculative Decoding
Jinze Li, Yixing Xu, Haiduo Huang +4
Speculative decoding (SPD) aims to accelerate the auto-regressive token generation process of a target Large Language Model (LLM). Some approaches employ a draft model with multipl…
Týr-the-Pruner: Structural Pruning LLMs via Global Sparsity Distribution Optimization
Guanchen Li, Yixing Xu, Zeping Li +4
Structural pruning enhances hardware-agnostic inference efficiency for large language models (LLMs) yet often fails to maintain comparable performance. Local pruning performs effic…
Jakiro: Boosting Speculative Decoding with Decoupled Multi-Head via MoE
Haiduo Huang, Fuwei Yang, Zhenhua Liu +7
Speculative decoding (SD) accelerates large language model inference by using a smaller draft model to predict multiple tokens, which are then verified in parallel by the larger ta…
SDXS: Real-Time One-Step Latent Diffusion Models with Image Conditions
Yuda Song, Zehao Sun, Xuanwu Yin
Recent advancements in diffusion models have positioned them at the forefront of image generation. Despite their superior performance, diffusion models are not without drawbacks; t…