DFlash: Block Diffusion for Flash Speculative Decoding
Like z-lab/dflash, 0 likes 0 Bookmark z-lab/dflash, 0 bookmarks 0 Speculative decoding has plenty of tricks. Which ones actually make LLMs faster? Explore 7 methods, 26 checkpoints, and a shared training and evaluation framework.
Like ernie-research/DCore, 0 likes 0 Bookmark ernie-research/DCore, 0 bookmarks 0 Official Implementation of DART (DART: Low-Latency Parallel Drafting with Continuity-Aware Tree Pruning for Speculative Decoding, EMNLP26).
Like fvliang/DART, 0 likes 0 Bookmark fvliang/DART, 0 bookmarks 0 [COLMW'26 Spotlight] AdaFlash: Adaptive Speculative Decoding via On-Policy Distilled Diffusion Drafters
Like ZinYY/AdaFlash, 0 likes 0 Bookmark ZinYY/AdaFlash, 0 bookmarks 0 A lightweight inference engine supporting speculative speculative decoding (SSD).
Like tanishqkumar/ssd, 0 likes 0 Bookmark tanishqkumar/ssd, 0 bookmarks 0 DeepSpec: a full-stack codebase for training and evaluating speculative decoding algorithms
Like deepseek-ai/DeepSpec, 0 likes 0 Bookmark deepseek-ai/DeepSpec, 0 bookmarks 0 A unified library for building, evaluating, and storing speculative decoding algorithms for LLM inference in vLLM
Like vllm-project/speculators, 0 likes 0 Bookmark vllm-project/speculators, 0 bookmarks 0 Official Implementation of EAGLE-1 (ICML'24), EAGLE-2 (EMNLP'24), and EAGLE-3 (NeurIPS'25).
Like SafeAILab/EAGLE, 0 likes 0 Bookmark SafeAILab/EAGLE, 0 bookmarks 0 JetSpec++/JetSpec: Causal Parallel (Tree) Drafting with Self Refinement
Like hao-ai-lab/JetSpec, 0 likes 0 Bookmark hao-ai-lab/JetSpec, 0 bookmarks 0 
