Fast-dLLM: Training-free Acceleration of Diffusion LLM by Enabling KV Cache and Parallel Decoding

Documenti analoghi

• Fast-dLLM v2: Efficient Block-Diffusion LLM
  di: Wu, Chengyue, et al.
  Pubblicazione: (2025)
• Sparse-dLLM: Accelerating Diffusion LLMs with Dynamic Cache Eviction
  di: Song, Yuerong, et al.
  Pubblicazione: (2025)
• dLLM-Cache: Accelerating Diffusion Large Language Models with Adaptive Caching
  di: Liu, Zhiyuan, et al.
  Pubblicazione: (2025)
• Streaming-dLLM: Accelerating Diffusion LLMs via Suffix Pruning and Dynamic Decoding
  di: Xiao, Zhongyu, et al.
  Pubblicazione: (2026)
• LoPA: Scaling dLLM Inference via Lookahead Parallel Decoding
  di: Xu, Chenkai, et al.
  Pubblicazione: (2025)