Fast-dLLM++
PulseAugur coverage of Fast-dLLM++ — every cluster mentioning Fast-dLLM++ across labs, papers, and developer communities, ranked by signal.
1 day(s) with sentiment data
-
New 'early-bird' decoding speeds up diffusion LLMs
Researchers have developed a new framework called "early-bird (EB)" decoding to significantly accelerate inference for diffusion large language models (dLLMs). This method addresses the inefficiency of dLLMs, which ofte…
-
New methods accelerate LLM inference speed via speculative decoding
Two new research papers introduce novel methods for accelerating the inference speed of large language models. The first paper, "ReTrace," proposes a technique that conditions each draft block on the rejected suffix fro…
-
Diffusion Language Models: Efficiency, Robustness, and Routing Innovations
Recent research explores advancements in diffusion language models (DLMs), focusing on improving their efficiency and robustness. One paper introduces Expert-Choice Routing as a superior alternative to Token-Choice Rout…
-
BlockServe framework boosts dLLM serving throughput by up to 10.6x
Researchers have developed BlockServe, a new framework designed to improve the efficiency of serving diffusion large language models (dLLMs). This system addresses the challenge of convergence heterogeneity in batch pro…
-
BlockServe framework boosts dLLM serving throughput by up to 10.6x
Researchers have developed BlockServe, a novel framework designed to improve the efficiency of serving diffusion large language models (dLLMs). This system addresses the challenge of heterogeneous convergence rates in b…
-
Sangam system optimizes serving for diffusion language models
Researchers have developed Sangam, a new serving system designed to efficiently handle diffusion language models (dLLMs). Unlike traditional autoregressive models, dLLMs generate text iteratively and have bidirectional …