Sched Fast Diffusion Decoding

Accelerate diffusion LLM decoding by halting when confidence stabilizes using progress-dependent thresholds. SchED achieves 3.8-4.0× speedup while retaining 99.8-100% quality—training-free and model-agnostic for any dLLM.

adu2021 Updated 7 repo stars

File contents

adu2021/skillxiv commit ae91fd0d94

Frequently asked questions

npx get-skill add adu2021/sched-fast-diffusion-decoding