Every story and comment from the Hacker News API, in a layout borrowed from 2001.
A comment on Lossless Acceleration of LLM via Adaptive N-Gram Parallel Decoding. Parent comment
You're either thinking of speculative encoding more generally, or Medusa: https://arxiv.org/abs/2401.10774
There are no comments on this comment yet.
You're either thinking of speculative encoding more generally, or Medusa: https://arxiv.org/abs/2401.10774