Draft page

This page is an outline. It describes what will be covered and is not yet complete technical documentation.

Decoding is limited by memory bandwidth rather than compute, so a forward pass that verifies several candidate tokens costs little more than one that produces a single token. Speculative decoding exploits that.

This page explains the mechanism and the conditions under which it actually pays off.

What you will learn

  • Why verifying several tokens costs little more than generating one.
  • How draft-and-verify preserves the target model’s output distribution.
  • Which draft model choices work, including model-free variants.
  • When speculative decoding fails to help or hurts.

This page is an outline. The subsections below are the planned structure; they are filled in as the handbook is written.

The bandwidth-bound decode phase

Draft

Draft and verify

Draft

Preserving the output distribution

Draft

Choosing a draft model

Draft

Acceptance rate and speedup

Draft

When it does not pay off

Draft