Implement speculative decoding: draft $k$ tokens, target verifies in one pass, ac

Hard

Inference & Serving4 / 5

Chapter 244 / 5

Inference & Serving

Implement speculative decoding: draft $k$ tokens, target verifies in one pass, ac

Hard

Implement speculative decoding: draft $k$ tokens, target verifies in one pass, accept/reject with the correct probabilistic rule.

Target interface

residual_sample(p, q, rng)

Implement the function/class skeleton in the editor. Any correct approach is accepted.

Hints0 / 2

Reference solution

Your own code stays in the editor.

solution.pyPython

Local draft

import numpy as np

def residual_sample(p, q, rng):
    raise NotImplementedError

def speculative_step(prefix, draft_dist, target_dists, k, rng):
    raise NotImplementedError

⌘/Ctrl↵to submit

AI review

No review yet

Submit your solution for a verdict, a score out of 100 and concrete feedback. Any correct approach passes.

Chapter 244 / 5

Inference & Serving

Implement speculative decoding: draft $k$ tokens, target verifies in one pass, ac

Hard

Implement speculative decoding: draft $k$ tokens, target verifies in one pass, accept/reject with the correct probabilistic rule.

Target interface

residual_sample(p, q, rng)

Implement the function/class skeleton in the editor. Any correct approach is accepted.

Hints0 / 2

Reference solution

Your own code stays in the editor.