Pith. sign in

REVIEW 1 cited by

Boosting Lossless Speculative Decoding via Feature Sampling and Partial Alignment Distillation

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2408.15562 v1 pith:F73MHN5I submitted 2024-08-28 cs.CL cs.LG

classification cs.CLcs.LG
keywords modeldecodingdrafttargetalignmentdistillationfspadlossless
verification ladder T0 review T1 audit T2 compute T3 formal

Signed reviews

No signed human review yet.

0 comments
read the original abstract

Lossless speculative decoding accelerates target large language model (LLM) inference by employing a lightweight draft model for generating tree-structured candidates, which are subsequently verified in parallel by the target LLM. Currently, effective approaches leverage feature-level rather than token-level autoregression within the draft model to facilitate more straightforward predictions and enhanced knowledge distillation. In this paper, we reassess these approaches and propose FSPAD (Feature Sampling and Partial Alignment Distillation for Lossless Speculative Decoding), which introduces two straightforward and effective components within the existing framework to boost lossless speculative decoding. Firstly, FSPAD utilizes token embeddings to sample features of the target LLM in high-dimensional space before feeding them into the draft model, due to the inherent uncertainty of the features preventing the draft model from obtaining the specific token output by the target LLM. Secondly, FSPAD introduces partial alignment distillation to weaken the draft model's connection between features and logits, aiming to reduce the conflict between feature alignment and logit confidence during training. Our experiments include both greedy and non-greedy decoding on the largest and smallest models from the Vicuna and LLaMA3-Instruct series, as well as tasks in multi-turn conversation, translation, summarization, question answering, mathematical reasoning, and retrieval-augmented generation. The results show that FSPAD outperforms the state-of-the-art method across all the aforementioned tasks and target LLMs.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. ZipAR: Parallel Auto-regressive Image Generation through Spatial Locality

    cs.CV 2024-12 conditional novelty 6.0 of 10

    ZipAR is a training-free decoding method that uses spatial locality to generate multiple visual tokens per forward pass, cutting autoregressive image generation steps by up to 91%.

Pith tools