Concept
SuffixDecoding
A model-free speculative decoding method that maintains a dynamic suffix tree and uses the longest suffix match of the current generated text against historical continuations as the draft. It is unaware of changes in the policy distribution during training.
Timeline (last 90 days)
Oct 26
The draft sources of SuffixDecoding and prompt-lookup decoding are static or semi-static text and are unaware of changes in the policy distribution during training.
Unverified50%
Oct 26
NgramRL's average acceptance length is up to 1.99 times higher than SuffixDecoding's.
Unverified50%