Multi-Token Prediction via Self-Distillation

Abstract: Existing techniques for accelerating LLM inference, such as speculative decoding, require training auxiliary speculator models and building and deploying complex inference pipelines. We consider a new approach for converting a pretrained autoregressive LLM from a slow single next token prediction model into a fast standalone multi-token prediction model using a simple online distillation objective. The final model retains the exact same implementation as the pretrained initial checkpoint and is deployable without the addition of any auxiliary verifier or other specialized inference code. On GSM8K, our method produces models that can decode more than $3\times$ faster on average at $<5\%$ drop in accuracy relative to single token decoding performance.