Faster Language Models with Better Multi-Token Prediction Using Tensor Decomposition

TL;DR

This paper introduces a tensor decomposition-based method for multi-token prediction in transformers, significantly improving inference speed and efficiency in text and code generation tasks while maintaining accuracy.

Contribution

It generalizes rank-1 tensor decomposition to a rank-r model for simultaneous multi-token prediction, integrating mixture of experts for efficient training.

Findings

Significant speedup in inference for text and code generation

Effective across various model sizes and training epochs

Low overhead for training and sampling

Abstract

We propose a new model for multi-token prediction in transformers, aiming to enhance sampling efficiency without compromising accuracy. Motivated by recent work that predicts the probabilities of subsequent tokens using multiple heads, we connect this approach to rank-$1$ canonical tensor decomposition. By generalizing it to a rank-$r$ canonical probability decomposition, we develop an improved model that predicts multiple tokens simultaneously. This model can also be interpreted as a mixture of experts, allowing us to leverage successful techniques from that domain for efficient and robust training. Importantly, the overall overhead for training and sampling remains low. Our method demonstrates significant improvements in inference speed for both text and code generation tasks, proving particularly beneficial within the self-speculative decoding paradigm. It maintains its effectiveness across various model sizes and training epochs, highlighting its robustness and scalability.