Fast Forwarding Low-Rank Training

TL;DR

Fast Forward is a novel optimization strategy that accelerates low-rank finetuning of language models by repeating recent optimizer steps, significantly reducing training FLOPs and time while maintaining performance.

Contribution

We introduce Fast Forward, a simple method that accelerates low-rank finetuning by repeating optimizer steps, achieving substantial computational savings.

Findings

Up to 87% reduction in FLOPs

Up to 81% reduction in training time

Maintains model performance during acceleration

Abstract

Parameter efficient finetuning methods like low-rank adaptation (LoRA) aim to reduce the computational costs of finetuning pretrained Language Models (LMs). Enabled by these low-rank settings, we propose an even more efficient optimization strategy: Fast Forward, a simple and effective approach to accelerate large segments of training. In a Fast Forward stage, we repeat the most recent optimizer step until the loss stops improving on a tiny validation set. By alternating between regular optimization steps and Fast Forward stages, Fast Forward provides up to an 87\% reduction in FLOPs and up to an 81\% reduction in train time over standard SGD with Adam. We validate Fast Forward by finetuning various models on different tasks and demonstrate that it speeds up training without compromising model performance. Additionally, we analyze when and how to apply Fast Forward.