DePT: Decomposed Prompt Tuning for Parameter-Efficient Fine-tuning
This work addresses efficiency problems in parameter-efficient fine-tuning for large language models, which is crucial for applications with heavy daily querying, though it is incremental as it builds on existing prompt tuning methods.
The authors tackled the issue of increased training and inference time and memory usage in prompt tuning for large language models by proposing Decomposed Prompt Tuning (DePT), which decomposes soft prompts into shorter prompts and low-rank matrices, achieving better performance while saving memory and time costs compared to vanilla prompt tuning and other PEFT methods, as demonstrated on 23 NLP and vision-language tasks.
Prompt tuning (PT), where a small amount of trainable soft (continuous) prompt vectors is affixed to the input of language models (LM), has shown promising results across various tasks and models for parameter-efficient fine-tuning (PEFT). PT stands out from other PEFT approaches because it maintains competitive performance with fewer trainable parameters and does not drastically scale up its parameters as the model size expands. However, PT introduces additional soft prompt tokens, leading to longer input sequences, which significantly impacts training and inference time and memory usage due to the Transformer's quadratic complexity. Particularly concerning for Large Language Models (LLMs) that face heavy daily querying. To address this issue, we propose Decomposed Prompt Tuning (DePT), which decomposes the soft prompt into a shorter soft prompt and a pair of low-rank matrices that are then optimised with two different learning rates. This allows DePT to achieve better performance while saving substantial memory and time costs compared to vanilla PT and its variants, without changing trainable parameter sizes. Through extensive experiments on 23 natural language processing (NLP) and vision-language (VL) tasks, we demonstrate that DePT outperforms state-of-the-art PEFT approaches, including the full fine-tuning baseline, in some scenarios. Additionally, we empirically show that DEPT grows more efficient as the model size increases. Our further study reveals that DePT integrates seamlessly with parameter-efficient transfer learning in the few-shot learning setting and highlights its adaptability to various model architectures and sizes.