huggingface--peft/docs/source/package_reference/hra.md

Bridging The Gap between Low-rank and Orthogonal Adaptation via Householder Reflection Adaptation (HRA)

Bridging The Gap between Low-rank and Orthogonal Adaptation via Householder Reflection Adaptation

HRA provides a new perspective connecting LoRA to OFT, which means it can harness the advantages of both strategies, by leveraging Householder reflections to reduce parameters and computation costs while penalizing the loss of pre-training knowledge. It consistently achieves better performance with fewer trainable parameters and outperforms state-of-the-art adapters across different models, including large language models (LLMs) and conditional image generators.

HRA constructs a chain of r trainable Householder reflections (HRs). Because the Householder reflection matrix is an orthogonal matrix and the product of orthogonal matrices is also an orthogonal matrix, HRA satisfies the theoretical guarantee of Orthogonal Finetuning (OFT). Meanwhile, HRA can also be viewed as a low-rank fine-tuning adapter. The higher r, the more trainable parameters, resulting in a larger model capacity and better performance. Besides, due to the chain structure, the orthogonality of HR planes impacts the capacity and regularity of HRA. To achieve a trade-off between the model capacity and regularity, an orthogonality regularizer of the HR planes is added to the loss function. The weight \(\lambda\) can control the strength of the regularizer.

The abstract from the paper is:

While following different technical routes, both low-rank and orthogonal adaptation techniques can efficiently adapt large-scale pre-training models in specific tasks or domains based on a small piece of trainable parameters. In this study, we bridge the gap between these two techniques, proposing a simple but effective adaptation method based on Householder reflections. Given a pre-trained model, our method fine-tunes its layers by multiplying each frozen weight matrix with an orthogonal matrix constructed by a chain of learnable Householder reflections (HRs). This HR-based orthogonal fine-tuning is equivalent to an adaptive low-rank adaptation. Moreover, we show that the orthogonality of the reflection planes corresponding to the HRs impacts the model capacity and regularity. The analysis motivates us to regularize the orthogonality of the HRs, leading to different implementations of the proposed Householder reflection adaptation (HRA) method. Compared with state-of-the-art methods, HRA achieves superior performance with fewer learnable parameters when adapting large language models and conditional image generators. The code is available at peft and HRA.

API

HRAConfig

autodoc tuners.hra.config.HRAConfig

HRAModel

autodoc tuners.hra.model.HRAModel