LoRA Provides Differential Privacy by Design via Random Sketching
Fuente:
arXiv
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911435773181952 |
|---|---|
| author | Malekmohammadi, Saber Farnadi, Golnoosh |
| author_facet | Malekmohammadi, Saber Farnadi, Golnoosh |
| contents | Low-rank adaptation of language models has been proposed to reduce the computational and memory overhead of fine-tuning pre-trained language models. LoRA incorporates trainable low-rank matrices into some parameters of the pre-trained model, called adapters. In this work, we show theoretically that the low-rank adaptation mechanism of LoRA is equivalent to fine-tuning adapters with noisy batch gradients, with the noise variance being a decreasing function of adaptation rank ($r$). Motivated by this understanding, we prove inherent differential privacy for LoRA when adaptation matrices $A_\ell$ are frozen. We show that various factors, e.g., the adaptation rank and batch size, affect the guaranteed privacy level. Our findings provide useful insights into LoRA and uncovers the reason behind the robustness of models fine-tuned with LoRA to privacy attacks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2409_17538 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | LoRA Provides Differential Privacy by Design via Random Sketching Malekmohammadi, Saber Farnadi, Golnoosh Machine Learning Artificial Intelligence Computation and Language Low-rank adaptation of language models has been proposed to reduce the computational and memory overhead of fine-tuning pre-trained language models. LoRA incorporates trainable low-rank matrices into some parameters of the pre-trained model, called adapters. In this work, we show theoretically that the low-rank adaptation mechanism of LoRA is equivalent to fine-tuning adapters with noisy batch gradients, with the noise variance being a decreasing function of adaptation rank ($r$). Motivated by this understanding, we prove inherent differential privacy for LoRA when adaptation matrices $A_\ell$ are frozen. We show that various factors, e.g., the adaptation rank and batch size, affect the guaranteed privacy level. Our findings provide useful insights into LoRA and uncovers the reason behind the robustness of models fine-tuned with LoRA to privacy attacks. |
| title | LoRA Provides Differential Privacy by Design via Random Sketching |
| topic | Machine Learning Artificial Intelligence Computation and Language |
| url | https://arxiv.org/abs/2409.17538 |