Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2502.00987 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909534679728128 |
|---|---|
| author | Albert, Paul Zhang, Frederic Z. Saratchandran, Hemanth Rodriguez-Opazo, Cristian Hengel, Anton van den Abbasnejad, Ehsan |
| author_facet | Albert, Paul Zhang, Frederic Z. Saratchandran, Hemanth Rodriguez-Opazo, Cristian Hengel, Anton van den Abbasnejad, Ehsan |
| contents | Low-Rank Adaptation (LoRA) and its variants have shown impressive results in reducing the number of trainable parameters and memory requirements of large transformer networks while maintaining fine-tuning performance. The low-rank nature of the weight update inherently limits the representation power of fine-tuned models, however, thus potentially compromising performance on complex tasks. This raises a critical question: when a performance gap between LoRA and standard fine-tuning is observed, is it due to the reduced number of trainable parameters or the rank deficiency? This paper aims to answer this question by introducing RandLoRA, a parameter-efficient method that performs full-rank updates using a learned linear combinations of low-rank, non-trainable random matrices. Our method limits the number of trainable parameters by restricting optimization to diagonal scaling matrices applied to the fixed random matrices. This allows us to effectively overcome the low-rank limitations while maintaining parameter and memory efficiency during training. Through extensive experimentation across vision, language, and vision-language benchmarks, we systematically evaluate the limitations of LoRA and existing random basis methods. Our findings reveal that full-rank updates are beneficial across vision and language tasks individually, and even more so for vision-language tasks, where RandLoRA significantly reduces -- and sometimes eliminates -- the performance gap between standard fine-tuning and LoRA, demonstrating its efficacy. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_00987 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | RandLoRA: Full-rank parameter-efficient fine-tuning of large models Albert, Paul Zhang, Frederic Z. Saratchandran, Hemanth Rodriguez-Opazo, Cristian Hengel, Anton van den Abbasnejad, Ehsan Computation and Language Artificial Intelligence Computer Vision and Pattern Recognition Low-Rank Adaptation (LoRA) and its variants have shown impressive results in reducing the number of trainable parameters and memory requirements of large transformer networks while maintaining fine-tuning performance. The low-rank nature of the weight update inherently limits the representation power of fine-tuned models, however, thus potentially compromising performance on complex tasks. This raises a critical question: when a performance gap between LoRA and standard fine-tuning is observed, is it due to the reduced number of trainable parameters or the rank deficiency? This paper aims to answer this question by introducing RandLoRA, a parameter-efficient method that performs full-rank updates using a learned linear combinations of low-rank, non-trainable random matrices. Our method limits the number of trainable parameters by restricting optimization to diagonal scaling matrices applied to the fixed random matrices. This allows us to effectively overcome the low-rank limitations while maintaining parameter and memory efficiency during training. Through extensive experimentation across vision, language, and vision-language benchmarks, we systematically evaluate the limitations of LoRA and existing random basis methods. Our findings reveal that full-rank updates are beneficial across vision and language tasks individually, and even more so for vision-language tasks, where RandLoRA significantly reduces -- and sometimes eliminates -- the performance gap between standard fine-tuning and LoRA, demonstrating its efficacy. |
| title | RandLoRA: Full-rank parameter-efficient fine-tuning of large models |
| topic | Computation and Language Artificial Intelligence Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2502.00987 |