An Empirical Study of Parameter Efficient Fine-tuning on Vision-Language Pre-train Model
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866909206646358016 |
|---|---|
| author | Tian, Yuxin Yang, Mouxing Li, Yunfan Liu, Dayiheng Ren, Xingzhang Peng, Xi Lv, Jiancheng |
| author_facet | Tian, Yuxin Yang, Mouxing Li, Yunfan Liu, Dayiheng Ren, Xingzhang Peng, Xi Lv, Jiancheng |
| contents | Recent studies applied Parameter Efficient Fine-Tuning techniques (PEFTs) to efficiently narrow the performance gap between pre-training and downstream. There are two important factors for various PEFTs, namely, the accessible data size and fine-tunable parameter size. A natural expectation for PEFTs is that the performance of various PEFTs is positively related to the data size and fine-tunable parameter size. However, according to the evaluation of five PEFTs on two downstream vision-language (VL) tasks, we find that such an intuition holds only if the downstream data and task are not consistent with pre-training. For downstream fine-tuning consistent with pre-training, data size no longer affects the performance, while the influence of fine-tunable parameter size is not monotonous. We believe such an observation could guide the choice of training strategy for various PEFTs. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2403_08433 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | An Empirical Study of Parameter Efficient Fine-tuning on Vision-Language Pre-train Model Tian, Yuxin Yang, Mouxing Li, Yunfan Liu, Dayiheng Ren, Xingzhang Peng, Xi Lv, Jiancheng Computer Vision and Pattern Recognition Recent studies applied Parameter Efficient Fine-Tuning techniques (PEFTs) to efficiently narrow the performance gap between pre-training and downstream. There are two important factors for various PEFTs, namely, the accessible data size and fine-tunable parameter size. A natural expectation for PEFTs is that the performance of various PEFTs is positively related to the data size and fine-tunable parameter size. However, according to the evaluation of five PEFTs on two downstream vision-language (VL) tasks, we find that such an intuition holds only if the downstream data and task are not consistent with pre-training. For downstream fine-tuning consistent with pre-training, data size no longer affects the performance, while the influence of fine-tunable parameter size is not monotonous. We believe such an observation could guide the choice of training strategy for various PEFTs. |
| title | An Empirical Study of Parameter Efficient Fine-tuning on Vision-Language Pre-train Model |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2403.08433 |