Parameter-Efficient Multi-Task Learning via Progressive Task-Specific Adaptation
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866917435457667072 |
|---|---|
| author | Gangwar, Neeraj Rangi, Anshuka Deshmukh, Rishabh Rahmanian, Holakou Dattatreya, Yesh Kani, Nickvash |
| author_facet | Gangwar, Neeraj Rangi, Anshuka Deshmukh, Rishabh Rahmanian, Holakou Dattatreya, Yesh Kani, Nickvash |
| contents | Parameter-efficient fine-tuning methods have emerged as a promising solution for adapting pre-trained models to various downstream tasks. While these methods perform well in single-task learning, extending them to multi-task learning exacerbates common issues, such as task interference and negative transfer, due to the limited number of trainable parameters. To address these challenges, we introduce progressive task-specific multi-task adaptation, a novel parameter-efficient approach for multi-task learning. Our approach introduces adapter modules that are shared in early layers and become increasingly task-specific in later layers. Additionally, we propose a gradient-based approach for computing task similarity and use this measure to allocate similar tasks to the shared adapter modules. To evaluate our approach, we adapt Swin and Pyramid Vision Transformers on PASCAL and NYUD-v2. On both datasets, our approach outperforms prior parameter-efficient multi-task methods while using fewer trainable parameters. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_19602 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Parameter-Efficient Multi-Task Learning via Progressive Task-Specific Adaptation Gangwar, Neeraj Rangi, Anshuka Deshmukh, Rishabh Rahmanian, Holakou Dattatreya, Yesh Kani, Nickvash Computer Vision and Pattern Recognition Parameter-efficient fine-tuning methods have emerged as a promising solution for adapting pre-trained models to various downstream tasks. While these methods perform well in single-task learning, extending them to multi-task learning exacerbates common issues, such as task interference and negative transfer, due to the limited number of trainable parameters. To address these challenges, we introduce progressive task-specific multi-task adaptation, a novel parameter-efficient approach for multi-task learning. Our approach introduces adapter modules that are shared in early layers and become increasingly task-specific in later layers. Additionally, we propose a gradient-based approach for computing task similarity and use this measure to allocate similar tasks to the shared adapter modules. To evaluate our approach, we adapt Swin and Pyramid Vision Transformers on PASCAL and NYUD-v2. On both datasets, our approach outperforms prior parameter-efficient multi-task methods while using fewer trainable parameters. |
| title | Parameter-Efficient Multi-Task Learning via Progressive Task-Specific Adaptation |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2509.19602 |