Finetuning Stellar Spectra Foundation Models with LoRA
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909709410238464 |
|---|---|
| author | Zhao, Xiaosheng Ting, Yuan-Sen Szalay, Alexander S. Huang, Yang |
| author_facet | Zhao, Xiaosheng Ting, Yuan-Sen Szalay, Alexander S. Huang, Yang |
| contents | Foundation models are beginning to impact stellar spectroscopy, where spectra encode rich physical information in a structured, language-like form. A key challenge is adapting these models across heterogeneous surveys with differing resolution and coverage. We apply Low-Rank Adaptation (LoRA) to fine-tune SpecCLIP--a contrastively pre-trained model on LAMOST and Gaia XP spectra--for downstream tasks on DESI Early Data Release (EDR) spectra. We show that LoRA enables few-shot learning on DESI, with performance varying by fine-tuned module and benefiting from Gaia XP knowledge embedded in the pre-trained model. Our results demonstrate that LoRA provides a lightweight and effective strategy for extending spectral foundation models to new instruments and survey domains. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2507_20972 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Finetuning Stellar Spectra Foundation Models with LoRA Zhao, Xiaosheng Ting, Yuan-Sen Szalay, Alexander S. Huang, Yang Instrumentation and Methods for Astrophysics Solar and Stellar Astrophysics Foundation models are beginning to impact stellar spectroscopy, where spectra encode rich physical information in a structured, language-like form. A key challenge is adapting these models across heterogeneous surveys with differing resolution and coverage. We apply Low-Rank Adaptation (LoRA) to fine-tune SpecCLIP--a contrastively pre-trained model on LAMOST and Gaia XP spectra--for downstream tasks on DESI Early Data Release (EDR) spectra. We show that LoRA enables few-shot learning on DESI, with performance varying by fine-tuned module and benefiting from Gaia XP knowledge embedded in the pre-trained model. Our results demonstrate that LoRA provides a lightweight and effective strategy for extending spectral foundation models to new instruments and survey domains. |
| title | Finetuning Stellar Spectra Foundation Models with LoRA |
| topic | Instrumentation and Methods for Astrophysics Solar and Stellar Astrophysics |
| url | https://arxiv.org/abs/2507.20972 |