Finetuning Stellar Spectra Foundation Models with LoRA

TL;DR

This paper demonstrates that Low-Rank Adaptation (LoRA) effectively fine-tunes stellar spectral foundation models, enabling few-shot learning across different surveys and instruments, thus broadening their applicability in astrophysics.

Contribution

The study introduces LoRA as a lightweight method to adapt spectral foundation models to new surveys and instruments, improving flexibility and performance.

Findings

LoRA enables effective few-shot learning on DESI spectra.

Pre-trained Gaia XP knowledge enhances fine-tuning results.

LoRA provides a lightweight adaptation strategy for spectral models.

Abstract

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.