Effect of fine-tuning sample size (0, 100, 200, 500, 1000, and 2069, where zero denotes zero-shot performance), evaluated on the testing sample. Shown are R² and the robustly estimated standard deviation of the residuals (σ) for [Fe/H] and [α/Fe], obtained after fine-tuning (residual-head fine-tuning for [Fe/H] and LoRA fine-tuning for [α/Fe]). For both metrics, we plot the mean and 1σ error bars from five independent runs. The scatterplots for the different sample sizes are shown in Figures A1 and A2.