A cheap way to fine-tune a model by training a small add-on layer instead of changing all of the model's weights.
Full fine-tuning updates billions of parameters, which needs serious hardware. LoRA (low-rank adaptation) freezes the original model and trains tiny adapter matrices alongside it — often under 1% of the size — that nudge the model's behavior.
The result: fine-tuning that runs on a single GPU, adapters you can swap like plugins, and one base model serving many customized variants. It's the main reason personalized open-weight models are affordable at all.