LoRA
Low-Rank Adaptation: fine-tuning tiny adapter matrices while freezing the base model — under 1% of parameters, single-GPU training, megabyte artifacts you can hot-swap per task.
Low-Rank Adaptation: fine-tuning tiny adapter matrices while freezing the base model — under 1% of parameters, single-GPU training, megabyte artifacts you can hot-swap per task.