arxiv preprint - One-for-All: Generalized LoRA for Parameter-Efficient Fine-tuning | AI Breakdown | Podwise