GraLoRA: Granular Low-Rank Adaptation for Parameter-Efficient Fine-Tuning

Re-assign community