rank self-pruning

on-device foundation model (1) heterogeneous rank adaptation (1) heterogeneous training (1) edge computing (1) low-rank adaptation (1) parameter-efficient fine-tuning (1) foundation model (1) federated learning (1) model compression (1)