Probe Pruning: Accelerating LLMs through Dynamic Pruning via Model-Probing

Open in new window