Revisiting Zeroth-Order Optimization for Memory-Efficient LLM Fine-Tuning: A Benchmark Paper • 2402.11592 • Published Feb 18 • 2