3 papers
cs.LG2026
SubZero+: Efficient Zeroth-Order LLM Fine-Tuning via Large Learning Rates
Ziming Yu, Shuyao Xiao, Xingyu Zhao +6
Zeroth-order (ZO) optimization enables backpropagation-free fine-tuning of large language models, but existing ZO methods suffer from high-variance gradient estimators, making conv…
cs.LG2024
Zeroth-Order Fine-Tuning of LLMs in Random Subspaces
Ziming Yu, Pan Zhou, Sike Wang +3
Fine-tuning Large Language Models (LLMs) has proven effective for a variety of downstream tasks. However, as LLMs grow in size, the memory demands for backpropagation become increa…
cs.LG2024
4-bit Shampoo for Memory-Efficient Network Training
Sike Wang, Pan Zhou, Jia Li +1
Second-order optimizers, maintaining a matrix termed a preconditioner, are superior to first-order optimizers in both theory and practice. The states forming the preconditioner and…