all AI news
Sparse MeZO: Less Parameters for Better Performance in Zeroth-Order LLM Fine-Tuning
Feb. 27, 2024, 5:41 a.m. | Yong Liu, Zirui Zhu, Chaoyu Gong, Minhao Cheng, Cho-Jui Hsieh, Yang You
cs.LG updates on arXiv.org arxiv.org
Abstract: While fine-tuning large language models (LLMs) for specific tasks often yields impressive results, it comes at the cost of memory inefficiency due to back-propagation in gradient-based training. Memory-efficient Zeroth-order (MeZO) optimizers, recently proposed to address this issue, only require forward passes during training, making them more memory-friendly. However, the quality of gradient estimates in zeroth order optimization often depends on the data dimensionality, potentially explaining why MeZO still exhibits significant performance drops compared to standard …
abstract arxiv cost cs.ai cs.cl cs.lg fine-tuning gradient issue language language models large language large language models llm llms making memory parameters performance propagation results specific tasks tasks training type
More from arxiv.org / cs.LG updates on arXiv.org
Jobs in AI, ML, Big Data
Senior Machine Learning Engineer
@ GPTZero | Toronto, Canada
Customer Data Analyst with Spanish
@ Michelin | Voluntari
HC Data Analyst - Senior
@ Leidos | 1662 Intelligence Community Campus - Bethesda MD
Healthcare Research & Data Analyst- Infectious, Niche, Rare Disease
@ Clarivate | Remote (121- Massachusetts)
Data Analyst (maternity leave cover)
@ Clarivate | R155-Belgrade
Sales Enablement Data Analyst (Remote)
@ CrowdStrike | USA TX Remote