Language-Guided Tuning: Configuration Optimization for Automated ML Research
Language-Guided Tuning is introduced, a framework that employs multi-agent Large Language Models to automatically optimize configurations through natural language reasoning to demonstrate substantial improvements over traditional optimization methods while maintaining high interpretability.
Yuxing Lu, Yucheng Hu, Nan Sun et al.
· 0 citations