Layer-wise Importance Matters: Less Memory for Better Performance in Parameter-efficient Fine-tuning of Large Language Models

Documenti analoghi

• GradOT: Training-free Gradient-preserving Offsite-tuning for Large Language Models
  di: Yao, Kai, et al.
  Pubblicazione: (2025)
• GRASS: Gradient-based Adaptive Layer-wise Importance Sampling for Memory-efficient Large Language Model Fine-tuning
  di: Tian, Kaiyuan, et al.
  Pubblicazione: (2026)
• ScaleOT: Privacy-utility-scalable Offsite-tuning with Dynamic LayerReplace and Selective Rank Compression
  di: Yao, Kai, et al.
  Pubblicazione: (2024)
• Less Could Be Better: Parameter-efficient Fine-tuning Advances Medical Vision Foundation Models
  di: Lian, Chenyu, et al.
  Pubblicazione: (2024)
• OPa-Ma: Text Guided Mamba for 360-degree Image Out-painting
  di: Gao, Penglei, et al.
  Pubblicazione: (2024)