Fine Tuned Weight

Fine-tuning pre-trained models involves adapting existing weights to new tasks, improving efficiency and performance compared to training from scratch. Current research focuses on optimizing fine-tuning strategies, including techniques like weight averaging across multiple fine-tuned models and identifying and selectively fine-tuning specific model components to enhance robustness and generalization. These advancements are significant for resource-constrained settings and improving the reliability and safety of large language models and other deep learning applications, particularly in areas like medical image analysis and adversarial robustness.

Papers

July 29, 2024

Segmenting Fetal Head with Efficient Fine-tuning Strategies in Low-resource Settings: an empirical study with U-Net
Fangyijie Wang, Guénolé Silvestre, Kathleen M. Curran
Fine Tuning Empirical Study U Net Low Resource Accurate Segmentation Ultrasound Data Fetal Motion Fine Tuned Weight

March 28, 2024

Model Stock: All we need is just a few fine-tuned models
Dong-Hwan Jang, Sangdoo Yun, Dongyoon Han
Full Model Pre Trained Model Large Pre Trained Model Fine Tuned Model Fine Tuned Weight

February 15, 2024

Recovering the Pre-Fine-Tuning Weights of Generative Models
Eliahu Horwitz, Jonathan Kahana, Yedid Hoshen
Generative Model Fine Tuning Pre Trained Model Generative Modeling Fine Tuned Model Fine Tuned Weight

January 22, 2024

WARM: On the Benefits of Weight Averaged Reward Models
Alexandre Ramé, Nino Vieillard, Léonard Hussenot, Robert Dadashi, Geoffrey Cideron, Olivier Bachem, Johan Ferret
Large Language Model Reinforcement Learning Complementary Benefit Fine Tuned Weight

August 1, 2023

Improving Generalization of Adversarial Training via Robust Critical Fine-Tuning
Kaijie Zhu, Jindong Wang, Xixu Hu, Xing Xie, Ge Yang
Native Robustness Strong Generalization Adversarial Training Adversarial Robustness Robust Fine Tuning Fine Tuned Weight

May 28, 2023

Efficient Storage of Fine-Tuned Models via Low-Rank Approximation of Weight Residuals
Simo Ryu, Seunghyun Seo, Jaejun Yoo
Low Rank Average Approximation Fine Tuned Model Residual Coding Fine Tuned Weight

December 20, 2022

Model Ratatouille: Recycling Diverse Models for Out-of-Distribution Generalization
Alexandre Ramé, Kartik Ahuja, Jianyu Zhang, Matthieu Cord, Léon Bottou, David Lopez-Paz
Machine Learning Model Foundation Model Distribution Generalization Auxiliary Task Diverse Model Finetuning Method Fine Tuned Weight

Fine Tuned Weight

Papers

Segmenting Fetal Head with Efficient Fine-tuning Strategies in Low-resource Settings: an empirical study with U-Net

Model Stock: All we need is just a few fine-tuned models

Recovering the Pre-Fine-Tuning Weights of Generative Models

WARM: On the Benefits of Weight Averaged Reward Models

Improving Generalization of Adversarial Training via Robust Critical Fine-Tuning

Efficient Storage of Fine-Tuned Models via Low-Rank Approximation of Weight Residuals

Model Ratatouille: Recycling Diverse Models for Out-of-Distribution Generalization