LiNeS: Post-training Layer Scaling Prevents Forgetting and Enhances Model Merging

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Wang, Ke, Dimitriadis, Nikolaos, Favero, Alessandro, Ortiz-Jimenez, Guillermo, Fleuret, Francois, Frossard, Pascal
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866913712145694720
author Wang, Ke
Dimitriadis, Nikolaos
Favero, Alessandro
Ortiz-Jimenez, Guillermo
Fleuret, Francois
Frossard, Pascal
author_facet Wang, Ke
Dimitriadis, Nikolaos
Favero, Alessandro
Ortiz-Jimenez, Guillermo
Fleuret, Francois
Frossard, Pascal
contents Fine-tuning pre-trained models has become the standard approach to endow them with specialized knowledge, but it poses fundamental challenges. In particular, \textit{(i)} fine-tuning often leads to catastrophic forgetting, where improvements on a target domain degrade generalization on other tasks, and \textit{(ii)} merging fine-tuned checkpoints from disparate tasks can lead to significant performance loss. To address these challenges, we introduce LiNeS, Layer-increasing Network Scaling, a post-training editing technique designed to preserve pre-trained generalization while enhancing fine-tuned task performance. LiNeS scales parameter updates linearly based on their layer depth within the network, maintaining shallow layers close to their pre-trained values to preserve general features while allowing deeper layers to retain task-specific representations. In multi-task model merging scenarios, layer-wise scaling of merged parameters reduces negative task interference. LiNeS demonstrates significant improvements in both single-task and multi-task settings across various benchmarks in vision and natural language processing. It mitigates forgetting, enhances out-of-distribution generalization, integrates seamlessly with existing multi-task model merging baselines improving their performance across benchmarks and model sizes, and can boost generalization when merging LLM policies aligned with different rewards via RLHF. Our method is simple to implement, computationally efficient and complementary to many existing techniques. Our source code is available at https://github.com/wang-kee/LiNeS
format Preprint
id arxiv_https___arxiv_org_abs_2410_17146
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle LiNeS: Post-training Layer Scaling Prevents Forgetting and Enhances Model Merging
Wang, Ke
Dimitriadis, Nikolaos
Favero, Alessandro
Ortiz-Jimenez, Guillermo
Fleuret, Francois
Frossard, Pascal
Machine Learning
Computer Vision and Pattern Recognition
Fine-tuning pre-trained models has become the standard approach to endow them with specialized knowledge, but it poses fundamental challenges. In particular, \textit{(i)} fine-tuning often leads to catastrophic forgetting, where improvements on a target domain degrade generalization on other tasks, and \textit{(ii)} merging fine-tuned checkpoints from disparate tasks can lead to significant performance loss. To address these challenges, we introduce LiNeS, Layer-increasing Network Scaling, a post-training editing technique designed to preserve pre-trained generalization while enhancing fine-tuned task performance. LiNeS scales parameter updates linearly based on their layer depth within the network, maintaining shallow layers close to their pre-trained values to preserve general features while allowing deeper layers to retain task-specific representations. In multi-task model merging scenarios, layer-wise scaling of merged parameters reduces negative task interference. LiNeS demonstrates significant improvements in both single-task and multi-task settings across various benchmarks in vision and natural language processing. It mitigates forgetting, enhances out-of-distribution generalization, integrates seamlessly with existing multi-task model merging baselines improving their performance across benchmarks and model sizes, and can boost generalization when merging LLM policies aligned with different rewards via RLHF. Our method is simple to implement, computationally efficient and complementary to many existing techniques. Our source code is available at https://github.com/wang-kee/LiNeS
title LiNeS: Post-training Layer Scaling Prevents Forgetting and Enhances Model Merging
topic Machine Learning
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2410.17146