How to Weight Multitask Finetuning? Fast Previews via Bayesian Model-Merging
Year: 2024
Type: preprint
Abstract: When finetuning multiple tasks altogether, it is important to carefully weigh them to get a good performance, but searching for good weights can be difficult and costly. Here, we propose to aid the search with fast previews to quickly get a rough idea of different reweighting options. We use model merging to create previews by simply reusing and averaging parameters of models trained on each task separately (no retraining required). To improve the quality of previews, we propose a Bayesian approach to design new merging strategies by using more flexible posteriors. We validate our findings on vision and natural-language transformers. Our work shows the benefits of model merging via Bayes to improve multitask finetuning.
Authors: Monzón Maldonado, Hugo, Möllenhoff, Thomas, Daheim, Nico, Gurevych, Iryna, Khan, Mohammad Emtiyaz
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W6910086468
Cites:
Cited by:
Related to: