Mergenetic: a Simple Evolutionary Model Merging Library

TL;DR

Mergenetic is an open-source library that facilitates evolutionary model merging, enabling flexible experimentation and achieving competitive results across tasks with minimal hardware.

Contribution

It introduces a flexible framework for evolutionary model merging, integrating various merging methods and fitness estimators to enhance performance and experimentation.

Findings

Achieves competitive results across multiple tasks and languages.

Supports flexible composition of merging strategies and evolutionary algorithms.

Operates efficiently on modest hardware.

Abstract

Model merging allows combining the capabilities of existing models into a new one - post hoc, without additional training. This has made it increasingly popular thanks to its low cost and the availability of libraries that support merging on consumer GPUs. Recent work shows that pairing merging with evolutionary algorithms can boost performance, but no framework currently supports flexible experimentation with such strategies in language models. We introduce Mergenetic, an open-source library for evolutionary model merging. Mergenetic enables easy composition of merging methods and evolutionary algorithms while incorporating lightweight fitness estimators to reduce evaluation costs. We describe its design and demonstrate that Mergenetic produces competitive results across tasks and languages using modest hardware.