# SakanaAI/evolutionary-model-merge

Official repository of Evolutionary Optimization of Model Merging Recipes

Repository: https://github.com/SakanaAI/evolutionary-model-merge
Canonical: https://ross.abutalabs.com/products/evolutionary-model-merge
Language: Python
License: Apache-2.0
License Family: permissive
Last push: 2024-11-29T04:39:39+00:00

## Health v2 (maintenance only)
Score: 16/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 8, longevity 65
- inputs: {"age_days": 911, "days_push": 642, "days_rel": 642, "gap_med": null, "n_releases_24m": 1}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1436, forks 125 (observed 2026-08-28T04:04:43.579894+00:00)

## What it is
Official repository for SakanaAI's Evolutionary Model Merge research, providing code and resources to reproduce paper evaluations of models created via evolutionary optimization of model merging recipes. It hosts merged models like EvoLLM-JP and EvoVLM-JP that combine multiple source LLMs and VLMs.

## Use cases
- merge multiple language models into one
- reproduce evolutionary model merging experiments
- create a VLM from existing models
- optimize model merge weights automatically
- evaluate merged Japanese LLMs
- apply neuroevolution to model fusion

## When to choose
- you want to experiment with evolutionary model merging techniques
- you need to reproduce the paper's merged models or evaluations
- you're researching automated model fusion methods

## When to avoid
- you need a production-ready model serving tool
- you want a general-purpose model merging library with polished APIs
- you lack GPU resources for large model experiments

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, llm-training, deep-learning
- domain: large-language-models, machine-learning, artificial-intelligence
- platform: python
- tags: model-merging, evolutionary-algorithms, neuroevolution, llm, vlm, research-code, huggingface, gpu, linux

## Member repositories
- SakanaAI/evolutionary-model-merge (main) score 16

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:43.579894+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:36:46.116453+00:00, confidence not recorded.
  - readme: https://github.com/SakanaAI/evolutionary-model-merge (fetched 2026-08-28T04:04:43.579894+00:00, sha 7af89fcc0417)
- Data as of 2026-08-30T08:39:29.467469+00:00.