jnormore

jnormore / cevolve

Public

Genetic algorithms for autonomous code optimization. The LLM imagines ideas, evolution discovers which combinations work best together.

16
0
100% credibility
Found Apr 15, 2026 at 16 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

cevolve automatically evolves optimal code configurations by breeding and testing AI-generated optimization ideas against user-defined benchmarks.

How It Works

1
🔍 Discover cevolve

You hear about a smart helper that automatically finds ways to make your program run faster and better.

2
⚙️ Share your code

You point it to your program files and tell it how to measure success, like speed or accuracy.

3
🧬 Magic evolution starts

It dreams up clever improvement ideas, mixes them in smart ways, and tests endless combinations to find winners.

4
📊 Watch it improve

Beautiful charts show your program's performance getting steadily better with each round of tests.

🏆 Optimized perfection

You get the winning recipe of tweaks that supercharge your code, ready to use forever.

Sign up to see the full architecture

3 more

Sign Up Free

Star Growth

See how this repo grew from 16 to 16 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is cevolve?

cevolve automates code optimization using genetic algorithms in Python, where an LLM proposes ideas like parameter tweaks or architectural changes, and evolution tests combinations via your benchmark script. Point it at files like train.py, specify a metric like val_bpb or time_ms, and it runs full sessions with `cevolve run`—applying edits, evaluating fitness, and generating convergence charts plus summaries. It's genetic algorithm optimization github style, but for real-world code like ML models or schedulers, solving manual hyperparameter hell.

Why is it gaining traction?

Unlike static genetic algorithms matlab toolboxes or basic genetic algorithm python libraries, cevolve pairs evolution with LLM-driven idea generation and code edits, handling rethink cycles to evolve the search space itself. The composable CLI (`init`, `next`, `eval`, `rethink`) lets agents or scripts plug in, while built-in TUI and synergy matrices reveal what works—perfect for genetic algorithms in search optimization and machine learning. Low stars (16) but hooks devs tired of grid search in genetic algorithm projects github.

Who should use this?

ML engineers tuning training loops or inference speed, algo traders optimizing backtests, and backend devs refining parsers or schedulers via genetic algorithm scheduling github. Ideal for anyone with a quick benchmark outputting METRIC lines, especially in genetic algorithms and engineering optimization workflows.

Verdict

Promising for genetic programming python github experiments, but at 1.0% credibility and 16 stars, it's early—docs are README-only, no tests visible. Prototype it for your next genetic algorithms pdf-inspired project; skip for production until more evals.

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.