#173173rd of 206 in ML frameworks and libraries
nevergrad
Python library for gradient-free optimization of continuous and discrete parameters
- Stars
- 4.2k
- License
- MIT
- Last commit
- Nov 2026
- Last release
- Apr 2025
- Language
- Python
4,168 on 22 Apr 20264,227 on 11 Nov 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
Nevergrad is a Python 3.8+ library that minimizes black-box functions without gradients. It provides optimizers such as NGOpt and differential evolution variants, plus a parametrization API for bounded, log-scaled, integer and categorical variables, including functions with keyword arguments. Typical use is hyperparameter search under a fixed evaluation budget.
Who it is for: Engineers tuning hyperparameters or black-box functions in Python
Strengths
- Parametrization mixes log-scaled, integer-cast, bounded and categorical variables in one search space
- Installs with pip; minimize() takes a parametrization and an evaluation budget
- NGOpt selects an algorithm automatically, so no optimizer tuning is needed to start
- MIT licensed, with a documentation site and published benchmark data
Weaknesses
- Last release is 2025-04-23, well behind recent commits
- README gives no benchmark numbers or comparison with other optimizer libraries
- Black-box optimizer only; not an LLM, agent or model-serving tool
- Authors describe the documentation as a work in progress
What it needs
- no GPU
What changed
From the release notes. No flag means the note does not say so, not that the upgrade is safe.
Also in ML frameworks and libraries
See all 206| Rank | Project | Score |
|---|---|---|
| 1 | transformersPython library defining open models for text, vision, audio and multimodal tasks | 94 out of 100 |
| 2 | ultralyticsPython library and CLI for training and running YOLO vision models | 94 out of 100 |
| 3 | pytorchPython tensor library with GPU support and tape-based autograd | 93 out of 100 |
| #4 | pytorch-image-modelsPyTorch library of image models, pretrained weights and training scripts | 93 out of 100 |
| #5 | supervisionPython toolkit for loading, annotating and counting computer vision detections | 92 out of 100 |