# microsoft/nni

An open source AutoML toolkit for automate machine learning lifecycle, including feature engineering, neural architecture search, model compression and hyper-parameter tuning.

Repository: https://github.com/microsoft/nni
Canonical: https://ross.abutalabs.com/products/nni
Homepage: https://nni.readthedocs.io
Language: Python
License: MIT
License Family: permissive
Topics: automl, deep-learning, neural-architecture-search, hyperparameter-optimization, distributed, bayesian-optimization, automated-machine-learning, machine-learning, machine-learning-algorithms, data-science, tensorflow, pytorch, neural-network, deep-neural-network, model-compression, feature-engineering, nas, python, hyperparameter-tuning, mlops
Archived: true
Last push: 2024-07-03T10:55:10+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 3015, "days_push": 791, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: archived
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 14361, forks 1850 (observed 2026-08-28T04:11:06.863757+00:00)

## What it is
NNI (Neural Network Intelligence) is an open-source AutoML toolkit from Microsoft that automates hyperparameter tuning, neural architecture search, model compression (pruning and quantization), and feature engineering. It supports PyTorch and TensorFlow and can run experiments locally or distributed across clusters.

## Use cases
- automate hyperparameter tuning for deep learning models
- run neural architecture search experiments
- prune and quantize models for faster inference
- compress a PyTorch model while preserving accuracy
- run distributed AutoML experiments across a cluster
- automate feature engineering for tabular data
- compare tuning algorithms like Bayesian optimization

## When to choose
- you want a mature, feature-rich AutoML toolkit covering HPO, NAS, and model compression in one package
- you work in PyTorch or TensorFlow and want plug-and-play tuning with a web dashboard
- you need distributed hyperparameter search across multiple machines

## When to avoid
- you need a lightweight single-purpose HPO tool like Optuna with minimal setup
- you require cutting-edge features or frequent updates, as development activity has slowed
- you work outside Python or with frameworks other than PyTorch/TensorFlow

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, benchmarking, data-science
- domain: machine-learning, deep-learning, data-science, artificial-intelligence
- platform: python, windows
- tags: automl, hyperparameter-optimization, neural-architecture-search, model-compression, pruning, quantization, feature-engineering, pytorch, tensorflow, distributed-training, linux, macos, gpu

## Member repositories
- microsoft/nni (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:11:06.863757+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 2026-08-29T17:12:37.956662+00:00, confidence not recorded.
  - readme: https://github.com/microsoft/nni (fetched 2026-08-28T04:11:06.863757+00:00, sha b07835dc2ed4)
  - homepage: https://nni.readthedocs.io (fetched 2026-08-29T08:06:09.720205+00:00, sha fc1cd38f37d3)
  - registry_pypi: https://pypi.org/pypi/nni/json (fetched 2026-08-29T08:06:09.730064+00:00, sha bfae64f96dc8)
- Data as of 2026-08-30T08:39:29.467469+00:00.