# tensorflow/adanet

Fast and flexible AutoML with learning guarantees.

Repository: https://github.com/tensorflow/adanet
Canonical: https://ross.abutalabs.com/products/adanet
Homepage: https://adanet.readthedocs.io
Language: Jupyter Notebook
License: Apache-2.0
License Family: permissive
Topics: automl, tensorflow, learning-theory, deep-learning, neural-architecture-search, gpu, machine-learning, ensemble, tpu, python, distributed-training
Archived: true
Last push: 2023-11-30T16:30:21+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 2988, "days_push": 1007, "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 3454, forks 522 (observed 2026-08-28T04:08:05.393111+00:00)

## What it is
AdaNet is a lightweight TensorFlow-based AutoML framework that automatically learns high-quality neural network architectures and ensembles with minimal expert intervention. It implements the AdaNet algorithm from ICML 2017, adaptively growing ensembles of subnetworks while providing theoretical learning guarantees.

## Use cases
- automatically train neural network models without manual architecture design
- learn ensembles of neural networks for better accuracy
- run neural architecture search with learning guarantees
- automate machine learning model selection on tabular or feature data
- extend AutoML with custom subnetwork architectures
- train models distributed across GPUs or TPUs

## When to choose
- you want automated model/ensemble training on TensorFlow with minimal tuning
- you need theoretically grounded AutoML with learning guarantees
- you want to research custom search spaces and subnetwork designs
- you need Keras/Estimator-compatible AutoML

## When to avoid
- you work primarily in PyTorch or non-TensorFlow ecosystems
- you need cutting-edge actively developed AutoML - the project is in maintenance mode
- you need transformer/LLM-focused AutoML
- you want a no-code AutoML service rather than a Python library

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, llm-training
- domain: machine-learning, deep-learning, artificial-intelligence
- platform: python, cross-platform
- tags: automl, neural-architecture-search, ensemble-learning, tensorflow, learning-theory, tpu, distributed-training, gpu

## Member repositories
- tensorflow/adanet (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:05.393111+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 2026-08-29T18:37:31.830978+00:00, confidence not recorded.
  - readme: https://github.com/tensorflow/adanet (fetched 2026-08-28T04:08:05.393111+00:00, sha 25be5565e960)
  - registry_pypi: https://pypi.org/pypi/adanet/json (fetched 2026-08-29T09:31:43.666849+00:00, sha e1e928b5982e)
- Data as of 2026-08-30T08:39:29.467469+00:00.