# kubeflow/katib

Automated Machine Learning on Kubernetes

Repository: https://github.com/kubeflow/katib
Canonical: https://ross.abutalabs.com/products/katib
Homepage: https://www.kubeflow.org/docs/components/katib
Language: Python
License: Apache-2.0
License Family: permissive
Topics: ai, automl, huggingface, hyperparameter-tuning, jax, kubeflow, kubernetes, llm, machine-learning, mlops, neural-architecture-search, pytorch, scikit-learn, tensorflow
Last push: 2026-08-06T23:09:31+00:00

## Health v2 (maintenance only)
Score: 71/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 96, release rhythm 22, longevity 100
- inputs: {"age_days": 3074, "days_push": 27, "days_rel": 307, "gap_med": 213, "n_releases_24m": 2}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1697, forks 536 (observed 2026-08-28T04:05:24.014372+00:00)

## What it is
Kubeflow Katib is a Kubernetes-native AutoML system for hyperparameter tuning, early stopping, and neural architecture search. It is ML-framework agnostic and can run training jobs via Kubernetes custom resources such as Kubeflow Training Operator, Argo Workflows, and Tekton Pipelines.

## Use cases
- tune hyperparameters of a PyTorch or TensorFlow model on Kubernetes
- run a hyperparameter sweep for an LLM fine-tuning job
- optimize RAG retrieval parameters like top_k and temperature
- perform neural architecture search on a cluster
- apply early stopping to cut wasted training runs
- sweep XGBoost model parameters in parallel
- plug a custom search algorithm into an AutoML pipeline

## When to choose
- you already run workloads on Kubernetes and want cluster-native AutoML
- you need framework-agnostic tuning across TensorFlow, PyTorch, XGBoost, or custom code
- you want to scale hyperparameter optimization with distributed trials and early stopping
- you are building a Kubeflow-based ML or GenAI platform

## When to avoid
- you have no Kubernetes cluster and just need local hyperparameter search
- you want a simple single-machine tuning library like Optuna
- you need a fully managed AutoML service without infrastructure setup

## Facets
- artifact type: service
- maturity: stable
- function: machine-learning, workflow-automation, scheduling, developer-tools
- domain: machine-learning, artificial-intelligence, large-language-models
- platform: cloud, self-hosted, python, go
- tags: automl, hyperparameter-tuning, neural-architecture-search, early-stopping, kubeflow, mlops, llm-fine-tuning, rag-tuning, containers, kubernetes, docker

## Member repositories
- kubeflow/katib (main) score 71

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:24.014372+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 2026-08-30T03:37:51.544803+00:00, confidence not recorded.
  - readme: https://github.com/kubeflow/katib (fetched 2026-08-28T04:05:24.014372+00:00, sha 1b31faeffb06)
  - homepage: https://www.kubeflow.org/docs/components/katib (fetched 2026-08-29T11:12:47.450355+00:00, sha 9037e726286f)
  - site_page: https://www.kubeflow.org/docs (fetched 2026-08-29T11:12:47.453458+00:00, sha 2d48c8aa19c2)
  - site_page: https://www.kubeflow.org/docs/about/community (fetched 2026-08-29T11:12:47.455529+00:00, sha 85bbf5395b12)
  - site_page: https://www.kubeflow.org/docs/genai (fetched 2026-08-29T11:12:47.457551+00:00, sha 0bf5a9037279)
  - site_page: https://www.kubeflow.org/docs/genai/use-cases (fetched 2026-08-29T11:12:47.459224+00:00, sha bfca00da3bad)
  - site_page: https://www.kubeflow.org/docs/started (fetched 2026-08-29T11:12:47.461503+00:00, sha 52d4becf41e9)
  - site_page: https://www.kubeflow.org/docs/started/introduction (fetched 2026-08-29T11:12:47.463657+00:00, sha 240c9135c386)
  - site_page: https://www.kubeflow.org/docs/started/architecture (fetched 2026-08-29T11:12:47.465894+00:00, sha eeb497b45570)
  - site_page: https://www.kubeflow.org/docs/started/installing-kubeflow (fetched 2026-08-29T11:12:47.467882+00:00, sha a977fcac93bf)
- Data as of 2026-08-30T08:39:29.467469+00:00.