# tensorflow/tpu

Reference models and tools for Cloud TPUs.

Repository: https://github.com/tensorflow/tpu
Canonical: https://ross.abutalabs.com/products/tpu
Homepage: https://cloud.google.com/tpu/
Language: Jupyter Notebook
License: Apache-2.0
License Family: permissive
Last push: 2026-06-22T19:40:55+00:00

## Health v2 (maintenance only)
Score: 72/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 88, release rhythm 35, longevity 100
- inputs: {"age_days": 3340, "days_push": 72, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 5278, forks 1752 (observed 2026-08-28T04:09:14.141935+00:00)

## What it is
A collection of reference models and tools for training machine learning models on Google Cloud TPUs, maintained as a public mirror by the TensorFlow team. It includes tutorials and utilities such as ctpu for getting started with Cloud TPU workflows.

## Use cases
- train models on cloud tpus
- learn how to use google cloud tpu
- find reference tensorflow models for tpu
- set up a tpu training environment
- benchmark models on tpu hardware

## When to choose
- you are training TensorFlow models on Google Cloud TPUs
- you want official reference implementations tuned for TPU hardware
- you need a guided tutorial to get started with Cloud TPU

## When to avoid
- you are training on GPUs or CPUs rather than TPUs
- you use PyTorch or JAX instead of TensorFlow
- you want to contribute pull requests, since the repo is a read-only mirror

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, llm-training, gpu-computing, benchmarking
- domain: deep-learning, machine-learning, cloud-computing, developer-tools
- platform: python, cloud
- tags: tpu, tensorflow, google-cloud, reference-models, jupyter-notebook, linux

## Member repositories
- tensorflow/tpu (main) score 72

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:14.141935+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 2026-08-29T17:59:17.991342+00:00, confidence not recorded.
  - readme: https://github.com/tensorflow/tpu (fetched 2026-08-28T04:09:14.141935+00:00, sha 20b34ed50cbf)
  - homepage: https://cloud.google.com/tpu/ (fetched 2026-08-29T08:54:08.183112+00:00, sha f6b88db53538)
- Data as of 2026-08-30T08:39:29.467469+00:00.