# tensorflow/benchmarks

A benchmark framework for Tensorflow

Repository: https://github.com/tensorflow/benchmarks
Canonical: https://ross.abutalabs.com/products/tensorflow-benchmarks
Language: Python
License: Apache-2.0
License Family: permissive
Archived: true
Last push: 2023-10-06T04:12:22+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 3612, "days_push": 1062, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, archived
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1148, forks 623 (observed 2026-08-28T04:03:46.043314+00:00)

## What it is
A benchmark framework for TensorFlow, containing PerfZero for running and reporting TensorFlow benchmarks and legacy CNN benchmark scripts for TensorFlow 1. It helps measure model training and inference performance across hardware setups.

## Use cases
- benchmark tensorflow model performance
- measure gpu training throughput for deep learning models
- run automated performance regression tests for tensorflow
- compare cnn benchmark results across hardware
- track tensorflow performance over time

## When to choose
- you need to benchmark TensorFlow model performance on GPUs or TPUs
- you want an automated framework for performance regression tracking
- you need reproducible CNN benchmark results for TensorFlow 1

## When to avoid
- you benchmark PyTorch or other non-TensorFlow frameworks
- you need maintained CNN benchmarks for TensorFlow 2 (tf_cnn_benchmarks is unmaintained)
- you want general-purpose microbenchmarking of arbitrary Python code

## Facets
- artifact type: framework
- maturity: maintenance
- function: benchmarking, machine-learning, deep-learning
- domain: machine-learning, deep-learning, performance, developer-tools
- platform: python
- tags: tensorflow, perfzero, cnn-benchmarks, performance-measurement, gpu-benchmarking, gpu, linux

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

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:46.043314+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 2026-08-30T06:33:52.337299+00:00, confidence not recorded.
  - readme: https://github.com/tensorflow/benchmarks (fetched 2026-08-28T04:03:46.043314+00:00, sha cc1aae64b642)
- Data as of 2026-08-30T08:39:29.467469+00:00.