# tensorpack/tensorpack

A Neural Net Training Interface on TensorFlow, with focus on speed + flexibility

Repository: https://github.com/tensorpack/tensorpack
Canonical: https://ross.abutalabs.com/products/tensorpack
Language: Python
License: Apache-2.0
License Family: permissive
Topics: tensorflow, deep-learning, reinforcement-learning, neural-networks, machine-learning
Last push: 2023-08-06T00:30:36+00:00

## Health v2 (maintenance only)
Score: 23/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 3904, "days_push": 1124, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 6286, forks 1778 (observed 2026-08-28T04:09:41.302885+00:00)

## What it is
Tensorpack is a high-level neural network training interface built on graph-mode TensorFlow, focused on training speed and flexibility for research. It provides efficient data loading via its dataflow library and off-the-shelf multi-GPU/distributed training, with reproducible implementations of well-known papers.

## Use cases
- train resnet on imagenet fast
- train faster r-cnn on coco
- speed up tensorflow training over keras
- multi-gpu distributed training in tensorflow
- reproduce research papers in computer vision
- efficient python data loading pipeline for deep learning
- reinforcement learning training framework

## When to choose
- you need maximum training throughput with graph-mode TensorFlow
- you want reproducible, research-quality implementations of vision papers
- you need scalable multi-GPU or distributed training out of the box
- your data pipeline is the bottleneck and you need fast Python data loading

## When to avoid
- you are starting a new project with PyTorch or modern TF2/Keras eager execution
- you need a simple high-level model wrapper for quick prototyping
- you require active community development and frequent updates

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, llm-training, data-science
- domain: deep-learning, machine-learning, computer-vision, reinforcement-learning
- platform: python, cross-platform
- tags: tensorflow, training-interface, dataflow, multi-gpu, model-zoo, research, gpu, linux

## Member repositories
- tensorpack/tensorpack (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:41.302885+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 2026-08-29T17:46:19.335781+00:00, confidence not recorded.
  - readme: https://github.com/tensorpack/tensorpack (fetched 2026-08-28T04:09:41.302885+00:00, sha 9cae430bb4f1)
  - registry_pypi: https://pypi.org/pypi/tensorpack/json (fetched 2026-08-29T08:42:59.938385+00:00, sha d5dca5527a65)
- Data as of 2026-08-30T08:39:29.467469+00:00.