# abhishekkrthakur/tez

Tez is a super-simple and lightweight Trainer for PyTorch. It also comes with many utils that you can use to tackle over 90% of deep learning projects in PyTorch.

Repository: https://github.com/abhishekkrthakur/tez
Canonical: https://ross.abutalabs.com/products/tez
Language: Python
License: Apache-2.0
License Family: permissive
Topics: tez, pytorch, deep-learning, deep-neural-networks, neural-networks
Last push: 2023-01-29T16:52:18+00:00

## Health v2 (maintenance only)
Score: 23/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 2119, "days_push": 1312, "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 1155, forks 139 (observed 2026-08-28T04:03:47.772934+00:00)

## What it is
Tez is a lightweight, simple trainer library for PyTorch that simplifies training loops while keeping users close to native PyTorch. It supports CPU, single GPU, multi-GPU, and TPU training and includes utilities covering most deep learning project needs.

## Use cases
- simplify pytorch training loops
- train neural networks on multiple gpus
- quickly prototype deep learning models
- train pytorch models on tpu
- reduce boilerplate in pytorch projects

## When to choose
- you want a minimal PyTorch trainer without heavy abstraction
- you need fast prototyping with customizable training
- you want CPU, GPU, multi-GPU, and TPU support in one lightweight library

## When to avoid
- you need a feature-rich framework like PyTorch Lightning with an active ecosystem
- you require long-term stability, as the project is early-stage with possible breaking changes
- you want to contribute via pull requests, which are currently not accepted

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, llm-training
- domain: deep-learning, machine-learning, developer-tools
- platform: python, cross-platform
- tags: pytorch, training-loop, trainer, neural-networks, tpu-support, lightweight-wrapper, gpu

## Member repositories
- abhishekkrthakur/tez (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:47.772934+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 2026-08-30T06:32:11.658610+00:00, confidence not recorded.
  - readme: https://github.com/abhishekkrthakur/tez (fetched 2026-08-28T04:03:47.772934+00:00, sha 1a870113e800)
  - registry_pypi: https://pypi.org/pypi/tez/json (fetched 2026-08-29T12:37:25.212382+00:00, sha ab19996cdbbe)
- Data as of 2026-08-30T08:39:29.467469+00:00.