TensorFlow SIG Build is a community group dedicated to the TensorFlow build process. This repository is a showcase of resources, guides, tools, and builds contributed by the community, for the community.
SIG Build is a community-led open source project. As such, the project depends on public contributions, bug-fixes, and documentation. Please see contribution guidelines for a guide on how to contribute. This project adheres to TensorFlow's code of conduct. By participating, you are expected to uphold this code.
- Public Mailing List
- SIG Monthly Meeting Notes
- Join our mailing list and receive calendar invites to the meeting.
Want to add your own project to this list? It's easy: check out CONTRIBUTING.md.
- TF SIG Build Dockerfiles: Standard Dockerfiles for TensorFlow builds, used internally at Google
- TensorFlow Runtime Dockerfiles: Simple Dockerfiles for running TensorFlow, with Jupyter variants.
- Manylinux 2014 Docker Images:
manylinux2014build environment for TensorFlow packages. - Distroless Dockerfiles: Distroless (info) TensorFlow images, which are smaller than TensorFlow's official images.
- DevInfra Windows RBE: Static snapshot of TF DevInfra's Windows Remote Build Execution images
- Golang Install Guide: Documentation for installing the Go bindings.
- ppc64le Builds: Dockerfiles and wheel build scripts for building TF on ppc64le.
- Raspberry Pi Builds: TensorFlow's old official docs for building on Raspberry Pi. Needs an owner.
- WSL2 GPU Guide: Instructions for enabling GPU with Tensorflow on a WSL2 virtual machine.
- Directory Template: Example short description.
- TF OSS Dashboard: Dashboard for all continuous statuses on TF GitHub Commits.
- Tekton CI: perfinion's experimental directory for using Tekton CI with TensorFlow
Amazing members of the TensorFlow community build, test, and package TensorFlow on more platforms than are supported by the official TensorFlow team. Please note that as community builds they are not supported by the TensorFlow team.
Want to add your own community builds to this list? It's easy: check out CONTRIBUTING.md.
| Owner | Build Type | Status | Artifacts |
|---|---|---|---|
| AMD | Linux AMD ROCm GPU Nightly | Nightly | |
| AMD | Linux AMD ROCm GPU Stable : TF 2.x | Release 2.12 | |
| AMD | Linux AMD ROCm GPU Stable : TF 1.x | Release 1.15 | |
| AMD | Linux AMD ZenDNN Plug-in CPU Stable : TF 2.x | No Badge | Release 2.x |
| IBM | Linux ppc64le CPU Nightly | Nightly | |
| IBM | Linux ppc64le CPU Stable: TF 1.x | Release 1.15 | |
| IBM | Linux ppc64le CPU Stable: TF 2.x | Release 2.x | |
| IBM | Linux ppc64le GPU Nightly | Nightly | |
| IBM | Linux ppc64le GPU Stable: TF 1.x | Release 1.15 | |
| IBM | Linux ppc64le GPU Stable: TF 2.x | Release 2.x | |
| IBM | Linux s390x Nightly | Nightly | |
| IBM | Linux s390x CPU Stable Release | Release | |
| Intel | Linux CPU with Intel oneDNN Stable Release 1.x | No Badge | Release 1.15 |
| Intel | Linux CPU with Intel oneDNN Stable Release 2.x | No Badge | Release 2.x |
| Intel | Windows CPU with Intel oneDNN Stable Release 2.x | No Badge | Release 2.x |
| Owner | Container Type | Status | Artifacts |
|---|---|---|---|
| Arm | TensorFlow AArch64 Neoverse-N1 CPU Stable | Static | Docker Hub |
| AMD | Linux ROCm GPU Stable | Static | Docker Hub |
| Intel | Linux CPU with Intel oneDNN Stable | Static | Docker Hub |