This repository contains custom builds of tensorflow. To install one of these on your system, download the correct file according to your version of python and gcc and run the following command.

pip install --ignore-installed --upgrade /path/to/binary.whl --user

pip install --ignore-installed --upgrade "Download URL" --user

All the binaries were compiled by passing

--config=opt

-O2 -Wno-sign-compare

Latest Builds

| TF | HW | OS | GCC | Python | Supports | | |--------|-----|----------------|--------------------|----------|---------------------------------|-----------------------------------------------------------------------------------------------------------------------------------------------------------| | 2.4.1 | CPU | Ubuntu 20.04 | 9.3.0 | 3.8.5 | FMA, AVX, AVX2, SSE4.1, SSE4.2 | Download | | 2.2.1 | CPU | macOS Catalina | clang-1200.0.26.2 | 3.8.6 | FMA, AVX, AVX2, SSE4.1, SSE4.2 | Download | | 2.2.0 | CPU | Ubuntu 18.04 | 7.5.0 | 3.7.7 | FMA, AVX, AVX2, SSE4.1, SSE4.2 | Download | | 2.1.0 | CPU | Ubuntu 16.04 | 5.4.0 | 3.5.2 | FMA, AVX, AVX2, SSE4.1, SSE4.2 | Download |

What flags are supported by my CPU?

To check the instruction sets supported by your CPU, check the output of the following commands:

# macOS
sysctl -a | grep "machdep.cpu.*features:"

Linux
cat /proc/cpuinfo | grep flags

If you find this project useful, consider donating whatever amount you value this project at to a (any) charity, please drop me an email if you do. I personally care a lot about improving access to high quality medical facilities and Rotary Ambala Cancer & General Hospital is my charity of choice. It is a donation supported hospital in my hometown and provides extremely affordable services to the local population.

The donation process on its page is suboptimal, if you prefer not going through the trouble, consider using the "Sponsor" button at the top or this Venmo QR code for @lakshayg on Venmo and I'll make sure your donations get to the hospital. Thanks!

Expand for Venmo QR Code

Visit tensorflow.org for instructions on building tensorflow for your machine.