Tensorflow on MacOS: Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA

Question

I tried to validate my tensorflow for my mac using these instructions https://www.tensorflow.org/install/install_mac#ValidateYourInstallation

but produce this result. is that ok? bad? how can i fix this? thanks

• sess = tf.Session()

Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA

• print(sess.run(hello))

b'Hello, TensorFlow!'

Mac OS version: MacOS High Sierra 10.13.6

here is the full installation and validation output:

usermacbook:tensorflowve someuser$ python3 -m venv . 
usermacbook:tensorflowve someuser$ source bin/activate
(tensorflowve) usermacbook:tensorflowve someuser$ pip3 install --upgrade tensorflow
Collecting tensorflow
  Downloading https://files.pythonhosted.org/packages/57/38/96ee343729363297d62311272adc28b466d30f74c87dc645badb0102af56/tensorflow-1.9.0-cp36-cp36m-macosx_10_11_x86_64.whl (48.9MB)
    100% |████████████████████████████████| 48.9MB 30kB/s 
Collecting termcolor>=1.1.0 (from tensorflow)
  Downloading https://files.pythonhosted.org/packages/8a/48/a76be51647d0eb9f10e2a4511bf3ffb8cc1e6b14e9e4fab46173aa79f981/termcolor-1.1.0.tar.gz
Collecting protobuf>=3.4.0 (from tensorflow)
  Downloading https://files.pythonhosted.org/packages/6d/7e/51c91b28cb8446ebd7231d375a2025bca4c59d15ddc0cf2dd0867b400cd7/protobuf-3.6.0-cp36-cp36m-macosx_10_6_intel.macosx_10_9_intel.macosx_10_9_x86_64.macosx_10_10_intel.macosx_10_10_x86_64.whl (1.2MB)
    100% |████████████████████████████████| 1.2MB 986kB/s 
Collecting absl-py>=0.1.6 (from tensorflow)
  Downloading https://files.pythonhosted.org/packages/96/5d/18feb90462c8edaae71305716c7e8bac479fc9dface63221f808a6b95880/absl-py-0.3.0.tar.gz (84kB)
    100% |████████████████████████████████| 92kB 3.7MB/s 
Collecting gast>=0.2.0 (from tensorflow)
  Downloading https://files.pythonhosted.org/packages/5c/78/ff794fcae2ce8aa6323e789d1f8b3b7765f601e7702726f430e814822b96/gast-0.2.0.tar.gz
Collecting grpcio>=1.8.6 (from tensorflow)
  Downloading https://files.pythonhosted.org/packages/c2/bc/31fb22c3dcccf018341a88bb9edc642a9b78bdedf9153bd3edc0d8e2f070/grpcio-1.14.0-cp36-cp36m-macosx_10_7_intel.whl (1.9MB)
    100% |████████████████████████████████| 1.9MB 654kB/s 
Collecting tensorboard<1.10.0,>=1.9.0 (from tensorflow)
  Downloading https://files.pythonhosted.org/packages/9e/1f/3da43860db614e294a034e42d4be5c8f7f0d2c75dc1c428c541116d8cdab/tensorboard-1.9.0-py3-none-any.whl (3.3MB)
    100% |████████████████████████████████| 3.3MB 425kB/s 
Collecting numpy>=1.13.3 (from tensorflow)
  Downloading https://files.pythonhosted.org/packages/6a/d5/218414f0f41cb3f183d55d68dbcd996fa3602d5849bdf2ad6c059e98fa68/numpy-1.15.0-cp36-cp36m-macosx_10_6_intel.macosx_10_9_intel.macosx_10_9_x86_64.macosx_10_10_intel.macosx_10_10_x86_64.whl (24.5MB)
    100% |████████████████████████████████| 24.5MB 59kB/s 
Collecting setuptools<=39.1.0 (from tensorflow)
  Downloading https://files.pythonhosted.org/packages/8c/10/79282747f9169f21c053c562a0baa21815a8c7879be97abd930dbcf862e8/setuptools-39.1.0-py2.py3-none-any.whl (566kB)
    100% |████████████████████████████████| 573kB 1.6MB/s 
Collecting six>=1.10.0 (from tensorflow)
  Downloading https://files.pythonhosted.org/packages/67/4b/141a581104b1f6397bfa78ac9d43d8ad29a7ca43ea90a2d863fe3056e86a/six-1.11.0-py2.py3-none-any.whl
Collecting astor>=0.6.0 (from tensorflow)
  Downloading https://files.pythonhosted.org/packages/35/6b/11530768cac581a12952a2aad00e1526b89d242d0b9f59534ef6e6a1752f/astor-0.7.1-py2.py3-none-any.whl
Collecting wheel>=0.26 (from tensorflow)
  Downloading https://files.pythonhosted.org/packages/81/30/e935244ca6165187ae8be876b6316ae201b71485538ffac1d718843025a9/wheel-0.31.1-py2.py3-none-any.whl (41kB)
    100% |████████████████████████████████| 51kB 4.0MB/s 
Collecting werkzeug>=0.11.10 (from tensorboard<1.10.0,>=1.9.0->tensorflow)
  Downloading https://files.pythonhosted.org/packages/20/c4/12e3e56473e52375aa29c4764e70d1b8f3efa6682bef8d0aae04fe335243/Werkzeug-0.14.1-py2.py3-none-any.whl (322kB)
    100% |████████████████████████████████| 327kB 2.0MB/s 
Collecting markdown>=2.6.8 (from tensorboard<1.10.0,>=1.9.0->tensorflow)
  Downloading https://files.pythonhosted.org/packages/6d/7d/488b90f470b96531a3f5788cf12a93332f543dbab13c423a5e7ce96a0493/Markdown-2.6.11-py2.py3-none-any.whl (78kB)
    100% |████████████████████████████████| 81kB 3.7MB/s 
Installing collected packages: termcolor, setuptools, six, protobuf, absl-py, gast, grpcio, numpy, werkzeug, markdown, wheel, tensorboard, astor, tensorflow
  Running setup.py install for termcolor ... done
  Found existing installation: setuptools 28.8.0
    Uninstalling setuptools-28.8.0:
      Successfully uninstalled setuptools-28.8.0
  Running setup.py install for absl-py ... done
  Running setup.py install for gast ... done
Successfully installed absl-py-0.3.0 astor-0.7.1 gast-0.2.0 grpcio-1.14.0 markdown-2.6.11 numpy-1.15.0 protobuf-3.6.0 setuptools-39.1.0 six-1.11.0 tensorboard-1.9.0 tensorflow-1.9.0 termcolor-1.1.0 werkzeug-0.14.1 wheel-0.31.1
You are using pip version 9.0.1, however version 18.0 is available.
You should consider upgrading via the 'pip install --upgrade pip' command.
(tensorflowve) usermacbook:tensorflowve someuser$ python
Python 3.6.3 (v3.6.3:2c5fed86e0, Oct  3 2017, 00:32:08) 
[GCC 4.2.1 (Apple Inc. build 5666) (dot 3)] on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> import tensorflow as tf
>>> hello = tf.constant('Hello, TensorFlow!')
>>> sess = tf.Session()
2018-08-04 09:28:35.905544: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA
>>> print(sess.run(hello))
b'Hello, TensorFlow!'

Answer 1

import os os.environ['TF_CPP_MIN_LOG_LEVEL']='2'
import tensorflow as tf

Answer 2

Tensorflow compiled with the advanced instructions can run 40% faster than the non-optimised code, so please don't ignore this message!

I would recommend building TF from source, here are the steps to download, compile and install v2.0 beta, including the link to the compiled .whl file.

First, uninstall the default version:

sudo pip3 uninstall protobuf
sudo pip3 uninstall tensorflow

In a temp folder, clone Tensorflow:

git clone https://github.com/tensorflow/tensorflow 
git checkout r2.0

Install the TensorFlow pip package dependencies:

pip3 install -U --user pip six numpy wheel setuptools mock future>=0.17.1
pip3 install -U --user keras_applications==1.0.6 --no-deps
pip3 install -U --user keras_preprocessing==1.0.5 --no-deps

Install Bazel, the build tool used to compile TensorFlow. After downloading bazel-0.26.0-installer-darwin-x86_64.sh:

chmod +x bazel-0.26.0-installer-darwin-x86_64.sh ./bazel-0.26.0-installer-darwin-x86_64.sh --user export PATH="$PATH:$HOME/bin" bazel version

Configure your system build by running the following at the root of your TensorFlow source tree:

./configure

Use bazel to make the TensorFlow package builder with CPU-only support:

bazel build -c opt --copt=-mavx --copt=-mavx2 --copt=-mfma --copt=-msse4.2 //tensorflow/tools/pip_package:build_pip_package

The bazel build command creates an executable named build_pip_package—this is the program that builds the pip package. Run the executable as shown below to build a .whl package in the /tmp/tensorflow_pkg directory.

To build from a release branch:

./bazel-bin/tensorflow/tools/pip_package/build_pip_package /tmp/tensorflow_pkg

The output wheel file is in: /tmp/tensorflow_pkg. You may download the file from here, and try to install it directly

pip3 install /tmp/tensorflow_pkg/tensorflow-2.0.0b1-cp37-cp37m-macosx_10_14_x86_64.whl 

cd out of that directory, and now running this should not produce any warning:

python3 -c "import tensorflow as tf;print(tf.reduce_sum(tf.random.normal([1000, 1000])))"

The original post was published here.

Answer 3

This is not an error. Tensorflow is running just fine.

It's a warning saying that Tensorflow is NOT taking advantage of some of the advanced instructions available on your CPU because its binary was compiled not to use these instructions (so it could run on older CPUs).

Your options are:

1. Ignore it.
2. Silence it (see the other answer).
3. Build the Tensorflow from source.

The #3 may make sense if you expect to perform a lot of computations on CPU: it could somewhat improve the performance. However, it would take time and effort, so I would not recommend it for a beginner: just use Tensorflow as-is.

P.S. b'Hello, TensorFlow!' is fine too. See: The print of string constant is always attached with 'b' inTensorFlow

Answer 4

Currently, I have encountered the same problem.

2018-10-01 10:43:40.926651: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA

Set environment variables before running.

Windows:

$ set TF_CPP_MIN_LOG_LEVEL=2

Linux/MacOS:

$ export TF_CPP_MIN_LOG_LEVEL=2

Through the above settings, I have solved the remaining error.