Hands-On Deep Learning Architectures with Python

上QQ阅读APP看书，第一时间看更新

Installing TensorFlow CPU

We mentioned earlier that TensorFlow is not currently stable with Python 3.7. Therefore, we will use Conda to install TensorFlow rather than using pip. Conda deals with it in a unique way. As soon as you issue the following command to install TensorFlow, Conda also downloads and installs the other required packages. It also restores Python to the version with which TensorFlow is known to be stable, say Python 3.6.8! Type the following commands into the Terminal window (except the mentioned comments) to install TensorFlow CPU:

# go to the deep_learning directory
   cd deep_learning

# activate the environment
   conda activate test_env

# install TensorFlow CPU
   conda install tensorflow

After the installation, you can run Python to check whether Conda has restored the version or not.

The commands mentioned are as per the time of publishing this book. You must know that the libraries and repositories keep on changing. In case the commands don't work properly, it is advised to check for the latest updates from the source.