I am trying to install TensorFlow in my local directory using the following command.
export TF_BINARY_URL=http://storage.googleapis.com/tensorflow/linux/cpu/tensorflow-0.11.0-cp27-none-linux_x86_64.whl pip install --install-option="--prefix=$PYTHONUSERBASE" --upgrade $TF_BINARY_URL
I am getting the following error:
IOError: [Errno 28] No space left on device
Then I did df to see the following:
Filesystem 1K-blocks Used Available Use% Mounted on tmpfs 10240 10240 0 100% /tmp tmpfs 10240 10240 0 100% /var/tmp
Is there a way I can install TF without the temp files being downloaded in /tmp or /var/tmp? Thanks.
Usually, You can set the environment variable 'TMPDIR' to use a different directory other than /tmp or /var/tmp and most programs will honour that.
You can perhaps try,
$ export TMPDIR=$HOME/tmp
and then start your 'pip install'
You might be able to use 'pip install -b /some/other/dir' which changes the build dir.
You can also change the wheel dir as can be seen here https://pip.pypa.io/en/stable/user_guide/#installation-bundles
Running pip help install will get you the other dir options as well.
-b, --build <dir> Directory to unpack packages into and build in. -t, --target <dir> Install packages into <dir>. By default this will not replace existing files/folders in <dir>. Use --upgrade to replace existing packages in <dir> with new versions. -d, --download <dir> Download packages into <dir> instead of installing them, regardless of what is already installed. --src <dir> Directory to check out editable projects into. The default in a virtualenv is "<venv path>/src". The default for global installs is "<current dir>/src".
