Updated TF installation instructions

- Updated TF installation instructions to reflect
  inclusion of GPU support in base `tensorflow`
  python package.
- Fixed `nvidia-driver` package name (and updated
  version number).
- Added changelog entry.

CHANGES.rst

@@ -34,6 +34,7 @@ Release history
   2.2 (due to a TensorFlow issue, see
   https://github.com/tensorflow/tensorflow/issues/39456). Loss/metric values
   will still be returned from the function as normal. (`#153`_)
+- Updated TensorFlow installation instruction in HTML documentation. (`#170`_)
 
 **Fixed**
 
@@ -53,6 +54,7 @@ Release history
 .. _#159: https://github.com/nengo/nengo-dl/pull/159
 .. _#161: https://github.com/nengo/nengo-dl/pull/161
 .. _#164: https://github.com/nengo/nengo-dl/pull/164
+.. _#170: https://github.com/nengo/nengo-dl/pull/170
 
 3.2.0 (April 2, 2020)
 ---------------------

docs/installation.rst

@@ -33,13 +33,13 @@ perform a developer installation:
 
 Installing TensorFlow
 ---------------------
-Use ``pip install tensorflow`` to install the minimal version of TensorFlow,
-or ``pip install tensorflow-gpu`` to include GPU support.
+Use ``pip install tensorflow`` to install the latest version of TensorFlow. GPU support
+is included in this package as of version 2.1.0.
 
 In order to use TensorFlow with GPU support you will need to install the appropriate
 Nvidia drivers and CUDA/cuDNN. The precise steps for accomplishing this will depend
 on your system. On Linux the correct Nvidia drivers (as of TensorFlow 2.2.0) can be
-installed via ``sudo apt install nvidia-drivers-430``, and on Windows simply using the
+installed via ``sudo apt install nvidia-driver-440``, and on Windows simply using the
 most up-to-date drivers should work.  For CUDA/cuDNN we recommend using
 `conda <https://docs.conda.io/projects/conda/en/latest/user-guide/install/>`_ to
 simplify the process. ``conda install tensorflow-gpu`` will install TensorFlow as