update readme to reference latest changes (#756)

Author: benfred

README.md

@@ -90,11 +90,12 @@ There are also several other articles about using Implicit to build recommendati
 
 #### Requirements
 
-This library requires SciPy version 0.16 or later and Python version 3.6 or later.
+This library requires SciPy version 0.16 or later and Python version 3.9 or later.
 
-GPU Support requires at least version 11 of the [NVidia CUDA Toolkit](https://developer.nvidia.com/cuda-downloads).
+GPU Support requires version 13 of the [NVidia CUDA
+Toolkit](https://developer.nvidia.com/cuda-downloads), and requires that RMM be installed `pip install rmm-cu13`
 
-This library is tested with Python 3.7, 3.8, 3.9, 3.10 and 3.11 on Ubuntu, OSX and Windows.
+This library is tested with Python 3.9 to 3.14 on Ubuntu, OSX and Windows.
 
 #### Benchmarks
 

docs/source/installation.rst

@@ -8,16 +8,16 @@ Implicit can be installed from pypi with:
     pip install implicit
 
 Installing with pip will use prebuilt binary wheels on x86_64 Linux, Windows
-and OSX. The wheels includes GPU support on Linux.
+and OSX. The wheels includes GPU support on Linux, for CUDA 13.
 
 There are also binary packages on conda-forge for Linux, Windows and OSX. These can be installed with:
 
 .. code-block:: bash
 
     conda install -c conda-forge implicit
 
-There are also GPU enabled packages on conda-forge for x86_64 Linux systems using either CUDA
-11.0, 11.1 or 11.2. The GPU packages can be installed with:
+There are also GPU enabled packages on conda-forge for x86_64 Linux systems.
+The GPU packages can be installed with:
 
 .. code-block:: bash
 
@@ -27,4 +27,4 @@ There are also GPU enabled packages on conda-forge for x86_64 Linux systems usin
 Requirements
 ------------
 
-This library requires SciPy version 0.16 or later.
+This library requires SciPy version 0.16 or later. GPU support requires RMM to be installed.