CUTLASS 3.1 Python interface documentation (#917)

* Add 12.1 Dockerfile

* Add 3.1 docs

python/docs/_sources/contribute.md.txt

@@ -0,0 +1,9 @@
+# Contributing
+
+Thank you for your interest in contributing to the CUTLASS Python interface. Based on the type of contribution, it will fall into two categories:
+
+1. You want to report a bug, feature request, or documentation issue
+    - File an [issue](https://github.com/NVIDIA/cutlass/issues/new/choose) describing what you encountered or what you want to see changed.
+    - The CUTLASS team will evaluate the issues and triage them, scheduling them for a release. If you believe the issue needs priority attention, comment on the issue to notify the team.
+2. You want to implement a feature or bug-fix
+    - We welcome contributions from the community. We recommend that you contribute via a [pull request](https://github.com/NVIDIA/cutlass/pulls). If you have questions about CUTLASS, consider asking a question via the [Discussions](https://github.com/NVIDIA/cutlass/discussions) tab. Please be sure to search through both existing issues and discussions to see whether your question has already been answered.

python/docs/_sources/cutlass.emit.rst.txt

@@ -0,0 +1,18 @@
+Emitters
+========
+
+Common
+------
+
+.. automodule:: cutlass.emit.common
+   :members:
+   :undoc-members:
+   :show-inheritance:
+
+PyTorch
+-------
+
+.. automodule:: cutlass.emit.pytorch
+   :members:
+   :undoc-members:
+   :show-inheritance:

python/docs/_sources/cutlass.op.rst.txt

@@ -0,0 +1,26 @@
+Operations
+==========
+
+GEMM
+----
+
+.. automodule:: cutlass.op.gemm
+   :members:
+   :undoc-members:
+   :show-inheritance:
+
+Grouped GEMM
+------------
+
+.. automodule:: cutlass.op.gemm_grouped
+   :members:
+   :undoc-members:
+   :show-inheritance:
+
+Operation
+---------
+
+.. automodule:: cutlass.op.op
+   :members:
+   :undoc-members:
+   :show-inheritance:

python/docs/_sources/cutlass.rst.txt

@@ -0,0 +1,36 @@
+CUTLASS
+=======
+
+Subpackages
+-----------
+
+.. toctree::
+   :maxdepth: 1
+
+   cutlass.emit
+   cutlass.op
+   cutlass.utils
+
+Epilogue
+--------
+
+.. automodule:: cutlass.epilogue
+   :members:
+   :undoc-members:
+   :show-inheritance:
+
+Library Defaults
+----------------
+
+.. automodule:: cutlass.library_defaults
+   :members:
+   :undoc-members:
+   :show-inheritance:
+
+Swizzle
+----------
+
+.. automodule:: cutlass.swizzle
+   :members:
+   :undoc-members:
+   :show-inheritance:

python/docs/_sources/cutlass.utils.rst.txt

@@ -0,0 +1,18 @@
+Utilities
+=========
+
+Checks
+------
+
+.. automodule:: cutlass.utils.check
+   :members:
+   :undoc-members:
+   :show-inheritance:
+
+Data Types
+----------
+
+.. automodule:: cutlass.utils.datatypes
+   :members:
+   :undoc-members:
+   :show-inheritance:

python/docs/_sources/examples.rst.txt

@@ -0,0 +1,9 @@
+Examples
+==================
+
+.. toctree::
+   :maxdepth: 5
+
+   Basic GEMM <externals/00_basic_gemm.nblink>
+   Epilogue <externals/01_epilogue.nblink>
+   PyTorch Extension <externals/02_pytorch_extension_grouped_gemm.nblink>

python/docs/_sources/externals/00_basic_gemm.nblink.txt

@@ -0,0 +1,3 @@
+{
+  "path": "./../../../../examples/python/00_basic_gemm.ipynb"
+}

python/docs/_sources/externals/01_epilogue.nblink.txt

@@ -0,0 +1,3 @@
+{
+  "path": "./../../../../examples/python/01_epilogue.ipynb"
+}

python/docs/_sources/externals/02_pytorch_extension_grouped_gemm.nblink.txt

@@ -0,0 +1,3 @@
+{
+  "path": "./../../../../examples/python/02_pytorch_extension_grouped_gemm.ipynb"
+}

python/docs/_sources/index.rst.txt

@@ -0,0 +1,55 @@
+.. CUTLASS Python interface documentation master file, created by
+   sphinx-quickstart on Mon Feb 13 17:57:39 2023.
+   You can adapt this file completely to your liking, but it should at least
+   contain the root `toctree` directive.
+
+.. include:: ../../README.md
+   :start-line: 1
+   :parser: markdown
+
+.. toctree::
+   :hidden:
+
+   Home <self>
+
+.. toctree::
+   :hidden:
+   :caption: Getting Started:
+
+   install.md
+   Getting Started <externals/00_basic_gemm.nblink>
+   contribute.md
+
+.. toctree::
+   :hidden:
+   :caption: Python Documentation:
+
+   modules.rst
+
+.. toctree::
+   :hidden:
+   :caption: Examples and Tutorials:
+
+   examples.rst
+
+.. toctree::
+   :hidden:
+   :caption: Advanced:
+
+.. toctree::
+   :hidden:
+   :caption: FAQ:
+
+.. toctree::
+   :hidden:
+   :caption: Reference:
+
+   Github <https://github.com/NVIDIA/cutlass>
+
+
+Indices and tables
+==================
+
+* :ref:`genindex`
+* :ref:`modindex`
+* :ref:`search`

python/docs/_sources/install.md.txt

@@ -0,0 +1,36 @@
+# Installation
+
+## Installing from source
+
+Installing from source requires the latest CUDA Toolkit that matches the major.minor of CUDA Python installed.
+
+Prior to installing the CUTLASS Python interface, one may optionally set the following environment variables:
+* `CUTLASS_PATH`: the path to the cloned CUTLASS repository
+* `CUDA_INSTALL_PATH`: the path to the installation of CUDA
+
+If these environment variables are not set, the installation process will infer them to be the following:
+* `CUTLASS_PATH`: one directory level above the current directory (i.e., `$(pwd)/..`)
+* `CUDA_INSTALL_PATH`: the directory holding `/bin/nvcc` for the first version of `nvcc` on `$PATH` (i.e., `which nvcc | awk -F'/bin/nvcc' '{print $1}'`)
+
+**NOTE:** The version of `cuda-python` installed must match the CUDA version in `CUDA_INSTALL_PATH`.
+
+### Installing a developer-mode package
+The CUTLASS Python interface can currently be installed via:
+```bash
+python setup.py develop --user
+```
+This will allow changes to the Python interface source to be reflected when using the Python interface.
+
+We plan to add support for installing via `python setup.py install` in a future release.
+
+## Docker
+To ensure that you have all of the necessary Python modules for running the examples using the
+CUTLASS Python interface, we recommend using one of the Docker images located in the docker directory.
+
+For example, to build and launch a container that uses CUDA 12.1 via an NGC PyTorch container, run:
+```bash
+docker build -t cutlass-cuda12.1:latest -f docker/Dockerfile-cuda12.1-pytorch .
+docker run --gpus all -it --rm cutlass-cuda12.1:latest
+```
+
+The CUTLASS Python interface has been tested with CUDA 11.8, 12.0, and 12.1 on Python 3.8.10 and 3.9.7.

python/docs/_sources/modules.rst.txt

@@ -0,0 +1,7 @@
+CUTLASS Python API
+==================
+
+.. toctree::
+   :maxdepth: 5
+
+   cutlass