From ad7f27b53f1328dd2e662fb20edd6313ecebe916 Mon Sep 17 00:00:00 2001 From: Hyunsu Cho Date: Sun, 13 Sep 2026 21:23:22 -0700 Subject: [PATCH 1/2] Remove RAPIDS branding from docs --- README.md | 30 ++++++++++++++------- cpp/include/nvforest/README.md | 6 +---- docs/source/build.rst | 48 ++++++++++----------------------- docs/source/getting_started.rst | 5 ++-- 4 files changed, 39 insertions(+), 50 deletions(-) diff --git a/README.md b/README.md index 56814a1..ec08694 100644 --- a/README.md +++ b/README.md @@ -1,6 +1,6 @@ # nvForest - Highly Optimized Decision Tree Inference -nvForest is a highly-optimized and lightweight [RAPIDS](https://rapids.ai/) library that enables fast inference for decision tree models on NVIDIA GPUs and CPUs. It does not train models; it runs inference on models trained elsewhere (e.g., XGBoost, LightGBM, scikit-learn, or cuML). +nvForest is a highly-optimized and lightweight library that enables fast inference for decision tree models on NVIDIA GPUs and CPUs. It does not train models; it runs inference on models trained elsewhere (e.g., XGBoost, LightGBM, scikit-learn, or cuML). nvForest uses [Treelite](https://treelite.readthedocs.io/) as the common format for importing tree models. You can load a model from a file or from an in-memory scikit-learn or Treelite object, then run predictions with a scikit-learn-like API. Setting `device="auto"` lets you deploy the same script on machines with or without GPUs. @@ -59,20 +59,32 @@ You can tune performance with `layout` (e.g., `depth_first`, `breadth_first`) an ## Installation -See [the RAPIDS Release Selector](https://docs.rapids.ai/install#selector) for the command line to install either nightly or official release nvForest packages via conda, pip, or Docker. +You can install nvForest using Pip or Conda. -## Build/Install from Source +```console +# Using Pip: need a suffix corresponding to your CUDA version, e.g. for CUDA 13: +$ pip install nvforest-cu13 +``` -See the build [guide](BUILD.md). +```console +# Using Conda: need to specify the rapidsai channel +$ conda install -c rapidsai -c conda-forge nvforest +``` -## Contributing +### System Requirements -We welcome contributions. For guidelines and how to get started, see the [RAPIDS contributing guide](https://docs.rapids.ai/contributing). +Please see the [Installation Guide](https://docs.rapids.ai/install/#system-requirements) +for NVIDIA CUDA-X libraries for data science for information about supported operating systems, +GPU drivers, and CUDA versions. + +## Build/Install from Source + +See the build [guide](BUILD.md). ## Contact -Find out more: [CUDA-X for Data Science](https://developer.nvidia.com/topics/ai/data-science/cuda-x-for-data-science) +Find out more: [NVIDIA CUDA-X for Data Science](https://developer.nvidia.com/topics/ai/data-science/cuda-x-for-data-science) -## Open GPU Data Science +## NVIDIA CUDA-X Libraries for Data Science -The RAPIDS suite of open source software libraries aims to enable execution of end-to-end data science and analytics pipelines entirely on GPUs. It relies on NVIDIA® CUDA® primitives for low-level compute optimization, exposing GPU parallelism and high-bandwidth memory through user-friendly Python interfaces. +The NVIDIA CUDA-X libraries for data science aim to enable execution of end-to-end data science and analytics pipelines entirely on GPUs. It relies on NVIDIA® CUDA® primitives for low-level compute optimization, exposing GPU parallelism and high-bandwidth memory through user-friendly Python interfaces. diff --git a/cpp/include/nvforest/README.md b/cpp/include/nvforest/README.md index 711ab33..de42e68 100644 --- a/cpp/include/nvforest/README.md +++ b/cpp/include/nvforest/README.md @@ -1,9 +1,5 @@ # nvForest Inference Library -RAPIDS nvForest Inference Library provides accelerated inference for -tree-based machine learning models. Unlike packages like XGBoost, -LightGBM, or even Scikit-Learn/cuML's random forest implementations, nvForest -cannot be used to _train_ forest models. Instead, its goal is to speed up -inference using forest models trained by all of those packages. +nvForest is a highly-optimized and lightweight library that enables fast inference for decision tree models on NVIDIA GPUs and CPUs. It does not train models; it runs inference on models trained elsewhere (e.g., XGBoost, LightGBM, scikit-learn, or cuML). This directory contains an implementation of nvForest which provides both CPU and GPU execution. Its GPU implementation also offers diff --git a/docs/source/build.rst b/docs/source/build.rst index fc4f1e8..e203161 100644 --- a/docs/source/build.rst +++ b/docs/source/build.rst @@ -5,46 +5,26 @@ Building from the source Setting up your build environment ================================= -To install nvForest from source, ensure the following dependencies are met: - -**Hardware needed to run nvForest.** -nvForest is part of RAPIDS and follows the RAPIDS support matrix. -See https://docs.rapids.ai/platform-support/. -It is possible to build and run nvForest on machines without a GPU; in such machines, nvForest will use the CPU to run inference. - -**Software dependencies.** -See https://docs.rapids.ai/platform-support/ for the list of required C++ compilers and Python interpreters. -In addition, nvForest requires Cython 3.0 or later. +Please see the `Installation Guide `_ +for NVIDIA CUDA-X libraries for data science for information about supported operating systems, +GPU drivers, and CUDA versions. .. note:: Building nvForest without GPU support - It is possible to build nvForest without GPU support; in this case, the CUDA toolkit is not required. - To build nvForest without GPU, set the CMake option ``NVFOREST_ENABLE_GPU=OFF``. - -**RAPIDS libraries.** -The nvForest code base is updated in tandem with the rest of RAPIDS. So to build the latest nvForest, you -should use the latest version of RAPIDS as well. (For example, nvForest 26.04 will require 26.04 version of -all RAPIDS packages.) - -**Python dependencies.** -Please see https://docs.rapids.ai/install/ for RAPIDS-wide version support. - -We aim to meet the `SPEC 0 guidelines `_ for minimal supported versions. - -**For development only.** - -* clang-format (= 20.1.8): enforces uniform C++ coding style; required for pre-commit hooks and CI checks. The packages ``clang=20`` and ``clang-tools=20`` from the conda-forge channel should be sufficient, if you are using conda. If not using conda, install the right version using your OS package manager. + It is possible to build nvForest without GPU support; in this case, the CUDA toolkit is not required. + To build nvForest without GPU, set the CMake option ``NVFOREST_ENABLE_GPU=OFF``. -.. note:: Use Conda to install all software dependencies +Using Conda to install all software dependencies +------------------------------------------------ - We highly recommend the use of Conda, a package manager that lets you obtain all necessary - software dependencies in a virtual environment. - We provide environment definition files ``conda/environments/all_*.yaml`` containing all software - dependencies for nvForest. +We highly recommend the use of Conda, a package manager that lets you obtain all necessary +software dependencies in a virtual environment. +We provide environment definition files ``conda/environments/all_*.yaml`` containing all software +dependencies for nvForest. - To create a development environment named ``nvforest_dev``, use the following commands. +To create a development environment named ``nvforest_dev``, use the following commands. - .. code-block:: console +.. code-block:: console $ conda create -n nvforest_dev python=3.13 $ conda env update -n nvforest_dev \ @@ -311,4 +291,4 @@ nvForest's cmake has the following configurable flags available: * - CMAKE_CUDA_ARCHITECTURES - List of GPU architectures, semicolon-separated - Empty - - List the GPU architectures to compile the GPU targets for. Set to "NATIVE" to auto detect GPU architecture of the system, set to "ALL" to compile for all RAPIDS supported archs. + - List the GPU architectures to compile the GPU targets for. Set to "NATIVE" to auto detect GPU architecture of the system, set to "ALL" to compile for all supported archs. diff --git a/docs/source/getting_started.rst b/docs/source/getting_started.rst index 8d48574..9a466f9 100644 --- a/docs/source/getting_started.rst +++ b/docs/source/getting_started.rst @@ -16,8 +16,9 @@ You can install nvForest using Pip or Conda. # Using Conda: need to specify the rapidsai channel $ conda install -c rapidsai -c conda-forge nvforest -You can also install nvForest as part of RAPIDS, a collection of libraries for GPU accelerated data science. -Visit https://docs.rapids.ai/install/ for more information. +Please see the `Installation Guide `_ +for NVIDIA CUDA-X libraries for data science for information about supported operating systems, +GPU drivers, and CUDA versions. nvForest with Python ==================== From 51adb3581c4e4a410d3252b9cc174d7a67d7cecb Mon Sep 17 00:00:00 2001 From: Hyunsu Cho Date: Mon, 14 Sep 2026 09:44:28 -0700 Subject: [PATCH 2/2] Fix hash for actions/checkout --- .github/workflows/link-checks.yaml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/.github/workflows/link-checks.yaml b/.github/workflows/link-checks.yaml index 96cb937..06ab5b9 100644 --- a/.github/workflows/link-checks.yaml +++ b/.github/workflows/link-checks.yaml @@ -20,7 +20,7 @@ jobs: contents: read steps: - name: Checkout repository - uses: actions/checkout@11d5960a326750d5838078e36cf38b85af677262 # v7.0.1 + uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1 with: persist-credentials: false - name: Restore lychee cache