How nvidia/cuml works
cuML is a GPU-accelerated machine-learning library with a native Python estimator API, a compatibility layer for accelerating existing scikit-learn/UMAP/HDBSCAN workflows, and selected distributed algorithms through Dask. The main user path is to create or supply data, fit/transform/predict with an estimator, and receive results in the configured output format. The graph preserves the README’s major capability areas and the documented CPU fallback. Many algorithm implementations are represented at their existing directories because their source was not sampled; edges reflect README statements or shown source references, not assumed internal call chains.
Main entry points: cuml Estimator API · estimator_proxy.py Estimator proxy · base.py Estimator runtime · outputs.py Output conversion
Components 23
GPU-native API
- Data generation
python/cuml/cuml/datasets - Estimator API
python/cuml/cuml
ML capabilities
- Clustering
python/cuml/cuml/cluster - Dimensionality reduction
python/cuml/cuml/decomposition - Regression and classification
python/cuml/cuml/linear_model - Preprocessing
python/cuml/cuml/preprocessing - Nearest neighbors
python/cuml/cuml/neighbors - Text features
python/cuml/cuml/feature_extraction/text.py - Model selection
python/cuml/cuml/model_selection - Model explanation
python/cuml/cuml/explainer - Metrics
python/cuml/cuml/metrics
Compatibility acceleration
- Accelerator entry
python/cuml/cuml/accel/__main__.py - Library patches
python/cuml/cuml/accel/_patches - Estimator proxy
python/cuml/cuml/accel/estimator_proxy.py - Profiling tools
python/cuml/cuml/accel/profilers.py
Distributed execution
- Dask algorithms
python/cuml/cuml/dask
Shared runtime
- Estimator runtime
python/cuml/cuml/internals/base.py - Input validation
python/cuml/cuml/internals/validation.py - Output conversion
python/cuml/cuml/internals/outputs.py - Global settings
python/cuml/cuml/internals/global_settings.py
Other components
- Python user
- Input data
- Model and results
Connections 28
- Python user → Data generation: generates data
README.md - Python user → Input data: supplies data
README.md - Data generation → Clustering: provides samples
README.md - Input data → Estimator API: fits or transforms
README.md - Estimator API → Model and results: returns results
README.md - Estimator API → Clustering: runs algorithms
README.md - Estimator API → Dimensionality reduction: runs algorithms
README.md - Estimator API → Regression and classification: runs algorithms
README.md - Estimator API → Preprocessing: runs transforms
README.md - Estimator API → Nearest neighbors: runs workflows
README.md - Estimator API → Text features: runs workflows
README.md - Estimator API → Model selection: runs workflows
README.md - Estimator API → Model explanation: runs workflows
README.md - Estimator API → Metrics: evaluates models
python/cuml/cuml/internals/mixins.py - Python user → Accelerator entry: starts accelerator
README.md - Accelerator entry → Library patches: patches libraries
- Library patches → Estimator proxy: wraps estimators
python/cuml/cuml/accel/_patches/sklearn/pipeline.py - Estimator proxy → Model and results: returns results
README.md - Estimator proxy → Estimator API: uses GPU estimators
python/cuml/cuml/accel/estimator_proxy.py - Estimator proxy → Python user: falls back to CPU
README.md - Python user → Profiling tools: checks execution
README.md - Python user → Dask algorithms: submits distributed work
README.md - Dask algorithms → Model and results: returns results
README.md - Estimator API → Estimator runtime: uses estimator base
python/cuml/cuml/internals/base.py - Estimator runtime → Input validation: validates inputs
- Estimator runtime → Output conversion: converts results
python/cuml/cuml/internals/base.py - Output conversion → Global settings: reads output mode
python/cuml/cuml/internals/outputs.py - Estimator proxy → Output conversion: converts outputs
python/cuml/cuml/accel/estimator_proxy.py