skills.rest
dotlab-hq
dotlab-hqOFFICIAL
Explore all skills by this author →
GitHub

cuml-machine-learning

Train GPU-accelerated machine learning models on tabular datasets with CPU fallback.

Updated Apr 12, 2026
Search Tags:#classification#gpu#machine-learning#dimensionality-reduction#rapids#cuml#tabular-data
One-click install
npx skills add https://github.com/dotlab-hq/torque --skill cuml-machine-learning-dotlab-hq
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: cuml-machine-learning
Source: https://github.com/dotlab-hq/torque/tree/main/.agents/skills/cuml-machine-learning
Command: npx skills add https://github.com/dotlab-hq/torque --skill cuml-machine-learning-dotlab-hq

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill reduces training and inference time for machine learning on large tabular datasets by leveraging NVIDIA cuML with a safe CPU fallback, making iterative modeling and experimentation practical on datasets where GPU acceleration yields significant speedups.

Core Features & Use Cases

  • GPU-accelerated training and inference for common supervised and unsupervised algorithms (classification, regression, clustering).
  • Compatibility and fallback: attempts cuDF/cuML and falls back to scikit-learn when GPU libraries are unavailable.
  • Dimensionality reduction & preprocessing: PCA, UMAP (GPU only), scaling and label encoding for large feature sets.
  • Use Case: Train a RandomForestClassifier on a 100k-row customer dataset with features cast to float32 to get rapid model iteration and feature importance ranking.

Quick Start

Use the cuml-machine-learning skill to train a GPU-accelerated model on your tabular dataset with numeric features cast to float32, fallback to scikit-learn if cuML is unavailable, and report train/test metrics.

Frequently Asked Questions about cuml-machine-learning

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I accelerate machine learning training on large tabular datasets?▼

To accelerate machine learning on large tabular datasets, use NVIDIA cuML for GPU-accelerated training and inference. This approach handles classification, regression, and clustering, falling back to scikit-learn if GPU libraries are unavailable.

Can I use GPU-accelerated UMAP and PCA for dimensionality reduction with cuML?▼

Yes, GPU-accelerated UMAP and PCA are supported for dimensionality reduction with cuML. This Skill applies these algorithms to large feature sets, though UMAP requires an NVIDIA GPU while other preprocessing tasks support CPU fallback.

Do I need an NVIDIA GPU to run cuML machine learning models?▼

An NVIDIA GPU is not strictly required to run cuML machine learning models. The Skill attempts to use cuDF and cuML for GPU acceleration but automatically falls back to scikit-learn on a CPU if the GPU libraries are unavailable.

What data format is required for GPU-accelerated machine learning with cuML?▼

GPU-accelerated machine learning with cuML requires dense numeric features cast to float32 or float64. Sparse matrices are not supported, and the dataset typically needs 10,000 or more rows to benefit from GPU speedups.

Why does my cuML training fall back to scikit-learn instead of using the GPU?▼

cuML training falls back to scikit-learn when NVIDIA GPU libraries are unavailable or improperly configured. The Skill safely defaults to CPU-based scikit-learn to ensure processing continues without interruption.

Related Skills

Explore More Data & Analytics Skills→
MrNeo01MrNeo01

cuml-machine-learning

Accelerate tabular machine learning with NVIDIA cuML GPU-accelerated training and inference.

Community
Intermediate
root-5root-5

validation

Compute ML metrics, cross-validate, and analyze residuals.

Community
Intermediate
sahiixxsahiixx
2

cuml-machine-learning

Run GPU-accelerated classification, regression, clustering, and dimensionality reduction on tabular data with cuML.

Community
Advanced
leesk212leesk212
1

cuml-machine-learning

Train and evaluate GPU-accelerated machine learning models on tabular data with cuML.

Community
Intermediate