What are those S3 classes made for? | Datasets, items, image and targets | Transform ordering | Classification and caption datasets and their items | Detection datasets and their items | Rotated box items | Segmentation datasets and their items | Slicing an image: sahi_split | Which function takes which class | Building an item by hand | What models return | Comparison with torchvision TVTensors (advanced)

A grid of images | Bounding boxes | Object detection with RF-DETR | Semantic segmentation masks | Keypoints

Check the system requirements | Install cuda.ml and its runtime | Inspect the backend | Fit and predict with a direct model API | Compute a lower-dimensional representation | Understand returned values | Choose the direct API or parsnip | Continue

Prepare the runtime | Train on GPU and deploy on CPU | Load external model formats | Select CPU or GPU inference | Predict regression and classification results | Serve predictions with plumber2 | Inspect a model | Inspect leaves and individual trees | Persist as a single R state

Choose a persistence format | Save directly to a file | Keep the state as raw bytes | Use a bundle | Export an nvForest checkpoint pair | Select the nvForest restore device | Treat model artifacts as trusted input

Available model specifications | Set the engine | Preprocess with a recipe | Tune common arguments | Linear regression routing | Choose between parsnip and the direct API

Install a prebuilt backend | Supported systems | Windows through WSL2 | Place or mirror the cache | Inspect, audit, or remove a backend | Build the complete backend from source | Use host dependencies | Troubleshooting

Data format | Data preparation | Node preparation rules for {tabnet} models | Avoid factor predictors | Avoid column name collision with reserved {data.tree} names | Avoid column named level_* to avoid collision with output data.tree names | Ensure the last hierarchy of the tree is the observation id | Ensure there is a root level in the hierarchy | Ensure there is no missing values in the hierarchical classes | Model building | Data set split | Model diagnostic | Model inference

Experiments | Datasets | Syn2 | Syn 4

Motivation | Missing-data dataset creation | Ames missings understanding | While keeping some room for freedom | Ames with missing data | Model pretraining | Variable importance with raw ames dataset | Variable importance with ames_missing dataset | Model training | Conclusion

Data preprocessing | Self-supervised training step | Continuing training with supervised task | Comparing against a model without pretraining

Overview | Quick Search | Example: Finding a Photovoltaic Dataset | Complete Catalog | Collections | Biology Collection (9 datasets) | Medical Collection (8 datasets) | Infrared Collection (4 datasets) | Damage Collection (3 datasets) | Underwater Collection (4 datasets) | Document Collection (6 datasets) | Usage Example | Dataset Statistics | Filtering and Exploration | Additional Resources | Citation

Windows | CPU | GPU | MacOS | Linux | Installing from pre-built binaries | Using cudatoolkit R packages | Environment variables | Troubleshooting | Debug installation messages | Large file download timeout | File based download | Installing older versions

Support for glue

Output | Using the chattr() function | Highlight the request in a script

Introduction | How imbalance is my problem ? | Solutions to improve imbalanced classification models | Using the AUC metric and pr_curve() plots | Case-weight | ROC_AUM loss | All together

Multiple optimizers | Fully flexible step | Next steps

Training a nn_module | The training loop | Metrics | Evaluate | Customizing with callbacks | Next steps

Example