Practical, license-reviewed examples for learning, testing, and demonstrating AnyLearning. Each recipe points to a small public dataset, explains the matching project type, and uses files that can be uploaded directly in the AnyLearning interface.
Note
AnyLearning is a local-first application. These examples download data to your machine; they do not upload it to an external training service.
| Task | Example | Dataset license |
|---|---|---|
| Image classification | Chest X-ray | CC BY 4.0 |
| Object detection | Safety helmet and jacket | Apache 2.0 |
| Semantic segmentation | Electron-microscopy particles | CC BY 4.0 |
| Handpose classification | ASL letters | Public Domain |
| Keypoint detection | Generated stick figures | CC0 1.0 |
| Keypoint detection | Vertebral landmarks | CC BY 4.0 |
| Structured AI | Tabular, text, and response evaluation | CC BY 4.0 / Apache 2.0 |
Install AnyLearning, then clone this repository:
git clone https://github.com/nrl-ai/anylearning-examples.git
cd anylearning-examples
python scripts/download_dataset.py helmet-jacket --output downloads/helmet-jacketCreate an Object Detection project and upload the generated train.zip,
val.zip, and test.zip files to their corresponding splits. Each example
directory contains the exact labels, import steps, and training suggestions.
List all available downloads with:
python scripts/download_dataset.py --listExamples should be reproducible, small enough to try locally, and backed by a
license that permits redistribution. Large archives remain in the
nrl-ai/anylearning-data
dataset repository. See SOURCES.md for provenance and licenses.
The code in this repository is licensed under Apache 2.0. Dataset licenses are independent and are recorded beside every example.