This repository contains a collection of CoreML demo apps, with optimized models for the Apple Neural Engine™️. It also hosts tutorials and other resources you can use in your own projects.
depth-anything.mp4 |
fastvit.mp4 |
DETR.mp4 |
The models showcased include:
| Sample Code | Task | Core ML packages |
|---|---|---|
| FastViT | Image Classification | apple/coreml-FastViT-T8 apple/coreml-FastViT-MA36 |
| Depth Anything V2 (small) | Monocular Depth Estimation | apple/coreml-depth-anything-v2-small |
| DETR (ResNet 50) | Semantic Segmentation | apple/coreml-detr-semantic-segmentation |
We leverage coremltools for testing and implementing these optimisations. You can read more about it here.
- How to convert Depth Anything v2 for GPU and Neural Engine. Source code notebook
