1.8.1

## 📋 Summary

RF-DETR 1.8.1 is a patch release with seven bug fixes and three quality-of-life improvements. The highlights: two compatibility crashes are fixed — one hitting environments that have `tensorflow` alongside NumPy 2.0, another hitting Albumentations 2.x users — and per-epoch metric tables show up again during training instead of being hidden behind the progress bar. No breaking changes, and nothing to migrate from v1.8.0.

## 🗒️ Notable changes

### 🌱 Changed

- **`pathlib.Path` accepted for path parameters** — path fields on `TrainConfig` and related config classes now accept `pathlib.Path` objects directly. ([#1124](https://github.com/roboflow/rf-detr/pull/1124))
- **Horizontal flip disabled for keypoint training** — flip augmentation is turned off for keypoint training for now, since it could produce incorrect labels without keypoint flip-pair reordering. ([#1122](https://github.com/roboflow/rf-detr/pull/1122))
- **Training metric plots** — improved plotting with optional error bands, AP@0.75 grouping, and configurable AP metric groups. ([#1122](https://github.com/roboflow/rf-detr/pull/1122))

### 🔧 Fixed

- **TensorBoard + NumPy 2.0 crash** — training falls back to CSV logging with a warning instead of crashing when TensorBoard can't be imported. ([#1123](https://github.com/roboflow/rf-detr/pull/1123))
- **Albumentations 2.x compatibility** — fixed a crash in horizontal-flip detection when using Albumentations 2.x. ([#1126](https://github.com/roboflow/rf-detr/pull/1126))
- **Metric tables hidden by the progress bar** — per-epoch metric tables now render correctly alongside the live training progress bar. ([#1128](https://github.com/roboflow/rf-detr/pull/1128))
- **Keypoint checkpoint selection** — best-checkpoint selection for keypoint models is now smoothed, avoiding spurious switches on noisy OKS metrics; smoothing also survives a training resume. ([#1122](https://github.com/roboflow/rf-detr/pull/1122))
- **Keypoint encoder evaluation** — fixed query routing in eval mode so keypoint predictions are computed correctly. ([#1135](https://github.com/roboflow/rf-detr/pull/1135))
- **Group DETR train-time evaluation** — fixed a crash during train-time evaluation in Group DETR mode. ([#1122](https://github.com/roboflow/rf-detr/pull/1122))
- **DINOv2 backbone config compatibility** — internal call sites updated from deprecated `config.use_return_dict` to `config.return_dict`. ([#1135](https://github.com/roboflow/rf-detr/pull/1135))

---

## 🏆 Contributors

Thanks to everyone who contributed to this release:

- **Jirka Borovec** ([@Borda](https://github.com/Borda), [LinkedIn](https://www.linkedin.com/in/jirka-borovec/)) — keypoint encoder fix, metric table display, training stability, TensorBoard compatibility
- **Juan Pablo Oberhauser** ([@jpoberhauser](https://github.com/jpoberhauser), [LinkedIn](https://www.linkedin.com/in/jp-oberhauser/)) — Albumentations 2.x compatibility
- **Brian Cong** ([@congbrian](https://github.com/congbrian)) — `pathlib.Path` config support
- **Stefan Schneider** ([@hinogi](https://github.com/hinogi)) — type hint modernisation

---

**Full changelog**: https://github.com/roboflow/rf-detr/compare/1.8.0...v1.8.1