# New Release: Ultralytics v8.3.198

**URL:** <https://community.ultralytics.com/t/new-release-ultralytics-v8-3-198/1455>\
**Category:** Discussion\
**Tags:** ultralytics-official, releases, announcements\
**Created:** [September 11, 2025, 1:12pm UTC](https://community.ultralytics.com/t/new-release-ultralytics-v8-3-198/1455 "2025-09-11T13:12:58Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![glenn-jocher](https://sea1.discourse-cdn.com/flex001/user_avatar/community.ultralytics.com/glenn-jocher/32/83_2.png) [@glenn-jocher](https://community.ultralytics.com/u/glenn-jocher)\
**Post date:** [September 11, 2025, 1:12pm UTC](https://community.ultralytics.com/t/new-release-ultralytics-v8-3-198/1455/1 "2025-09-11T13:12:58Z")

</div>

# Ultralytics v8.3.198 — Smarter tuning, sturdier training, simpler exports 🚀

Summary  
Ultralytics v8.3.198 delivers a major upgrade to hyperparameter tuning with BLX-α crossover, unified metric plotting across tasks, safer training defaults, and multiple robustness fixes (NMS, DDP loss, Intel GPU checks). You’ll also find a cleaner export API and flexible `torch.compile` modes to fit your workflow. If you’re training or deploying YOLO today, this release is a win. 🙌

Highlights

- Smarter tuning: BLX-α crossover + adaptive mutation for faster convergence and stronger results.
- Unified metrics: Consistent `plot_results` across detect/segment/pose/classify.
- Safer, sturdier: NMS correctness, DDP unwrap before loss, environment-safe Intel GPU checks.
- Simpler exports: Export functions now return just the output path string.
- Flexible compile: `compile` accepts True/False or mode strings across train/val/predict.

Tip: YOLO11 is the latest stable and recommended Ultralytics model. If you’re considering YOLO12 or YOLO13, we recommend sticking with YOLO11 for best accuracy-speed-stability.

New Features

- Hyperparameter Tuner (priority)
  - Introduces BLX-α gene crossover for exploration across top parents, improving search quality, as detailed in PR [#22038](https://github.com/ultralytics/ultralytics/pull/22038). This also adds adaptive mutation sigma, safer bounds, `close_mosaic` in the search space, and uses only mAP@0.5:0.95 for fitness to align with common benchmarks. Runtime hygiene cleans GPU memory between iterations and improves resume/CSV/Mongo handling. 🎯

Improvements

- Exporter API cleanup
  - Export functions now return just the output path string (TF SavedModel also returns the Keras model) for lighter, cleaner pipelines, implemented in PR [#22009](https://github.com/ultralytics/ultralytics/pull/22009). 📦

- torch.compile flexibility
  - `compile` now supports True/False or mode strings `"default" | "reduce-overhead" | "max-autotune"` across train/val/predict, thanks to PR [#21999](https://github.com/ultralytics/ultralytics/pull/21999). ⚙

- Unified results plotting
  - `plot_results` auto-detects metrics/losses and works for all tasks with logic centralized in `BaseTrainer`, as unified in PR [#22026](https://github.com/ultralytics/ultralytics/pull/22026). 📈

- Training robustness
  - DDP/compiled models are unwrapped before loss calculation to avoid wrapper-related issues, addressed in PR [#22016](https://github.com/ultralytics/ultralytics/pull/22016). 🧩

- NMS correctness
  - Fixed early-exit and sorting in TorchNMS to reduce false positives and improve stability, delivered in PR [#22014](https://github.com/ultralytics/ultralytics/pull/22014). ✅

Bug Fixes and Stability

- Segmentation
  - Correct mask resizing/shape handling in validation (e.g., `mask_ratio=1`) and faster, correct plotting with `overlap_mask`, fixed across PRs [#22037](https://github.com/ultralytics/ultralytics/pull/22037) and [#22031](https://github.com/ultralytics/ultralytics/pull/22031). 🖼

- Environment resilience
  - Dataset cache auto-recovers on `ModuleNotFoundError` in PR [#22025](https://github.com/ultralytics/ultralytics/pull/22025) and Intel GPU discovery now catches all exceptions to prevent environment crashes, per PR [#22034](https://github.com/ultralytics/ultralytics/pull/22034).

- Docs & configs
  - Clearer configs and tracker YAMLs, improved compile args docs, corrected detection boxes column order (track\_id position), and a lighter Quickstart (Seaborn no longer required), improved across PRs [#22011](https://github.com/ultralytics/ultralytics/pull/22011), [#22028](https://github.com/ultralytics/ultralytics/pull/22028), and [#22035](https://github.com/ultralytics/ultralytics/pull/22035).
  - The Construction-PPE dataset docs now include a “Business Value” section to help justify real-world ROI, added in PR [#22029](https://github.com/ultralytics/ultralytics/pull/22029). 👷📈

Quick Examples

- Tuning with the improved Tuner:

```python
from ultralytics import YOLO

model = YOLO("yolo11s.yaml")
model.tune(
    device=0,
    data="coco128.yaml",
    optimizer="AdamW",
    epochs=100,
    batch=8,
    compile=False,
    plots=False,
    val=False,
    save=False,
    workers=16,
    project="tune-yolo11s-scratch-coco128-100e",
    iterations=1000,
)

```

- Using compile modes:

```python
from ultralytics import YOLO

model = YOLO("yolo11n.pt")
model.train(data="coco8.yaml", epochs=3, compile="reduce-overhead") # or "default", "max-autotune", True/False

```

- Export now returns a file path:

```python
from ultralytics import YOLO

model = YOLO("yolo11n.pt")
onnx_path = model.export(format="onnx") # 'onnx_path' is a string

```

Upgrade

- Update to the latest release with:

```bash
pip install -U ultralytics

```

- You can read the full release notes in the v8.3.198 entry on the Ultralytics Releases page, available via the dedicated link to the [v8.3.198 release](https://github.com/ultralytics/ultralytics/releases/tag/v8.3.198).

What’s Changed (PRs)

- Clean up `Exporter` and remove unnecessary `None` placeholder in PR [#22009](https://github.com/ultralytics/ultralytics/pull/22009) by [Laughing-q](https://github.com/Laughing-q).
- Allow settable compile mode for `torch.compile` in PR [#21999](https://github.com/ultralytics/ultralytics/pull/21999) by [Laughing-q](https://github.com/Laughing-q).
- Improve config YAMLs in PR [#22011](https://github.com/ultralytics/ultralytics/pull/22011) by [glenn-jocher](https://github.com/glenn-jocher).
- Unwrap DDP model before loss calculation in PR [#22016](https://github.com/ultralytics/ultralytics/pull/22016) by [Y-T-G](https://github.com/Y-T-G).
- Fix `TorchNMS.nms()` early exit logic in PR [#22014](https://github.com/ultralytics/ultralytics/pull/22014) by [Y-T-G](https://github.com/Y-T-G).
- Revise detection boxes argument documentation in PR [#22028](https://github.com/ultralytics/ultralytics/pull/22028) by [daniel-mooney](https://github.com/daniel-mooney).
- Fix mask plotting when number of objects equals number of images in a batch in PR [#22031](https://github.com/ultralytics/ultralytics/pull/22031) by [Y-T-G](https://github.com/Y-T-G).
- Remove Seaborn from manual installation dependencies in Quickstart docs via PR [#22035](https://github.com/ultralytics/ultralytics/pull/22035) by [onuralpszr](https://github.com/onuralpszr).
- Catch all exceptions during Intel GPU discovery in PR [#22034](https://github.com/ultralytics/ultralytics/pull/22034) by [Y-T-G](https://github.com/Y-T-G).
- Add Business Value section to Construction-PPE dataset docs in PR [#22029](https://github.com/ultralytics/ultralytics/pull/22029) by [UltralyticsAbi](https://github.com/UltralyticsAbi).
- Reset labels cache when loading with incompatible NumPy version in PR [#22025](https://github.com/ultralytics/ultralytics/pull/22025) by [Y-T-G](https://github.com/Y-T-G).
- Faster length retrieval for `torch.Tensor` using `shape[0]` in PR [#22021](https://github.com/ultralytics/ultralytics/pull/22021) by [Laughing-q](https://github.com/Laughing-q).
- Clean up and unify `plot_results` across detect/segment/pose/obb/classify in PR [#22026](https://github.com/ultralytics/ultralytics/pull/22026) by [Laughing-q](https://github.com/Laughing-q).
- Fix mask shape mismatch when validating with `mask_ratio=1` in PR [#22037](https://github.com/ultralytics/ultralytics/pull/22037) by [Laughing-q](https://github.com/Laughing-q).
- Improve Tuner with BLX-α gene crossover in PR [#22038](https://github.com/ultralytics/ultralytics/pull/22038) by [glenn-jocher](https://github.com/glenn-jocher).

New Contributors

- We’re excited to welcome [daniel-mooney](https://github.com/daniel-mooney), who contributed via PR [#22028](https://github.com/ultralytics/ultralytics/pull/22028).
- A warm welcome to [UltralyticsAbi](https://github.com/UltralyticsAbi), who contributed via PR [#22029](https://github.com/ultralytics/ultralytics/pull/22029).

Try it and Share Feedback

- You can explore the complete diff in the [full changelog for v8.3.198](https://github.com/ultralytics/ultralytics/compare/v8.3.197...v8.3.198), and if you spot anything, please start a conversation in [Ultralytics Discussions](https://github.com/orgs/ultralytics/discussions) or open an issue on the repository.
- Huge thanks to our community and contributors for making each release better—your feedback helps us refine YOLO for everyone. 💛