# New Release: Ultralytics v8.4.178 **URL:** \ **Category:** Discussion\ **Tags:** releases, announcements, ultralytics-official\ **Created:** [October 16, 2026, 1:09pm UTC](https://community.ultralytics.com/t/new-release-ultralytics-v8-4-178/2267 "2026-10-16T13:09:57Z")\ **Posts on this page:** 1\ **Page:** 1
**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:** [October 16, 2026, 1:09pm UTC](https://community.ultralytics.com/t/new-release-ultralytics-v8-4-178/2267/1 "2026-10-16T13:09:57Z")
# 🚀 Ultralytics v8.4.178: More Reliable Platform Links and Stronger Workflows **Ultralytics YOLO v8.4.178 is here!** This release improves Platform training links and strengthens validation, tracking, exports, and training integrations. It also introduces a benchmark resource covering **37 real-world detection datasets**. ## 🌟 Highlights - **Reliable training links:** [PR #26699 by @glenn-jocher](https://github.com/ultralytics/ultralytics/pull/26699) uses the model URL returned by Ultralytics Platform, reflecting the final owner, project, and model name. If the server provides no URL, no link is printed. - **Broader evaluation coverage:** [PR #26691 by @Laughing-q](https://github.com/ultralytics/ultralytics/pull/26691) adds the Ultralytics Benchmark page with a fixed evaluation protocol across 37 detection datasets, including medical, retail, industrial, and aerial imagery. This is an evaluation resource—not a new model architecture. - **Better YOLO-NAS compatibility:** [PR #26670 by @raimbekovm](https://github.com/ultralytics/ultralytics/pull/26670) fixes loading by model name, prediction, validation, export, and ReID tracking. ## 🎯 Model, Tracking, and Validation Fixes - **Persistent prompts and object IDs:** [PR #26676 by @raimbekovm](https://github.com/ultralytics/ultralytics/pull/26676) keeps FastSAM prompts active across later frames and preserves SAM2 object IDs when masks are blank. - **More consistent oriented-box tracking:** [PR #26680 by @raimbekovm](https://github.com/ultralytics/ultralytics/pull/26680) handles equivalent box representations without disrupting tracking. - **Correct static-export validation:** [PR #26690 by @Nikhi00718](https://github.com/ultralytics/ultralytics/pull/26690) respects recorded rectangular input shapes. - **Safer SAM mask prompts:** [PR #26673 by @cainiao33](https://github.com/ultralytics/ultralytics/pull/26673) fixes a crash by downsampling prompts to the encoder’s low-resolution input size. - **Preserved segmentation detail:** [PR #26667 by @Vaishnavi220506](https://github.com/ultralytics/ultralytics/pull/26667) retains distinct grounding polygons that share the same bounding box. ## 📦 Export and Hardware Improvements - **TorchScript on Apple MPS:** [PR #26693 by @raimbekovm](https://github.com/ultralytics/ultralytics/pull/26693) fixes loading by converting traced `float64` scalars to `float32`. - **Dynamic ONNX on MPS:** [PR #26686 by @raimbekovm](https://github.com/ultralytics/ultralytics/pull/26686) keeps dynamic-shape nodes off the CoreML execution provider. - **Full-integer LiteRT inputs:** [PR #26674 by @aswanth-07](https://github.com/ultralytics/ultralytics/pull/26674) clamps input values before casting. - **Ascend NPU training:** [PR #26687 by @Eric-Lotus](https://github.com/ultralytics/ultralytics/pull/26687) fixes an unsupported reduction in `TaskAlignedAssigner`. Welcome to your first contribution! 🎉 ## 📊 Platform and Experiment Reporting - **Aligned class labels:** [PR #26665 by @JaviChulvi](https://github.com/ultralytics/ultralytics/pull/26665) sends class indices with Platform validation curves when validation data omits classes. - **Task-specific curves:** [PR #26678 by @JaviChulvi](https://github.com/ultralytics/ultralytics/pull/26678) sends segmentation and pose validation curves from the correct task. - **Correct Comet confusion matrices:** [PR #26669 by @raimbekovm](https://github.com/ultralytics/ultralytics/pull/26669) fixes matrix orientation and labels. - **Cleaner TensorBoard teardown:** [PR #26682 by @Vaishnavi220506](https://github.com/ultralytics/ultralytics/pull/26682) closes writers after training. - **Default MLflow tracking:** [PR #26688 by @raimbekovm](https://github.com/ultralytics/ultralytics/pull/26688) uses MLflow’s default tracking store when no tracking URI is configured. ## 🛠 Training, Data, and Workflow Reliability - **Safer Docker publishing:** [PR #26666 by @glenn-jocher](https://github.com/ultralytics/ultralytics/pull/26666) prevents overwriting image tags built from newer commits. - **Preserved optimizer precision:** [PR #26672 by @cainiao33](https://github.com/ultralytics/ultralytics/pull/26672) retains Adam’s `exp_avg` momentum in FP32 during optimizer FP16 conversion. - **UTF-8 datasets:** [PR #26681 by @Nikhi00718](https://github.com/ultralytics/ultralytics/pull/26681) explicitly reads NDJSON datasets with UTF-8 encoding. - **More robust source loading:** [PR #26671 by @cainiao33](https://github.com/ultralytics/ultralytics/pull/26671) handles UTF-8 BOM source manifests and adds a URL timeout to `autocast_list`. - **Screenshot inputs:** [PR #26664 by @Nikhi00718](https://github.com/ultralytics/ultralytics/pull/26664) fixes prediction with monitor and region parameters. - **Configuration directories:** [PR #26692 by @raimbekovm](https://github.com/ultralytics/ultralytics/pull/26692) creates a missing `YOLO_CONFIG_DIR` instead of falling back to `/tmp`. - **Correct ZIP names:** [PR #26694 by @Vaishnavi220506](https://github.com/ultralytics/ultralytics/pull/26694) preserves full directory names when creating archives. - **Clearer CLI errors:** [PR #26675 by @glenn-jocher](https://github.com/ultralytics/ultralytics/pull/26675) simplifies argument errors and excludes callables from suggestions. - **Cleaner solution output:** [PR #26697 by @MohammadHijjawi97](https://github.com/ultralytics/ultralytics/pull/26697) skips empty workout lists when printing `SolutionResults`. ## 📚 Documentation Updates - **Annotation model coverage:** [PR #26668 by @glenn-jocher](https://github.com/ultralytics/ultralytics/pull/26668) lists all Platform class-prompted annotation models. - **Image-generation guidance:** [PR #26683 by @glenn-jocher](https://github.com/ultralytics/ultralytics/pull/26683) documents paid image-generation models. - **Platform workflows:** [PR #26677 by @raimbekovm](https://github.com/ultralytics/ultralytics/pull/26677) documents Explore filters, AI-assisted training diagnosis, and API details. - **Improved visuals:** [PR #26684 by @glenn-jocher](https://github.com/ultralytics/ultralytics/pull/26684) adds updated Ultralytics Academy task and guide images. - **Documentation corrections:** [PR #26685 by @glenn-jocher](https://github.com/ultralytics/ultralytics/pull/26685) fixes two figures and retraining-chart alt text. ## 💬 Try It and Share Your Feedback Install this release with: ```bash pip install --upgrade ultralytics==8.4.178 ``` Explore the [v8.4.178 release notes](https://github.com/ultralytics/ultralytics/releases/tag/v8.4.178) for the release overview, or review the [full changelog](https://github.com/ultralytics/ultralytics/compare/v8.4.177...v8.4.178) for every change. **How does v8.4.178 work with your models, exports, and Platform workflows?** Share your results and any issues below. Thanks to the YOLO community and Ultralytics team for keeping these improvements coming! 🙌