ultralytics 8.3.13 on Python PyPI

🌟 Summary

The 'v8.3.13' release primarily focuses on refactoring the Segment Anything Model (SAM) to enhance prompt handling and documentation clarity.

📘 Documentation improvements with updated examples for the model method.
🔄 Code refactoring introduces a new function _prepare_prompts to manage prompt preparations in SAM.
📦 Enhanced prompt handling for bounding boxes, points, labels, and masks.
🗑️ Cleaned up duplicated logging in YOLO training for simpler outputs.
💤 Added a 60-second delay to improve deployment stability in GitHub Actions.
📊 Fixed subplot indexing for clearer result plots in classification, segmentation, and pose.
🍇 Updated Raspberry Pi guide for YOLO11.
📈 With every precision metric, ensure it's returned for consistency with or without plotting.

📝 Documentation Clarity: Clearer examples help users better understand model functionalities.
🔧 Efficiency & Maintainability: By moving prompt processing to a new function, the code is now easier to maintain and modify in the future.
🚀 User Experience: Improved multiple points and labels handling in SAM provides a better user experience for image segmentation tasks.
📉 Logging: Cleaner training logs mean users can more easily follow the training progress without unnecessary clutter.
🛠️ Stability & Reliability: Delaying deployment steps helps ensure smoother execution in CI/CD workflows, thus increasing user trust in automated deployments.
📊 Visual Clarity: Fixes to plotting result in more intuitive and easily interpretable visual outputs, aiding better analysis and decision-making.
🥝 Guides Update: Raspberry Pi users can leverage the latest YOLO11 advances with tailored guidance, benefiting from performance improvements.

Full Changelog: v8.3.12...v8.3.13