Metrics YOLOv8

[Simon334]

Is it possible that the metrics that are generated when training the YOLOv8 model have signed x and y axes. If so, how can this be done?

[UltralyticsAssistant]

👋 Hello @Simon334, thank you for your interest in Ultralytics 🚀! Metrics play a crucial role in model evaluation, and your question touches on an interesting topic.

If this is a ❓ Question regarding custom metrics or training, please provide as much detail as possible, including examples of dataset images, training configurations, and logs, so we can gain better insights into the scenario.

If this is a 🐛 Bug Report, it would be very helpful to have a minimum reproducible example (MRE) which includes a clear description of how to reproduce the issue.

For best practices, ensure your environment is updated with the latest ultralytics package and its requirements in a Python>=3.8 setup. You can verify performance in environments such as Colab, Kaggle, or other platforms you may be using.

An Ultralytics engineer will review your discussion soon and assist you further! Feel free to join our community spaces like Discord, Discourse, or Reddit for additional insights from other members of the Ultralytics community 🎧.

[glenn-jocher]

@Simon334 the metrics generated during YOLOv8 training, such as loss and mAP, don't inherently include signed x and y axes. If you need custom metrics or signed axes for specific tasks, you'll need to modify the training script or post-process the outputs. For guidance, you can refer to the Ultralytics documentation: https://docs.ultralytics.com.