data.yaml names.keys() integer assert

[glenn-jocher]

Check to avoid #10180

Signed-off-by: Glenn Jocher glenn.jocher@ultralytics.com

🛠️ PR Summary

_{Made with ❤️ by Ultralytics Actions}

🌟 Summary

Improved data integrity checks for dataset labels in YOLOv5.

📊 Key Changes

• Converted the label format in the dataset configuration from a list or tuple to a dictionary if not already in dictionary form.
• Added an assertion to ensure that all keys in the data['names'] dictionary are integers, reinforcing consistent label formatting.

🎯 Purpose & Impact

• Ensures that the names of the classes in the dataset configuration (data.yaml) are always stored in a dictionary format with integer keys, which is more reliable and less prone to errors. 🛠️
• The added integer key assertion prevents potential issues with class indexing, leading to more robust and predictable behavior when training models. 📈
• Users will benefit from more stable input data handling, which can improve the training process's efficiency and accuracy. 👩‍💻👨‍💻