fix: try 2 - prevent logging config clobbering

[rkechols]

Fix for issue #10132

• Previous behavior: loading this repository with torch.hub.load clobbers the existing logging configuration by modifying the root logger's configuration.
• New behavior: loading this repository with torch.hub.load only clobbers the logging configuration for logger yolov5 and its descendants. This is done in a way compatible with Google Colab.

The new set_logging function uses "disable_existing_loggers": False so that any existing logging configurations are left unchanged. It also uses "propagate": False on the logger in question so that any log records sent to this logger are not also sent through any handlers from pre-existing logging configurations (which would duplicate messages).

🛠️ PR Summary

_{Made with ❤️ by Ultralytics Actions}

🌟 Summary

Improved logging configurability in yolov5 utils.

📊 Key Changes

• Added logging.config import to manage logging configurations.
• Created a new constant LOGGING_NAME set to "yolov5" to unify logger naming.
• Redefined set_logging function to use logging.config.dictConfig for logging setup.
• Streamlined the process of defining log formatters, handlers, and loggers.

🎯 Purpose & Impact

• ✨ Enhanced Customization: Utilizing dictConfig allows for more advanced logging customization and potential future expansion of logging capabilities.
• 🔄 Consistency and Maintenance: Introduction of LOGGING_NAME aids in maintaining consistent logger references across the codebase.
• 🛡️ Isolation: Newly configured loggers do not propagate messages to parent loggers, reducing potential logging noise.
• 🔍 Debugging and Control: The changes provide better control over the logging levels and output formats, which can aid in debugging and monitoring the application behavior, especially in multi-GPU training scenarios.

[glenn-jocher]

@rkechols looks good! PR is merged. Thank you for your contributions to YOLOv5 🚀 and Vision AI ⭐