New global TQDM_BAR_FORMAT

[glenn-jocher]

@AyushExel centralized TQDM bar format defined in general.py and used everywhere. Slightly simpler format, removed iteration and time remaining.

🛠️ PR Summary

_{Made with ❤️ by Ultralytics Actions}

🌟 Summary

Refactor progress bar format in YOLOv5's training and validation scripts.

📊 Key Changes

• Introduced a new global variable TQDM_BAR_FORMAT to standardize the tqdm progress bar format.
• Replaced hardcoded tqdm bar format strings with the new global TQDM_BAR_FORMAT variable across multiple scripts (train.py, validate.py, classify/train.py, classify/val.py, segment/train.py, segment/val.py, utils/autoanchor.py, and utils/dataloaders.py).

🎯 Purpose & Impact

• Consistency and Maintainability: The use of a global variable for the progress bar format ensures consistency across different modules and makes future changes to the format simpler and more maintainable.
• Enhanced User Experience: The updated progress bar format includes additional information like total number and elapsed time, providing a more informative and user-friendly interface during model training and validation.
• Potential Decrease in Visual Clutter: By standardizing the progress bar, the visual output during long training processes is potentially less cluttered, with important metrics being easier to read.