Add git info to cls, seg checkpoints

[glenn-jocher]

🛠️ PR Summary

_{Made with ❤️ by Ultralytics Actions}

🌟 Summary

Enhanced training scripts to include Git repository information in model checkpoints.

📊 Key Changes

• Added GIT constant (containing remote, branch, commit info) to the saved checkpoints in train.py, classify/train.py, and segment/train.py.
• Removed some commented-out Weights & Biases related arguments from segment/train.py.

🎯 Purpose & Impact

• Purpose: To provide better tracking and reproducibility of model training by including the repository's state within the saved checkpoints.
• Impact: Users can now identify which exact version of the code was used to train a model, making it easier to replicate results or diagnose issues related to specific commits. 🕵️‍♂️✅
• The removed W&B code lines indicate a leaner configuration in segment/train.py, potentially indicating a shift away from specific W&B features or a cleanup of the code.