Pass LOCAL_RANK to torch_distributed_zero_first()

[qiningonline]

resolve #5111

The change proposed in this pull request:

• passing local_rank to torch_distributed_zero_first, to avoid CUDA device index error

Details of the issue are listed in the issue description and follow-up comment.

🛠️ PR Summary

_{Made with ❤️ by Ultralytics Actions}

🌟 Summary

Enhancement of multi-GPU training support in YOLOv5.

📊 Key Changes

• Replaced RANK with LOCAL_RANK to better support distributed training on multiple GPUs.

🎯 Purpose & Impact

• 🤝 Improved Compatibility: These changes enhance the compatibility with distributed training frameworks, allowing for more efficient utilization of multiple GPUs.
• ⚡ Increased Efficiency: Better support for local GPU rank handling potentially leads to improved parallel processing performance and reduced training times.
• 👨‍💻 Developer Experience: For developers using YOLOv5 for training on multiple GPUs, this update simplifies processes and could help avoid common issues related to device ranking in distributed environments.

[github-actions]

👋 Hello @qiningonline, thank you for submitting a 🚀 PR! To allow your work to be integrated as seamlessly as possible, we advise you to:

• ✅ Verify your PR is up-to-date with origin/master. If your PR is behind origin/master an automatic GitHub actions rebase may be attempted by including the /rebase command in a comment body, or by running the following code, replacing 'feature' with the name of your local branch:

git remote add upstream https://github.com/ultralytics/yolov5.git
git fetch upstream
git checkout feature  # <----- replace 'feature' with local branch name
git merge upstream/master
git push -u origin -f

• ✅ Verify all Continuous Integration (CI) checks are passing.
• ✅ Reduce changes to the absolute minimum required for your bug fix or feature addition. "It is not daily increase but daily decrease, hack away the unessential. The closer to the source, the less wastage there is." -Bruce Lee

[glenn-jocher]

@qiningonline PR is merged. Thank you for your contributions to YOLOv5 🚀 and Vision AI ⭐