Revert torch.hub.load() test

[glenn-jocher]

Temporarily reverts #4978 until torch 1.10 is released, which should resolve urllib.error.HTTPError: HTTP Error 403: rate limit exceeded errors generated by torch hub from GitHub actions runners that is currently causing CI failures.

See pytorch/vision#4156 for details.

🛠️ PR Summary

_{Made with ❤️ by Ultralytics Actions}

🌟 Summary

Improving CI workflow robustness by noting a known HTTP 403 error. 🛠️

📊 Key Changes

• Commented out the line in CI testing that loads a model from Ultralytics's GitHub to avoid a known urllib.error.HTTPError: HTTP Error 403: rate limit exceeded.
• Added a comment explaining the issue and indicating it will be resolved in the upcoming torch==1.10.0 release.

🎯 Purpose & Impact

• 🔗 The change prevents CI from failing due to GitHub rate limiting, which is important for maintaining a smooth continuous integration process.
• 🛡️ This is a temporary workaround until the issue is resolved in PyTorch's future release, ensuring developers are aware of the limitation and the CI testing does not get obstructed by external HTTP errors.
• 🚀 For users, this means better reliability in the project's development workflow, potentially leading to quicker releases and bug fixes since CI plays a crucial role in the quality control process.