Update README.md

[glenn-jocher]

🛠️ PR Summary

_{Made with ❤️ by Ultralytics Actions}

🌟 Summary

Enhancements to THOP: PyPI badges added, improved error handling in benchmarking.

📊 Key Changes

• Added PyPI version, downloads, and Python version badges to README for better visibility and information access.
• Introduced error handling in evaluate_famous_models.py to gracefully manage failures when processing certain models.

🎯 Purpose & Impact

• Improved Visibility: The new badges on the README make it easier for users to find THOP on PyPI, understand its compatibility, and gauge its popularity.
• Robust Benchmarking: The addition of try-except blocks in benchmark scripts ensures the tool can handle and report errors more gracefully. This improvement helps developers identify and troubleshoot issues with specific models without crashing the entire benchmark process.

🚀 These changes make THOP more user-friendly and reliable for developers profiling their PyTorch models, potentially increasing its adoption and user satisfaction.