Detect.inplace=False for multithread-safe inference

[glenn-jocher]

Detect.inplace=False for safe multithread inference when loading YOLOv5 AutoShape models with PyTorch Hub.

May resolve #8565

🛠️ PR Summary

_{Made with ❤️ by Ultralytics Actions}

🌟 Summary

Improved thread safety for multi-threaded inference in the YOLOv5 model.

📊 Key Changes

• Set Detect.inplace attribute to False during model creation for safer multithreading.
• Tidy-up in the yolo.py comments without changing functionality (changed 'in-place' to 'inplace').

🎯 Purpose & Impact

• 🧵 The modification ensures thread safety, reducing the risk of issues when the YOLOv5 model is used in environments with multiple threads making inferences simultaneously.
• 🚀 Potential impact includes more robust performance in multi-threaded applications, improving the model's reliability and usability in production environments.