Add Dockerfile descriptions

[glenn-jocher]

🛠️ PR Summary

_{Made with ❤️ by Ultralytics Actions}

🌟 Summary

Enhancements to Ultralytics YOLOv5 Docker images for better clarity and optimization.

📊 Key Changes

• Updated Dockerfile descriptions for clarity on the purpose and optimization of each image:
  • Dockerfile: Now specifies that the image is CUDA-optimized for YOLOv5 training and inference using single/multi-GPU.
  • Dockerfile-arm64: Indicates compatibility with ARM architectures, such as Apple M1, Jetson Nano, and Raspberry Pi.
  • Dockerfile-cpu: Points out that the image is optimized for CPU deployments, including ONNX, OpenVINO, and PyTorch.

🎯 Purpose & Impact

• These changes provide clearer documentation and guidance for users looking to utilize Docker images tailored to their hardware setup.
• They help ensure that users can select the most appropriate Docker image for their specific needs, potentially improving their experience with YOLOv5 applications.
• By optimizing images for specific architectures and use cases, users can expect better performance and efficiency, whether they are using GPUs, ARM-based CPUs, or traditional CPUs. 🚀