Update ClassificationModel default training imgsz=224

[glenn-jocher]

To match classify/val.py and classify/predict.py

Helps #9462

Signed-off-by: Glenn Jocher glenn.jocher@ultralytics.com

🛠️ PR Summary

_{Made with ❤️ by Ultralytics Actions}

🌟 Summary

YOLOv5 classification model now trains with higher resolution images by default.

📊 Key Changes

• The default input image size for training has been changed from 128 to 224 pixels.
• The example usage in the documentation has been updated to reflect this new default image size.

🎯 Purpose & Impact

• 🎨 Enhances the model's ability to learn from higher resolution images, which can potentially improve classification accuracy.
• 🧠 May impact computational requirements, as higher resolution images consume more memory and processing power.
• 📈 Users should see an improvement in the model performance on classification tasks with adequate hardware.