YOLO11 validation without augmentation.

[NS-Nik]

Hi!

I have trained yolo11m.pt for my custom objects but I'm facing some issues during validation. I implement the aggressive augmentation settings during the model training. However, I don't want to use it for the calculation of the validation metrics. I tried to switch it off with the 'augment=False' setting. But the augmentation is still implemented.

metrics = model.val(augment=False, iou=0.25)

Any ideas how to turn of the augmentation for the validation?

[UltralyticsAssistant]

👋 Hello @NS-Nik, thank you for reaching out to the Ultralytics community 🚀! Your issue with disabling augmentation during validation is a valuable topic. We invite you to explore our Docs where you might find insight into validation settings. Specifically, our Python Usage guide could be of assistance for similar issues.

If this is a 🐛 Bug Report, please help us by providing a minimum reproducible example, which will assist in debugging the situation more efficiently.

Moreover, if you have custom training ❓ questions, please share detailed information, including dataset samples and training logs. Double-check to ensure you're following the guidelines in our Tips for Best Training Results.

To engage with the community, consider joining our Discord 🎧 for real-time interactions, or for deeper dives, tap into our Discourse or our Subreddit to fellowship with other members.

Upgrade

To ensure you’re running the latest version, upgrade the ultralytics package including all requirements in a Python>=3.8 environment with PyTorch>=1.8:

pip install -U ultralytics

Environments

Your YOLO model can operate in verified environments offering a range of dependencies:

• Notebooks with free GPU:
• Google Cloud via Deep Learning VM. Visit GCP Quickstart Guide
• Amazon via Deep Learning AMI. Visit AWS Quickstart Guide
• Docker Image: Check the Docker Quickstart Guide

Status

This badge displays the status of Ultralytics CI, which confirms the operation of all YOLO Modes and Tasks across multiple OS environments.

Don't worry, this is an automated response 🤖, but rest assured, an Ultralytics engineer will be here to assist you further soon!

[glenn-jocher]

@NS-Nik hi! To ensure no augmentation is applied during validation, confirm that the validation function is correctly configured without any augmentation settings in your YAML or script. Also, check that there are no residual augmentation parameters overriding your augment=False setting. If the issue persists, try updating to the latest version of the Ultralytics package, as there may have been fixes or improvements. For further assistance, feel free to review the YOLO Common Issues documentation.

[NS-Nik]

HI! @glenn-jocher
I created the .yaml file where I defined paths to my validation images and the passed it to the "data" argument and it worked!

Thank you for the help!

[glenn-jocher]

@NS-Nik i'm glad to hear that it worked for you! If you have any more questions or need further assistance, feel free to reach out. Kudos to the YOLO community and the Ultralytics team for their support!

[Skillnoob]

Augmentation isn't used during validation and the augment parameter is only for prediction, to answer your other points.