is there a max limit to --imgsz ?

[mamdouhhz]

Search before asking

• I have searched the YOLOv5 issues and discussions and found no similar questions.

Question

!python train.py --imgsz 2880 --batch 16 --epochs 400 --data /content/drive/MyDrive/DATASET/data.yaml --weights yolov5m.pt --device 0

training does not start when i set imgsz to 2880, but starts when the imgsz value is 640 (the default), why does this happen?
my dataset is dental x-ray images with dimensions around 2953 × 1316 from the dentex grand challenges.

Additional

can it be memory limitations ?

No response

[github-actions]

👋 Hello @mamdouhhz, thank you for your interest in YOLOv5 🚀! Please visit our ⭐️ Tutorials to get started, where you can find quickstart guides for simple tasks like Custom Data Training all the way to advanced concepts like Hyperparameter Evolution.

If this is a 🐛 Bug Report, please provide a minimum reproducible example to help us debug it.

If this is a custom training ❓ Question, please provide as much information as possible, including dataset image examples and training logs, and verify you are following our Tips for Best Training Results.

Requirements

Python>=3.8.0 with all requirements.txt installed including PyTorch>=1.8. To get started:

git clone https://github.com/ultralytics/yolov5  # clone
cd yolov5
pip install -r requirements.txt  # install

Environments

YOLOv5 may be run in any of the following up-to-date verified environments (with all dependencies including CUDA/CUDNN, Python and PyTorch preinstalled):

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

Status

If this badge is green, all YOLOv5 GitHub Actions Continuous Integration (CI) tests are currently passing. CI tests verify correct operation of YOLOv5 training, validation, inference, export and benchmarks on macOS, Windows, and Ubuntu every 24 hours and on every commit.

Introducing YOLOv8 🚀

We're excited to announce the launch of our latest state-of-the-art (SOTA) object detection model for 2023 - YOLOv8 🚀!

Designed to be fast, accurate, and easy to use, YOLOv8 is an ideal choice for a wide range of object detection, image segmentation and image classification tasks. With YOLOv8, you'll be able to quickly and accurately detect objects in real-time, streamline your workflows, and achieve new levels of accuracy in your projects.

Check out our YOLOv8 Docs for details and get started with:

pip install ultralytics

[glenn-jocher]

Hello there! 👋

Yes, you're onto something with memory limitations. The --imgsz parameter in YOLOv5 specifies the image size to resize your images to during training. When you set a very large --imgsz, such as 2880, it significantly increases GPU memory requirements. Your training likely doesn't start with --imgsz 2880 due to insufficient GPU memory to handle such large images.

A workaround is to use a smaller --imgsz that fits within your GPU's memory limits. You can experiment starting with lower sizes and gradually increasing until you find the maximum size that works for your setup. Additionally, reducing the batch size can help accommodate larger image sizes, as it also reduces memory consumption.

Feel free to consult the docs for more insights on managing resource usage during training.

Let us know if you have any more questions. Happy training!

[mamdouhhz]

thank you for your response, i have another question

the dimensions of my dataset's images are not the same they range from 1976 x 976 up to 3076 x 1536 and all images are labelled, if i resized the images so that all images are the same size and to be smaller in size then the labels will not work on the new image dimensions, how to solve this problem ? the labels are bbox coordinates and diesease class

[glenn-jocher]

Hello again! 😊

No worries, YOLOv5 handles varying image sizes and label rescaling automatically. When you specify --imgsz, YOLOv5 resizes your images to that size for training while appropriately scaling the bounding box coordinates in your labels, ensuring they still accurately represent the object locations in the resized images.

Just proceed with training by setting your desired --imgsz parameter, and YOLOv5 will take care of the rest. No manual resizing or label adjustment needed on your part!

Remember, though, to choose an --imgsz that balances between your GPU memory limitations and the need to maintain detail for accurate detection, especially for high-resolution datasets like yours.

Happy to help if you have more questions. Keep up the great work!

[mamdouhhz]

Hello again! 😊

No worries, YOLOv5 handles varying image sizes and label rescaling automatically. When you specify --imgsz, YOLOv5 resizes your images to that size for training while appropriately scaling the bounding box coordinates in your labels, ensuring they still accurately represent the object locations in the resized images.

Just proceed with training by setting your desired --imgsz parameter, and YOLOv5 will take care of the rest. No manual resizing or label adjustment needed on your part!

Remember, though, to choose an --imgsz that balances between your GPU memory limitations and the need to maintain detail for accurate detection, especially for high-resolution datasets like yours.

Happy to help if you have more questions. Keep up the great work!

Thank you very much

[glenn-jocher]

You're welcome! 😄 If you have any more questions or need further assistance as you proceed, feel free to reach out. Happy training!

[github-actions]

👋 Hello there! We wanted to give you a friendly reminder that this issue has not had any recent activity and may be closed soon, but don't worry - you can always reopen it if needed. If you still have any questions or concerns, please feel free to let us know how we can help.

For additional resources and information, please see the links below:

• Docs: https://docs.ultralytics.com
• HUB: https://hub.ultralytics.com
• Community: https://community.ultralytics.com

Feel free to inform us of any other issues you discover or feature requests that come to mind in the future. Pull Requests (PRs) are also always welcomed!

Thank you for your contributions to YOLO 🚀 and Vision AI ⭐