Error in W&B Callback about ultralytics

[github-actions]

👋 Hello @Pikus16, thank you for your interest in Ultralytics YOLOv8 🚀! We recommend a visit to the Docs for new users where you can find many Python and CLI usage examples and where many of the most common questions may already be answered.

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.

Join the vibrant Ultralytics Discord 🎧 community for real-time conversations and collaborations. This platform offers a perfect space to inquire, showcase your work, and connect with fellow Ultralytics users.

Install

Pip install the ultralytics package including all requirements in a Python>=3.8 environment with PyTorch>=1.8.

pip install ultralytics

Environments

YOLOv8 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 Ultralytics CI tests are currently passing. CI tests verify correct operation of all YOLOv8 Modes and Tasks on macOS, Windows, and Ubuntu every 24 hours and on every commit.

[glenn-jocher]

It seems like the error might be occurring due to concurrent modification of the plots dictionary while it's being iterated over in the W&B callback. This can happen if the training process or another part of the code modifies the plots dictionary during logging to W&B.

A potential workaround could be to capture a shallow copy of plots.items() before the loop starts in your callback function. Here's a quick modification you can try:

def _log_plots(plots, step):
    items = list(plots.items())  # take a snapshot of items
    for name, params in items:
        # your logging code here

This change ensures that you're iterating over a static list of items, and any modifications to the original plots dictionary during iteration won't affect the loop. Although it feels a bit hacky, it is a common practice to avoid such runtime errors and should resolve the issue. Give this a try and see if it stabilizes your training logging!

[Pikus16]

Makes sense, thank you for your response! I submitted a PR with your change in it here: #12779

[glenn-jocher]

@Pikus16 PR merged, thank you for your contributions!

[Pikus16]

Thank you for all your help!

[glenn-jocher]

@Pikus16 you're welcome! If you have any more questions or need further assistance down the road, feel free to reach out. Happy coding! 😊