Update default.yaml to include the save_dir parameter.

[jbdebelle]

The goal of your commit is to address an issue encountered during validation, specifically related to the save_dir argument not utilizing the value passed to the function as intended. Instead, it utilizes an incremented path, leading to unexpected behavior.

This issue stems from the get_cfg function, where the condition if "save_dir" not in cfg: sets save_dir to None if it's not already defined in the configuration. As a result, the intended value of save_dir passed to the function is not used, leading to the undesired behavior during validation.

🛠️ PR Summary

_{Made with ❤️ by Ultralytics Actions}

🌟 Summary

Introducing a customizable save directory for results in Ultralytics configurations! 🚀

📊 Key Changes

• Added a new configuration option save_dir to specify the directory where results should be saved.

🎯 Purpose & Impact

• Flexibility: Users can now specify a custom directory to save their results, enhancing flexibility and organization of output data. 📁
• Improved Usability: Makes it easier for users to manage and locate their output files, streamlining workflows and analysis. 🛠️
• Potential Impact: This change could significantly improve user experience by providing more control over where results are stored, potentially leading to better data management and analysis practices. 🌈

[github-actions]

CLA Assistant Lite bot All Contributors have signed the CLA. ✅

[github-actions]

👋 Hello @jbdebelle, thank you for submitting an Ultralytics YOLOv8 🚀 PR! To allow your work to be integrated as seamlessly as possible, we advise you to:

• ✅ Verify your PR is up-to-date with ultralytics/ultralytics main branch. If your PR is behind you can update your code by clicking the 'Update branch' button or by running git pull and git merge main locally.
• ✅ Verify all YOLOv8 Continuous Integration (CI) checks are passing.
• ✅ Update YOLOv8 Docs for any new or updated features.
• ✅ Reduce changes to the absolute minimum required for your bug fix or feature addition. "It is not daily increase but daily decrease, hack away the unessential. The closer to the source, the less wastage there is." — Bruce Lee

See our Contributing Guide for details and let us know if you have any questions!

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 69.93%. Comparing base (fcfc44e) to head (4cd2df9).

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #10153      +/-   ##
==========================================
- Coverage   73.46%   69.93%   -3.53%     
==========================================
  Files         122      122              
  Lines       15622    15622              
==========================================
- Hits        11476    10926     -550     
- Misses       4146     4696     +550     

Flag	Coverage Δ
Benchmarks	35.35% <ø> (-0.20%)	⬇️
GPU	37.28% <ø> (-0.02%)	⬇️
Tests	66.13% <ø> (-3.86%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[jbdebelle]

I have read the CLA Document and I sign the CLA

[jbdebelle]

recheck

[glenn-jocher]

@jbdebelle certainly! 😊 It looks like you're requesting a re-evaluation of a particular aspect or issue. To assist you effectively, could you please provide more specific details about what you need rechecked? If it's pertaining to a potential bug or an update in code or documentation, a brief code snippet or direct link to the concerned part would be immensely helpful!

[jbdebelle]

args = dict( model=model, imgsz=size, max_det=max_det, data=data, split=split, conf=conf, iou=iou, task=task, save_dir=save_dir, ) return YOLO(model).val(**args)

In the code above is I specify a custom save_dir path, it will not be taken into account. 
By adding save_dir in the default.yaml file, this mitigates this issue. 

[glenn-jocher]

Thanks for pointing this out! 😊

Indeed, including save_dir in your default.yaml and passing it through your code as shown, should direct your results to the specified directory. Just ensure the save_dir is defined in your default.yaml like so:

save_dir: 'path/to/your/directory'

And the snippet you've provided should work as intended. If it doesn't, making sure that default.yaml is correctly loaded before running your validation will be key.

Keep us posted if you encounter any other issues! 🛠️

[jbdebelle]

Then shouldn't my PR be merged into the main? This means that the current implementation doesn't support the use of the save_dir argument. My suggested fix will resolve this issue.

[glenn-jocher]

Thanks for bringing this to our attention! 🌟 You're right; your PR addresses the save_dir argument not being effectively used as intended. This is an important fix, and we appreciate your effort to improve the functionality. We'll review the proposed changes closely and proceed with merging if everything aligns well with our project goals and standards. Please bear with us as we go through this process.

Keep contributing! Your input helps make our project better for everyone. 👍

[n1mmy]

+1. I encountered the same issue of save_dir not working. Modifying default.yaml worked for me to fix the issue. Thanks for this!

[glenn-jocher]

Hi there! 🎉 I'm glad to hear that modifying default.yaml resolved the save_dir issue for you. Thanks for sharing your successful workaround—it's really helpful for the community. If anyone else encounters this, they can apply the same fix by adding:

save_dir: 'desired/path/for/output'

to their default.yaml. Happy coding and keep the feedback coming! 👍

[luca-depe]

Hi @glenn-jocher! Will this be merged soon?

[luca-depe]

Nice catch!

[glenn-jocher]

@luca-depe hello!

Thanks for reaching out. We're currently reviewing the changes to ensure everything aligns perfectly with our standards. If all goes well, it should be merged shortly. We appreciate your patience and enthusiasm! 🚀

Stay tuned!