Fix single-image dataset failure mode

[PacificDou]

Bugfix for https://github.com/ultralytics/ultralytics/issues/11693

🛠️ PR Summary

_{Made with ❤️ by Ultralytics Actions}

🌟 Summary

Enhancements in RTDETR functionality and data buffering for improved training and prediction.

📊 Key Changes

• RTDETR Test Update: The command for testing RTDETR functionality now includes an additional argument (fraction=0.25) to enable single-image training by processing only a fraction of the dataset.
• Data Buffer Modification: Updated the condition to prevent the data buffer from being empty before removal actions are initiated, ensuring at least one item remains for processing efficiency.

🎯 Purpose & Impact

• Enhanced Training Efficiency: By allowing single-image training through a fraction of the dataset, developers can test and iterate faster, saving time and computational resources. 🚀
• Improved Data Handling: The data buffer update ensures smoother processing by maintaining an optimal number of images in the buffer, enhancing overall performance and stability. 📈

These changes aim to streamline development workflows and improve the system's reliability for users, potentially contributing to better model performance and user satisfaction.

[PacificDou]

I have read the CLA Document and I sign the CLA

[codecov[bot]]

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Project coverage is 68.76%. Comparing base (b6415e4) to head (900f7a9).

Additional details and impacted files

```diff @@ Coverage Diff @@ ## main #11694 +/- ## ========================================== - Coverage 70.65% 68.76% -1.90% ========================================== Files 122 122 Lines 15591 15591 ========================================== - Hits 11016 10721 -295 - Misses 4575 4870 +295 ``` | [Flag](https://app.codecov.io/gh/ultralytics/ultralytics/pull/11694/flags?src=pr&el=flags&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=ultralytics) | Coverage Δ | | |---|---|---| | [Benchmarks](https://app.codecov.io/gh/ultralytics/ultralytics/pull/11694/flags?src=pr&el=flag&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=ultralytics) | `?` | | | [GPU](https://app.codecov.io/gh/ultralytics/ultralytics/pull/11694/flags?src=pr&el=flag&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=ultralytics) | `37.38% <0.00%> (ø)` | | | [Tests](https://app.codecov.io/gh/ultralytics/ultralytics/pull/11694/flags?src=pr&el=flag&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=ultralytics) | `66.73% <100.00%> (-0.01%)` | :arrow_down: | Flags with carried forward coverage won't be shown. [Click here](https://docs.codecov.io/docs/carryforward-flags?utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=ultralytics#carryforward-flags-in-the-pull-request-comment) to find out more.

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

[glenn-jocher]

@PacificDou thanks for the PR! I see you are right, this throws an error as COCO8 has only 4 train images:

yolo train data=coco8.yaml fraction=0.25

[glenn-jocher]

@PacificDou ok I've tested your fix (it works!), simplified it a bit and added a test for this failure mode. Running CI now and will merge after.

[PacificDou]

@glenn-jocher Wonderful! Thanks a lot for the prompt reply!

[glenn-jocher]

@PacificDou PR merged! Thank you for your contributions :)