Local dataset quality check
Validate a YOLO dataset before training
YOLO exports can contain malformed boxes, inconsistent classes, missing images, or unexpected annotation types.
What to check
Check a dataset ZIP to confirm its detected YOLO subtype, class coverage, box geometry, and image inventory.
Why it matters
Invalid coordinates and missing media can waste training runs or quietly distort evaluation.
When Pro helps
Pro raises the file limit to 1 GB and adds saved repair sessions.
What your report can look like
YOLO detection export
An illustrative example of the findings and repair guidance Labelyst produces for your dataset.
Status: Ready with warnings
Score: 84/100
2 findings
Findings
- 2 boxes extend outside the image boundary
- 1 class name differs only by case
Passed check
- No exact cross-split duplicates found
Suggested next step: Review the two clipped boxes before training.
Rules: standard profile · local check
Illustrative local result. It is not a certification.
Check it before training
No account or upload is required for a Free check. Analysis, repairs, and reports stay on your device.