I have a dataset where there was 5608 images and it downloaded the entire dataset with no issues. This version was from 1/24. I added about 400 more images and the version is from 1/25. The download of the dataset is missing a significant amount of images. Possibly a bug that was introduced in an update done on the back end? Any help on getting this resolved would be greatly appreciated.
Project ID: v3-body-afhwu
This is happening currently to me.
When I create a new version of my dataset (no new images were included, just changed annotations by a team mate) and I download it, only half of the contents are downloaded, although it is clearly seen in the web that it contains more images than that.
The new .zip compared to the older versions weights exactly half of its original size and it contains fewer images.
Same issue on my workspace, tried to duplicate project, rebalancing splits etc. This is so frustrating and credits consuming. The more I generate versions, the less I get in my folders/zip files
yes, this is happening to me as well
Iām glad to see Iām not the only one having this issue. It appears others have commented on a similar post but it was for feature request. Hoping that the roboflow team will be able to fix this tomorrow once they are back at work.