I have a dataset of 540 images that I used to train some models (with YOLOv11) and auto label unnannoted images as well, but from the previous month to this month the Auto Label stoped working. When I upload new images and click on the “Auto-Label Entire Batch” and select my model, the message “There was an error running your Automatic Labeling test. Please try again.” poops up. I already trainned a new model, refreshed the page and uploaded other images but the message continues to pop up.
Describe your question/issue here! (delete this when you post)
- Project Type: Instance Segmentation
- Operating System & Browser: Google Chrome
- Project Universe Link or Workspace/Project ID: Fixed quadrats PT 2
- Do you grant Roboflow Support permission to access your Workspace for troubleshooting? (Yes/No): Yes
Hi @André_Anjo ,
I looked into your workspace, seems to work for me if you have confidence above 20%. At 10% there are likely way too many detections and it doesn’t work. Could you try that?
Thanks, Erik
Hello, I have always put the confidence threshold at 50%, but even when I put 20% or any other % it still doens’t work.
Thanks, André
Hi @André_Anjo ,
I see, and to confirm, you’re using fixed-quadrats-pt-2-qjt6o/64 for the model? As the popup is in the way.
Thanks, Erik
Yes, I was using fixed-quadrats-pt-2-qjt6o/64. But I tried other models as well and none worked. I even went to the Model tab to see each model and a pop up message appeared saying “Oops something went wrong, there might be issues with this model” in all I selected.
This problem only occured after I uploaded annoted images to Roboflow, where the classes switched colour. I put the original colour back on the Class tab but when I try to Auto label the class colours are different from the ones I selected on the Class tab.
I even duplicated the project, and the same problem happened on the duplicated project.
Thanks, André
