I have successfully annotated images in a dataset using Vertex AI(AutomL) and trained a model, achieving good accuracy in the development project. However, when I transferred the same dataset with annotations from the development project to the production project, the model is not predicting accurately.
I would appreciate any assistance in understanding why there is a difference in predictions between the two projects. The accuracy in the development project is satisfactory, but the model is not performing similarly in the production project. If anyone has insights or suggestions on potential reasons for this discrepancy and how to address it, I would be grateful for your help.
Thank you!
I need any help, solution or suggestion regarding the problem