You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
On using resume feature of mxrcnn to start training my model again after some number of epochs, it produced really bad inferences. Whereas when I trained the whole model in one go, the results we satisfactory. I think there is some issue in the resume feature of mxrcnn pipeline.
The text was updated successfully, but these errors were encountered:
Thank you for pointing out the issue, model might not be loading correct weights or checkpoint information while resuming. We will try to resolve the error soon
On using resume feature of mxrcnn to start training my model again after some number of epochs, it produced really bad inferences. Whereas when I trained the whole model in one go, the results we satisfactory. I think there is some issue in the resume feature of mxrcnn pipeline.
The text was updated successfully, but these errors were encountered: