Is there a better way to calculate loss for multi-task DNN modeling?

Question

Suppose there are over one thousand tasks in the multi-task deep learning. More than a thousand columns of labels. Each task (column) has a specific weight in this case. It