If we execute ivadomed training on Torch 1.9+ we run into this particular import error, mainly because most existing code bases are only tested with Torch 1.8.X. A potential solution has been this as ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results