Zhengchun Liu

I also have the same problem with 1.12, it is extremely slow for the 1st epoch with Conv, no cnn (only lstm) is fine

@DmitryUlyanov how do you explain the mask in the loss function? it seems intuitive to me that the mask are 1s for missing pixels because we want to predict those...

> > @abhishekaich27 how do you explain the mask in the loss function? > > it seems intuitive to me that the mask are 1s for missing pixels because we...

> **like this format! `summary(net, [(1, 128, 128), (1, 128, 128)])`** > > If it doesn't work, plz show more code and error description. it works! Thanks

@dianarg @slabasan Thanks very much for your response! sorry for my silence. We want to do some evaluations using MSR (control energy consumption) but the resource provider does not want...