skip-gram-pytorch
skip-gram-pytorch copied to clipboard
Loss doesnt decrease?
Why do i feel that the loss isnt decreasing? It started at 13 and then jumped to 8 and now i see random losses like 7 or 6 or 5 or 8 again... is that normal?
I think in file word2vec.py line 37 it should NOT be “optim.SGD(model.parameters(),lr=0.2)” but "SGA"(However, pytorch have not SGA) So, the solution is flipping the sign of objective function.