Conv1D implemented in Pytorch doesn't work !

thtran97 / Deep-Learning-for-Audio

Audio Classification & Sound Event Detection in Pytorch

5 stars 0 forks source link

Closed thtran97 closed 4 years ago

thtran97 commented 4 years ago

Loss doesnt seem decreasing every epoch.... Accuracy are always about 2-3% !!!

thtran97 commented 4 years ago

Softmax layer is not necessary at the final layer of model when using CELoss Criterion

thtran97 commented 4 years ago

Pay attention to put softmax layer when predicting the result in order to compute accuaracy