search

jfzhang95 / pytorch-video-recognition

PyTorch implemented C3D, R3D, R2Plus1D models for video activity recognition.
MIT License
1.16k stars 250 forks source link

nn.CrossEntropyLoss should work with logit scores #7

Closed coderSkyChen closed 5 years ago

coderSkyChen commented 5 years ago

if you use nn.CrossEntropyLoss the input should be the scores before softmax layer. So you should remove the softmax layer.

jfzhang95 commented 5 years ago

Hi, Three models implemented in this repo did not contain softmax layer.