Kensho Hara

How do you evaluate the accuracy? If you get the accuracy in val.log, the accuracy is the clip-level accuracy whereas the accuracy reported in my paper is the video-level accuracy,...

Could you tell me how do you execute when the error occurs.

What did you mean in the question?

Which model do you use? Do you use ResNet-34 with shortcut type A? If you use ResNet-34 with shortcut type B, please specify --resnet_shortcut B.

It seems to be a problem of your resnet-34-kinetics-cpu.pth. Please download the file and try it again.

Using --pretrain_path initializes the last fc layer. So, the training of the last fc may need more epochs. If you want to use my pretrained model, you should just use...

Could you tell me your settings of options in the execution?

@CCV-Edward I'm glad the problem solved! @xiadingZ As mentioned by @CCV-Edward , you should set --resnet_shortcut B to use the models except for resnet-18 and -34.

Related to [this issue](https://github.com/kenshohara/3D-ResNets-PyTorch/issues/32), ```resnet-34-kinetics-cpu.pth``` includes some problems. Please download the latest file and try it again.

When did you download the file? I update it 7 days ago. If you tried the latest file, my code or model still include some bugs.