search

TengdaHan / CoCLR

[NeurIPS'20] Self-supervised Co-Training for Video Representation Learning. Tengda Han, Weidi Xie, Andrew Zisserman.
Apache License 2.0
286 stars 32 forks source link

config for linear probe #19

Closed wbxzky closed 3 years ago

wbxzky commented 3 years ago

Hi Tengda, Thanks for sharing the code! I have a question about your linear probe setting. I use your released UCF101 pretrained rgb model to train a fc layer. To do that I run the command with your code's default setting: python3 main_classifier.py --train_what last --pretrain /../CoCLR-ucf101-rgb-128-s3d-ep182.tar

After that I got a new pretrained model and I reload it to evaluate the s3d backbone: python3 main_classifier.py --test /../epoch9.pth.tar --ten_crop

Finally I got Acc@1 67.9 on UCF-101 test split 1, which is lower than your report result 70.2. So I want to ask is there something wrong with my training and testing processing? could you please share me your config and steps to train a fc layer for linear probe? Thanks, looking forward to your reply.

junmin98 commented 3 years ago

Hello. I'm trying to do exactly what you asked i.e. python3 main_classifier.py --test {pretrained_model_path} --ten_crop ... I got the following error. Could you please tell me how I solved it?

File "main_classifier.py", line 490, in test_10crop vname = vname[0] UnboundLocalError: local variable 'vname' referenced before assignment

junmin98 commented 3 years ago

Hello. I'm trying to do exactly what you asked i.e. python3 main_classifier.py --test {pretrained_model_path} --ten_crop ... I got the following error. Could you please tell me how I solved it?

File "main_classifier.py", line 490, in test_10crop vname = vname[0] UnboundLocalError: local variable 'vname' referenced before assignment

I solved this problem by referring to the MemDPC code.

TengdaHan commented 3 years ago

I upload detailed instructions in readme now. Sorry for the delay. @wbxzky I think reducing lr by step scheduler is important. The 9-th epoch you evaluated may not be well-trained.