Closed greathope closed 4 years ago
ok, I found it in your paddlepaddle code, so I just close it.
Thanks @greathope . We have tried different training methods and then combined the trained model. In the paper, we mainly report the method to train the best single model (based on validation score).
I have read your paper and code roughly, but have few questions. Your code used SE-ResNeXt101, but ResNeXt101 in your paper; your code did not use Block3; your learning rate is 0.02 instead 0.01 in your paper; the lr scheduler is MultiStepLR instead Cosine; the feature dim is 4092 instead 1024; May I ask why?