search

zhmiao / OpenLongTailRecognition-OLTR

Pytorch implementation for "Large-Scale Long-Tailed Recognition in an Open World" (CVPR 2019 ORAL)
BSD 3-Clause "New" or "Revised" License
839 stars 128 forks source link

the accuracy of the train and val #61

Closed cl475143764 closed 3 years ago

cl475143764 commented 4 years ago

thanks for such a beautiful work. but i have a question about the acc between the train and val. in the train phase, the network could converge finally and got a high acc on the train_set, but in the test phase, when i used the test-set, the network's acc became very low, so i tried to use the train data to verify the weights is right, but the result was that it was also so worse, i tried to check what's wrong, but i had no idea about it, could you please give me a advice? thanks a lot!

zhmiao commented 4 years ago

Hello @cl475143764, thank you very much for asking and sorry for the late reply! I think this might because the model is not properly saved. Did you change anything in the original code?

zhmiao commented 3 years ago

As it has been over three months, I will close this issue for now. If you have any more questions, you are welcome to reopen it. Thanks.