search

kjunelee / MetaOptNet

Meta-Learning with Differentiable Convex Optimization (CVPR 2019 Oral)
Apache License 2.0
517 stars 95 forks source link

val and test set in FC100 #15

Closed AceChuse closed 4 years ago

AceChuse commented 4 years ago

Thanks for your code! I think it is really great work. But I have a problem about FC100. During the training, the training loss is not problem， gradually decrease as the number of epoch increases. However, accuracy on val and test set remains constant around 50%. Here is the version of my package Python 3.6.3 torch 1.0.1.post2 torchnet 0.0.4 qpth 0.0.13

kjunelee commented 4 years ago

Hello. Thanks for your interest in our work.

I think your result seems okay as the performance of our system for FC100 is around 40 % for 1-shot and 55 % for 5-shot as shown in Table 2 of our paper. As a note, you will observe a jump in performance at epoch 21 where we reduce the learning rate to 1/10.

AceChuse commented 4 years ago

Thank you for your reply! I see what your mean. I will try something else.。