Yichong Xu
Results
3
comments of
Yichong Xu
@ipod825 Have you tried your implementation using clamp and l_t.detach()? I tried that and got a very high performance on 6 glimpses, 8*8, 1 scale setting, around 0.58%. Paper reported...
Thanks for the response! However I only get accuracy 82.05% on SNLI with the setting in the config file. Have you run the experiment using that file?
We will release that part of code soon.