Yanbin Liu
Yanbin Liu
I am also considering this issue. I think you can refer to this original theano code. https://github.com/yoonkim/CNN_sentence/blob/master/conv_net_sentence.py#L227 It seems except for the first layer, all weights are clipped. But I...
It's great. I will keep an eye on this project.
@bertinetto @jakesnell I also reproduced the experiments using tensorflow with the same setting and parameters, still got similar results. I wonder if I miss some implementation or parameter details?
@dnlcrl I think the results above (97.4 | 99.3 | 95.4 | 98.8) cite the original prototypical paper of ICLR (https://openreview.net/references/pdf?id=BJ-3bnVmg), which is different setting with this git repo. The...
Hi, @yannlif I think validation cannot be combined into the training set for fair comparisons. Also, in this paper, the authors said: "We follow their procedure by training on the...
@yannlif I got it. I think you are right for Omniglot. Thanks for this information. I will try again.
> > 想问下有没有pytorch实现版本 > > 我在做了，如果顺利的话会发布的 请问，pytorch版本GMAN可否release或者能否推荐目前比较好的实现。 谢谢！
Our model is a general framework for ConvNet and ResNet. This repo aims to provide simple code for reproducing the results in our ICLR19 paper. For ResNet12, I think you...
In our paper, we monitor the best validation performance and choose the corresponding iteration to test the model. This iteration may differ according to different training processes with random seeds.
I think early stopping should be applied to balance.