Wangchunshu Zhou
Wangchunshu Zhou
I've followed the hyper parameters presented in the paper but only got accuracy of ~83.0. Have you figured out how to get the reported result? Thanks!
I think it is one batch, which is a common practice in GANs.
interesting. Did you figured out why?
Hi @Harry-zzh First, could you please share on which dataset you conduct your experiments? If it is some small datasets, 3% variation may indeed come from different random seeds. Otherwise,...
@Harry-zzh Can you share the exact command for your best result on the task? Also, can you share the results on the dev set of the GLUE benchmark? You can...
Hi, that was my mistake. The teacher is initialized by pretrained BERT (well-read student). But using fine-tuned teacher should be able to achieve similar performance. First I think you should...
For further questions, maybe you can send me an email with your wechat ID to [email protected] so that I can offer further guidance and help more promptly and conveniently.
> Hi @KaiDMML! Thank you for making FakeNewsNet avalable. > > I'm trying to apply the dataset using only the news content, but I'm finding some problems when reproducing the...
I also found this problem, when the test data and reference data is the same, self-bleu is always 1. However many papers in this domain use it as a diversity...
Hey I'm not getting "too many requests in one hour" message even with many access tokens and my own proxy server. I think it's an IP-based rate limit. But it...