-
Hi，Thanks for your code, it's really useful for most nlp researchers and thank you again.
And when I run this code, it's often interrupted by network error after download a little files, I thought …
-
Hi, thank you very much for the implementation!
I'm trying to compare your implementation with the official TF BERT head-to-head with the Gutenberg dataset (since the BookCorpus dataset is no longe…
-
Hi Sanyam,
Are there already some saved models based on the BookCorpus we can use with your code?
Thank you.
meoti updated
6 years ago
-
Hi, can you please give more information about the pretrained corpus that you used. The DPP paper only mentions Huge corpus, but he is not specific.
Moreover, do you have results without the pre-tra…
-
I'm trying to train a custom decoder. I am using the bookcorpus data with science fiction novels. According to the docs:
https://github.com/ryankiros/skip-thoughts/blob/master/decoding/README.md
"We…