yzh119 / bpt Goto Github PK

View Code? Open in Web Editor NEW

124.0 124.0 20.0 65 KB

Source code of paper "BP-Transformer: Modelling Long-Range Context via Binary Partitioning"

License: MIT License

Shell 0.50% Python 80.82% Cuda 14.95% C++ 3.72%

bpt's People

Stargazers

Watchers

bpt's Issues

How to use BPT with a custom dataset for a machine translation task?

Thanks for this great artefact. I am interested to apply this library for a machine translation task. However, in my case, I have to feed a custom dataset. Furthermore, I have to add customized nodes and edges as my task is slightly different.

Do you think it is possible with BPT? Would be great if I can discuss this with you in greater detail.

BPT vs BERT

你好

BP-Transformer这篇论文非常棒，我很喜欢这篇论文

关于Text Classification类的任务，
请问你们有否做过和BERT或者ALBERT之间的比较试验？

Publication/Peer Review?

Hi cool work!

Do you have any plans to publish it under peer-review? I was only able to find a preprint version.

Thanks!

I think there is a bug in the implementation of bpc

According to the material I have find from here and here, bpc=log2(NLL).
But in the implementation in your code, I found that bpc = NLL / log2.
Is there something wrong for the calculation of bpc, or I have missed anything?

Recommend Projects