Marcin Elantkowski

Taking ideas from [Deep Learning for NLP](http://ruder.io/deep-learning-nlp-best-practices/index.html#wordembeddings), things I believe would be nice and easy to try out on `LSTM` model: - [x] self-attention on the sequence of items -...

Oh, yes, sorry. I'll submit the PR ASAP.

Looks cool, thank you! Perhaps to be extra safe we could add some check with a nice error message?

I'm sorry, I didn't intend this to sound offensive. It *seemed* broken, since doing something quite intuitive resulted in an obscure pytorch error, but I agree I'm at fault here....

Thanks a lot for the explanation! I'll be happy to open a PR adding a small section with `pmap`

Apologies, I forgot about this issue, I'll get to it this week and open a PR once #243 is merged.

Hello @KevinMusgrave , which of the issues tagged with `new algorithm request` do you think would be the most beneficial to have implemented? I'm looking to contribute to some open...