Losswise/losswise-python

question about your blog about LSTM

tinymindxx opened this issue · 1 comments

Sorry for create an issue in this project and thanks for your great work on the blog about lstm(http://nicodjimenez.github.io/2014/08/08/lstm.html) . I am afraid that you have already deprecated the project of lstm (https://github.com/nicodjimenez/lstm) ,so i open an issue here.
Sorry for this again.

in your blog (http://nicodjimenez.github.io/2014/08/08/lstm.html), derivative of s(t) is get by this equation,
image
i wonder why the left of this equation is equal to the right.
in my opinion, left = 0.5*right. and why we need the top_diff_s if we could calculate
image directly.
image

I am new to nn and lstm, and i will appreciate if you could give some tips on this question, thanks!

Hi, thanks for reaching out! Let's continue this discussion on the LSTM blog.