-
Notifications
You must be signed in to change notification settings - Fork 250
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Incorporate gru-nonlin #15
Conversation
@@ -250,6 +250,39 @@ def param_init_gru(options, params, prefix='gru', nin=None, dim=None): | |||
return params | |||
|
|||
|
|||
def param_init_gru_nonlin(options, params, prefix='gru', |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
How about defining U_nl
, b_nl
, Ux_nl
and bx_nl
in param_init_gru_cond
after calling param_init_gru
, instead of making a new function param_init_gru_nonlin
?
@jakezhaojb once the one comment I made is resolved, I'll merge it right away. |
Yes! I just did. Please review it. |
nin=None, dim=None, dimctx=None): | ||
nin=None, dim=None, dimctx=None, | ||
nin_nonlin=None, dim_nonlin=None, | ||
hiero_nonlin=False): |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
hiero_nonlin
is not used in this code. Please remove any related entries.
Yeah done. |
Thanks, @jakezhaojb |
This is basically an incorporation of the `nonlin' trick that introduced from [here].(https://github.com/laulysta/nmt). Please review before merging it into master.