Skip to content
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.

Sign up for GitHub

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

Jump to bottom

Adding weight initialization options #223

Merged
merged 6 commits into from
Nov 3, 2018
Merged

Adding weight initialization options #223

merged 6 commits into from
Nov 3, 2018

Conversation

mwcvitkovic
Copy link
Contributor

Feature / Fix

  • Giving user the ability to specify a torch.nn.init initialization function for their network in the net section of spec files.

kengz
kengz approved these changes Nov 3, 2018
View reviewed changes
Copy link
Owner

@kengz kengz left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

beautiful, this brings proper weight init functionality to the lab!
will merge to staging first, then:

  • some minor stylefix (name fxn to fn per conventions in the lab)
  • apply the new init_layers function to all networks

slm_lab/agent/net/net_util.py Show resolved Hide resolved
@kengz kengz changed the base branch from master to staging November 3, 2018 05:34
@kengz kengz merged commit fa7abbe into kengz:staging Nov 3, 2018
@kengz kengz mentioned this pull request Nov 3, 2018
Merged
kengz added a commit that referenced this pull request Nov 3, 2018
@kengz
Add weight initialization to master (#225)
73b777a 
* Adding weight initialization options (#223)

* Adding vizdoom functionality (hat tip @joelouismarino)

* Typos

* Typo/bug

* Adding weight initialization options in Net specification

* Code cleanup

* extend init_fxn to other mlp

* extend recurrent layer init fxn comment

* extend init_fxn doc in convnet

* rename fxn to fn
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@kengz kengz kengz approved these changes

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
2 participants
@mwcvitkovic @kengz