Enh/optimizer #6
Open
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
mschmidt87
reviewed
Jul 31, 2018
Contributor
mschmidt87
left a comment
mschmidt87
left a comment
There was a problem hiding this comment.
Very nice. Would be nice if torch.autograd could automatically compute natural gradients.
| optimizer_mu.step() | ||
|
|
||
| # manually update sigma | ||
| sigma *= np.exp(learning_rate_sigma / 2. * np.dot(utility, s ** 2 - 1)) |
Contributor
There was a problem hiding this comment.
Would it make sense here to update log(sigma) with the pytorch-optimizer? Or is that a bad idea due to numerical stability issues?
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This PR introduces pytorch optimizers to snes to update mu of the search distribution. instead of doing plain stochastic gradient descent, this opens the possibility to use any of the optimizers available in pytorch (https://pytorch.org/docs/stable/optim.html)
@mschmidt87 please have a look and let me know your thoughts on this