change LBFGS's default tolerance_grad to 1e-7 #25240
Closed
Conversation
This file contains 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
There was a problem hiding this comment.
@vincentqb has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.
|
@vincentqb merged this pull request in 1b7f7aa. |
pytorchmergebot
pushed a commit
that referenced
this pull request
Apr 22, 2023
LBFGS' `tolerance_grad` parameter has had a default value of `1e-7` since #25240. The doc string wasn't updated in that PR to match the change https://github.com/pytorch/pytorch/blob/main/torch/optim/lbfgs.py#L207. no open issue for it, just happened to set it to 1e-7 and was surprised my results didn't change :-) eventually noticed inconsistency in the doc and seemed like an easy opportunity to figure out how to contribute. Pull Request resolved: #99792 Approved by: https://github.com/janeyx99
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
Hi,
I noticed after v1.2.0 the implement of LBFGS optimizer has been changed. In this new implement, the return condition has been changed from the sum of the gradients to the max value in the gradients (see:
pytorch/torch/optim/lbfgs.py
Line 313 in b15d914
So, I came up this pull request to suggest that changing this tolerance_grad to a smaller value