Fixup AccuracyLayer like SoftmaxLossLayer in #1970 #2076
Merged
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Outdated suggestions cannot be applied.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.
There has been some understandable confusion (e.g. #2063) over my changes to
AccuracyLayer
in #1970 -- I'm not sure what I was thinking when I added the confusingCHECK_LE
-- I don't think it broke any unit tests or examples, but it may have broken something for someone and probably confused a lot of people. Sorry about that. Regardless, this makesAccuracyLayer
with ND blobs behave likeSoftmaxLossLayer
, including adding anaxis
param to allow for labels with "spatial" axes. (@BlGene's changes in #1942 to addignore_label
are orthogonal and still deserve a rebase.)Edit: oops, it looks like @BlGene actually did extend this to spatial axes in #1942; sorry @BlGene -- making AccuracyLayer treat its axes just like SoftmaxLossLayer seemed like the clearest path to quickly fixing the confusion I'd created with my previous changes.