Drop error if a Caffe layer references an invalid bottom #576
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
| @@ -1490,3 +1501,38 @@ def get_network_desc(self): | |||
| return text description of model | |||
| """ | |||
| return text_format.MessageToString(self.network) | |||
|
|
|||
| def net_sanity_check(self, net, phase): | |||
There was a problem hiding this comment.
Can we make this a classmethod so it's easy to write tests for?
There was a problem hiding this comment.
Sure, or perhaps a staticmethod since we're not making any reference to the class?
And I need to write tests too, thanks for reminding me :-)
There was a problem hiding this comment.
Yeah you're right - I meant staticmethod.
This should help address the case when a layer is added to manipulate e.g. the label blob. In that case the layer must include a directive like:
include {
phase: TRAIN
}
Otherwise Caffe will drop an error during inference.
A sanity check is added to DIGITS to catch these errors earlier in the process (i.e. before training).
gheinrich
force-pushed
the
dev/caffe-net-sanity-check
branch
from
0c592d5
to
d3cbdca
Compare
|
I have updated the patch to make the new method static and I have added tests. |
Nice! |
lukeyeager
added a commit
that referenced
this pull request
Feb 12, 2016
Drop error if a Caffe layer references an invalid bottom
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Outdated suggestions cannot be applied.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.
This should help address the case when a layer is added to manipulate e.g. the label blob. In that case the layer must include a directive like:
include {
phase: TRAIN
}
Otherwise Caffe will drop an error during inference.
A sanity check is added to DIGITS to catch these errors earlier in the process (i.e. before training).