Move root_net_ check in net constructor #4806
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
|
Heads-up: the |
|
That's a large PR, may take a while. It may make sense merging this one first. Similar complaints here: #4851 |
|
This should be unnecessary with the merge of #4563, but thank you for proposing a fix. @junshi15 and @SIshijima could you check that #4563 in fact fixes the issue? |
|
@shelhamer I won't have time to verify the recurrent layer at this moment as I am quite busy at work. Let me close this PR for now. If we see problems in the future, we will re-visit it. |
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Outdated suggestions cannot be applied.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 PR extends recurrent_layer to the multi-gpu settings.
Currently, the recurrent_layer constructs a net internally (the unrolled net), https://github.com/BVLC/caffe/blob/master/src/caffe/layers/recurrent_layer.cpp#L108.
No root_net_ is set in that constructor. In a multi-gpu setting, worker solver will create an unrolled net too, however it will fail at the line below, https://github.com/BVLC/caffe/blob/master/src/caffe/net.cpp#L50-L51, because the worker solver is not a root solver and the net does not have a root_net.
Since root_net is only used in share_from_root, this PR moves the check to where share_from_root is defined.