Fix crash on map 1000 #36
Merged
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
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Outdated suggestions cannot be applied.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 fixes the known bug
#1
AlexeyAB/darknet#8669
AlexeyAB/darknet#8766
I could reproduce it with current master on first try
This fix is not perfect since we loose some performance by not using CUDNN_CONVOLUTION_FWD_ALGO_IMPLICIT_PRECOMP_GEMM. However in our tests the difference during training was very low. (And if the last ms matters in production intereference on gpu ppl would use fp16 optimized freamworks like opencv anyway)
The mid/long term fix will be to improve the cudnn <-> darknet interface but the current status is filled with so many warnings that it seems wasted to search for this one line when there are probably many bugs in this part of the code.
Tested with current cuda and cudnn version.