Fix for NaN's when training DASM with ambiguous sequences #93
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
willdumm
commented
Dec 10, 2024
| @@ -195,11 +196,10 @@ def loss_of_batch(self, batch): | |||
| # logit space, so we are set up for using the cross entropy loss. | |||
| # However we have to mask out the sites that are not substituted, i.e. | |||
| # the sites for which aa_subs_indicator is 0. | |||
| subs_mask = aa_subs_indicator == 1 | |||
| subs_mask = (aa_subs_indicator == 1) & mask | |||
There was a problem hiding this comment.
This is it, the whole Nan/inf fix! Could you have a look around here and make sure this looks reasonable to you, too?
There was a problem hiding this comment.
Nice sleuthing! It seems good to me... am I missing something?
There was a problem hiding this comment.
I don't think so, just wanted extra eyes on it! Thanks
matsen
approved these changes
Dec 10, 2024
| @@ -195,11 +196,10 @@ def loss_of_batch(self, batch): | |||
| # logit space, so we are set up for using the cross entropy loss. | |||
| # However we have to mask out the sites that are not substituted, i.e. | |||
| # the sites for which aa_subs_indicator is 0. | |||
| subs_mask = aa_subs_indicator == 1 | |||
| subs_mask = (aa_subs_indicator == 1) & mask | |||
There was a problem hiding this comment.
Nice sleuthing! It seems good to me... am I missing something?
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Outdated suggestions cannot be applied.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.
The fix is one line, applying a mask that should have been applied from the very beginning, but previously didn't matter much since the data didn't contain ambiguities.
Also adds parallelized neutral model application, moved to the cpu. This makes running some notebooks much faster, and also speeds up setup for training Snakemake runs.