Loss gen tokens #3677
Merged
Loss gen tokens #3677
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
mvpatel2000
reviewed
Oct 23, 2024
mvpatel2000
approved these changes
Oct 23, 2024
This was referenced Oct 24, 2024
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Outdated suggestions cannot be applied.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.
What does this PR do?
This PR allows differentiation between loss generating tokens and total tokens, by allowing
get_num_tokens_in_batchto return a dict withtotalandloss_generatingkeys. This will only have an impact if you are using theaccumulate_batch_on_tokensflag. This does not break or change any existing functionality, as by default, if you are just returning a single number fromget_num_tokens_in_batch, we will accumulate using that value. Things will only change if the user adjusts theirget_num_tokens_in_batchto be a dictionary.Related PR in LLM Foundry: mosaicml/llm-foundry#1610
Manual testing (using Foundry draft PR):
Before with all tokens loss generating, microbatching is deterministic:
Before with not all tokens loss generating, microbatching is not deterministic:
After with not all tokens loss generating, microbatching is deterministic:
