Merged
Conversation
This file contains hidden or 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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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 introduces a new bucketing strategy called 'Two-Bits Bucketing' to the
tokenizers/bucketpackage.Changes:
tokenizers/bucketpackage:ByTwoBitBucket(batchSize, minSentenceLength int): Configures the bucketizer to use buckets of sentence-length sized to the next value that can be represented with 2 bits (e.g., 1, 2, 3, 4, 6, 8, 12, 16, ...).ByTwoBitBucketBudget(tokensBudget, minSentenceLength int): Similar toByTwoBitBucket, but adjusts the batch size to fit within a fixed tokens budget.TwoBitBucketLen(unpaddedLen int): Helper function to calculate the smallest size >=unpaddedLenthat uses only the two highest bits.docs/CHANGELOG.mdto include the new strategy.tokenizers/bucket/bucket_test.go.About Two-Bits Bucketing:
This is a '2-bit semi-log bucketing' where each size is separated from the other by a factor of 1.5 or 1.333 alternatingly (averaging ~1.414 or sqrt(2)). This results in bucket sizes that are 'friendlier' for binary addressing and memory pages while minimizing padding by providing more granular bucket steps than pure powers of 2.