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Outdated suggestions cannot be applied.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 adds a speed optimization:
For longer texts, tokenization can take up 90-95% of our time, because we tokenize the entire text. However, we usually only take the first N (usually 512) tokens of text. So, it makes sense to only really tokenize the first X tokens of a text. We therefore truncate texts to N * median(token_lengths). A test on wikipedia shows that this doesn't lead to any unnecessary truncation, i.e., it never truncated to a length < 512 tokens.