Speed up tagger loading: remove IndexMap, new -> with_capacity #66
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
|
This is very good news! 25% is already a noticeable improvement, sorry for dropping the ball on this :> |
|
No worries. As of release 0.6.2 you should see the speedup :) |
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Outdated suggestions cannot be applied.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.
Hey @drahnr I've had a go at speeding up loading the Tokenizer today.
I did two things:
IndexMapwithVec<(WordIdInt, PosIdInt)>as discussed in Improve loading speed (of regex?) - cli usecase #56. This makes the most difference.new->with_capacityby storing the lengths, this makes a very small but measurable difference (a couple %).Overall I get a 25% speedup, which is something at least. I experimented a bit with parallelization, particularly setting some "anchor" points in the FST and splitting the work in chunks where each chunk iterators from one anchor point to the next, but it seems the speedup from that is nullified by the merge we have to do afterwards.
Maybe there's some more smarter ways to further speed this up, but I couldn't think of anything.