FSRS7 + ADR + Optimization#426
Open
JSchoreels wants to merge 47 commits into
Open
Conversation
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
…se, added penalty flag
…FSRS7 penalty/training integration and fractional-interval plumbing, while preserving legacy FSRS6 compatibility semantics.
… longer will be rounded anyway in anki
This reverts commit a334c8a.
Contributor
|
@L-M-Sherlock the final Rust version of FSRS-7 will be ready today or tomorrow. It will be final for real for real. I have squeezed everything I possibly could from FSRS-7, both accuracy-wise and speed-wise. I'm just doing some final hyperparameter and default parameter tuning now. |
Contributor
|
The final FSRS-7 version is done: https://github.com/Expertium/fsrs-rs-speed-autoresearch |
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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 build on top of #395 and #412.
It mainly use @Expertium performance points, with a specific implementation of NEON for macos, and also improve by 15-20% ADR time computation.