search

KGrewal1 / optimisers

A collection of optimisers for use with candle
MIT License
26 stars 2 forks source link

Lbfgs dev #4

Closed KGrewal1 closed 7 months ago

KGrewal1 commented 7 months ago

Adds in an implementation of LBFGS: this is not equivalent to the pytorch implementation as each step is one change (as opposed to torch where it is up to max_iter). This is functional but can still be optimised more.

codecov-commenter commented 7 months ago

Codecov Report

Attention: 47 lines in your changes are missing coverage. Please review.

Comparison is base (ebcbd32) 99.18% compared to head (7f7ca1a) 94.81%.

Files Patch % Lines
src/lbfgs.rs 83.33% 24 Missing :warning:
src/lbfgs/strong_wolfe.rs 86.11% 20 Missing :warning:
src/lib.rs 0.00% 3 Missing :warning:

:exclamation: Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files ```diff @@ Coverage Diff @@ ## master #4 +/- ## ========================================== - Coverage 99.18% 94.81% -4.37% ========================================== Files 8 11 +3 Lines 732 1023 +291 ========================================== + Hits 726 970 +244 - Misses 6 53 +47 ```

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.