search

aleximmer / Laplace

Laplace approximations for Deep Learning.
https://aleximmer.github.io/Laplace
MIT License
458 stars 71 forks source link

Running metrics #135

Closed wiseodd closed 1 year ago

wiseodd commented 1 year ago

Support running metrics in cross validation of prior precision. Rationale: gathering data then computing metric is very expensive if n_data is very large.

There are some failed tests for LLLaplace, but I think they are unrelated.

wiseodd commented 1 year ago

@runame please check again!

runame commented 1 year ago

LGTM, have you checked the tests?

wiseodd commented 1 year ago

Updated the test! @runame