Fix/loss ekfac - Githubissues

aai-institute / pyDVL

pyDVL is a library of stable implementations of algorithms for data valuation and influence function computation

https://pydvl.org

GNU Lesser General Public License v3.0

89 stars 9 forks source link

Closed schroedk closed 6 months ago

schroedk commented 6 months ago

Fix loss usage in Ekfac implementation, different loss for gradient computation (cross-entropy) and Hessian computation (empirical cross-entropy)

[ ] Wrote Unit tests (if necessary)
[ ] Updated Documentation (if necessary)
[ ] Updated Changelog
[ ] If notebooks were added/changed, added boilerplate cells are tagged with "tags": ["hide"] or "tags": ["hide-input"]