search

dilyabareeva / quanda

A toolkit for quantitative evaluation of data attribution methods.
https://quanda.readthedocs.io
MIT License
33 stars 0 forks source link

Minor fixes #153

Closed aski02 closed 2 months ago

aski02 commented 2 months ago

Changes

  1. Readded captum_tracincp_fast_rand_proj_self_influence to init (so that it can be imported)
  2. Added test for captum_tracincp_fast_rand_proj_self_influence
  3. Made model_id and cache_id optional for captum_tracincp_fast_rand_proj_self_influence (since it is also optional for all other TracInCP methods)
  4. Updated docstrings for captum_tracincp_fast_rand_proj_self_influence accordingly