issues
search
dilyabareeva
/
quanda
A toolkit for quantitative evaluation of data attribution methods.
https://quanda.readthedocs.io
MIT License
33
stars
0
forks
source link
Self-Influence kwargs
#95
Closed
dilyabareeva
closed
3 months ago
dilyabareeva
commented
3 months ago
remove kwargs from self-influence methods in explainers
decide on the batch size: either add it as an argument to every self-influence methods and ignore it sometimes
or remove batch size everywhere and just decide on some hard-coded value in case it is needed
maybe consider adding batch size as an explainer explainer init kwarg