You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
once we add the notion of Filter and Scorer to EPP, I envisage that orgs that use GIE will have their own filters and scorers in some internal repo.
while GIE is not expected to support dynamic changes of filters/scorers, I believe we do want to allow an easy way to setup different filters/scorers in a fork. That is, I expect the filters/scorers to get registered in the code in a single place (e.g., single file), such that others can change their filters/scorers without drifting from upstream.
With that in mind, the main logic in a scorer (and/or filter) is usually based on some metric(s). so basically if one wants to define his own scorer, he might need to define also additional metric(s).
same as I expect that filters/scorers setup can be easily replaced without drifting from upstream, I think we should allow a very easy, simple and straight forward way to define the set of metrics (that eventually filter/scorer will use) that EPP collects.
The text was updated successfully, but these errors were encountered:
once we add the notion of
Filter
andScorer
to EPP, I envisage that orgs that use GIE will have their own filters and scorers in some internal repo.while GIE is not expected to support dynamic changes of filters/scorers, I believe we do want to allow an easy way to setup different filters/scorers in a fork. That is, I expect the filters/scorers to get registered in the code in a single place (e.g., single file), such that others can change their filters/scorers without drifting from upstream.
With that in mind, the main logic in a scorer (and/or filter) is usually based on some metric(s). so basically if one wants to define his own scorer, he might need to define also additional metric(s).
same as I expect that filters/scorers setup can be easily replaced without drifting from upstream, I think we should allow a very easy, simple and straight forward way to define the set of metrics (that eventually filter/scorer will use) that EPP collects.
The text was updated successfully, but these errors were encountered: