CHOOSEDevtools & code
Data on how Jev @typesafeai has performed against an existing classifier.
Dushyant Garg@gargdush
Data on how Jev @typesafeai has performed against an existing classifier.
Its lighting fast, this is great but our comparison sorta fails cause our models also give us free text output that is crucial.
Cheap, but then again cant say for certain until free text poisons the mix.
Fairly precise, I think with more tweaking I can get it to be more precises but I noticed that the accuracy for 1 question vs 4 questions stayed fairly similar