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
________________________________
From: Marco Diomedi ***@***.***>
Sent: Thursday, May 30, 2024 6:10:10 PM
To: marcotcr/lime ***@***.***>
Cc: Subscribed ***@***.***>
Subject: [marcotcr/lime] Question: using LIME for text generation (Issue #747)
Good evening,
is it possible to use LIME for text generation tasks? Or only for classification?
Thanks
—
Reply to this email directly, view it on GitHub<#747>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/APH5BTM5MLXFMXL43KJP5RDZE5MPFAVCNFSM6AAAAABIRJ7G3WVHI2DSMVQWIX3LMV43ASLTON2WKOZSGMZDMMJVG42DCMY>.
You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>
Most text generation is carried via classification of tokens, with a probability score for each token based on how likely it is to be the correct token at that position. You should be able to hence use LIME just as you would use it normally, and get a feature importance score for every token generated
Good evening,
is it possible to use LIME for text generation tasks? Or only for classification?
Thanks
The text was updated successfully, but these errors were encountered: