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
It would be very helpful if SpanQuestions could be used for ChatFields.
We’re currently using Argilla to audit a conversational chatbot. In this case, the user may ask multiple questions in an interaction, so we can’t simply use a pair of text fields. And there are several cases where we need span highlighting (to identify wrong answers, PII, etc.)
Currently, if a SpanQuestion is attached to a ChatField, the ChatField UI simply won’t appear in the annotation—users see the raw dictionary instead.
Our current workaround is to parse the chat conversation into a markdown text field, but that has obvious downsides, including more effort to parse the subsequent results.
The text was updated successfully, but these errors were encountered:
It would be very helpful if SpanQuestions could be used for ChatFields.
We’re currently using Argilla to audit a conversational chatbot. In this case, the user may ask multiple questions in an interaction, so we can’t simply use a pair of text fields. And there are several cases where we need span highlighting (to identify wrong answers, PII, etc.)
Currently, if a SpanQuestion is attached to a ChatField, the ChatField UI simply won’t appear in the annotation—users see the raw dictionary instead.
Our current workaround is to parse the chat conversation into a markdown text field, but that has obvious downsides, including more effort to parse the subsequent results.
The text was updated successfully, but these errors were encountered: