2 comments

[ 29.6 ms ] story [ 68.1 ms ] thread
This is really great and fills a gap in current search.

Is this going to be open source? Any more tech details on how you trained it and how you would manage additional datasets? (I asked a similar question on Twitter.)