alt
Hacker News
satvikpendem
•
today at 3:14 AM
•
0 replies
•
view on HN
It's a little too late for that, all the models train on prompts and responses.