Hacker News new | ask | show | jobs
by trashcan2137 79 days ago

  and the EOS is "<turn|>". "<|channel>thought\n" is also used for the thinking trace!
Can someone explain this to me? Why is this faux-XML important here?
2 comments

That’s how the model is trained to signal the end to its generation and to indicate its thinking.
These are likely individual tokens. They are super common.