logoalt Hacker News

SwellJoeyesterday at 10:10 PM3 repliesview on HN

Everybody builds one. And, then they usually figure out that making the model fill its context with a bunch of memories hurts performance more often than it helps.


Replies

esafakyesterday at 11:40 PM

That's why I always ask: got benchmarks?

show 1 reply
zaydmulaniyesterday at 11:19 PM

[dead]

kdkdkdjdksksntoday at 12:17 AM

[dead]