Y
Hacker News
new
|
ask
|
show
|
jobs
by
davely
313 days ago
Not entirely unexpected given that these models have 2x - 5x more parameters than the 120b gpt-oss model.
It’s impressive their performance gets so close given the size!