Hacker News
new
past
comments
ask
show
jobs
points
by
mikodin
3 hours ago
|
comments
by
simcop2387
2 hours ago
|
[-]
Probably Kev and/or the decider models. Kev is trained on one of the 4B qwen models, similar for decider but it ranges from 0.8B through to the 35B-A3B model so far I believe.
reply