Y
Hacker News
new
|
ask
|
show
|
jobs
by
GaggiX
871 days ago
Well for LLaVa-1.6 we know it was trained only on English multi-modal data.