1 comment

[ 0.24 ms ] story [ 14.3 ms ] thread
We trained a frame captioning model that's 2x faster and 17x cheaper than Claude 4 Sonnet