lucataco / kosmos-2
Grounding Multimodal Large Language Models to the World
Run time and cost
This model runs on Nvidia A40 GPU hardware. Predictions typically complete within 12 seconds. The predict time for this model varies significantly based on the inputs.