>1B tokens/minute/GPU by combining query planner and inference engine

(modal.com)

1 points | by birdculture 8 hours ago ago