Omnivision-968M: Vision Language Model with 9x Tokens Reduction for Edge Devices

(nexa.ai)

62 points | by BUFU 14 hours ago ago

10 comments